ABM: How to Define Your Ideal Customer Profile (ICP)

Cool, so thanks everyone for being here. This is meetup number 2 for 2019 for the Demand Generation Meetup Group. So we are recording the meetup, you're gonna see the video in a few days with the slides. So we always share this content so you can review it, take a look, or share it with So this meetup today is focused on building the ideal customer profile and the target account list. So there's gonna be a lot of very good content, different angles on this topic. We are already setting up the next meetup, which is gonna be about a very good topic, which is CAC, or customer acquisition cost analysis, understanding that. That's the T, TBG. So for today, I'm your host. My name is Franco. I've been running this group now for 3 years, but we are starting only since last year that we are scaling this and really running events. I'm the founder of a company called SaaS MQL, which where we help SaaS companies with demand generation and account-based marketing. To generate opportunities from target accounts. So today, this is the agenda. The first speaker is going to be Colin Funnestuehr, who is the CEO of Predictable Revenue. He's going to talk about how you can nail a niche. Very important topic at the beginning when you're getting started. And then, Gail Moran from Mintigo, she's going to talk about how you can automate the creation of the ideal customer profile using AI. And then Gil from Metadata, the CEO of Metadata, is going to talk to us about how you can activate those target accounts, again using AI and also automation. And then I'll close the meetup talking about how you can track this progress on your target accounts using different platforms, different approaches. And then we'll have some Q&A at the end. So our sponsors, as you very well know by now, Metadata has been our kind sponsor now since last year. Um, Gil, you want to say a couple of words or you want to leave it to the end? And then our host today, Upstage, great location. Thanks again for hosting. Um, really, really good space. So thank you again for both sponsors. So my usual page about people hiring on demand gen. So I received these two requests, one company called Sidetracker looking for a demand marketing manager, and a company called Tanium looking for a marketing automation manager. And, uh, here another person hiring at Influx Data, right?

Influx Data, and it's headquartered in San Francisco, San Marcos Airport.

So all looking for demand gen people in San Francisco. So if you have, uh, if you're interested in any of this role, contact me or go directly here. I can direct, uh, direct you to the right person.

Do they have to be, um, on-site or can they work distantly?

In these roles?

Yeah.

Uh, that I don't know. I know that's a good question. I know some of them might be open for remote work, but usually they're looking like when they have a location, normally they're looking for a logo.

I have a company called SaaS in QL.

Okay, so let me introduce the first speaker, who is Colin Stewart, CEO of a company called Predictable Revenue. Please welcome Colin.

So I think I've got like the, the for dummies version of the talks because I think everybody else after me is getting progressively more advanced and more advanced. We go from— this is the easy and sort of analog version, and I think Gail's gonna— she's gonna one-up me, I'm sure. So I'm looking forward to hearing your talk. My name is Colin. I'm one of the co-founders of Predictable Revenue. I'm gonna basically talk about know your niche. It's a term— so my co-founder's name is Aaron Ross.

He wrote a book called Predictable Revenue.

Second book was called Impossible to Inevitable. He wrote that with Jason Jason Lemkin from SaaSter. And we talk about nailing a niche, it's basically a combined persona and ideal customer profile exercise. And I think it's one of the things that the customers that we've seen, that we've worked with, you know, no matter what size or scale, just, they just haven't done it, they haven't put the work in. And so I'm gonna share an exercise that we typically run them through, and you can do this in, you know, an hour and a half, or you can do it in a couple of day session. I'll sort of show the, you know, the quick and the longer versions. I want to keep it 20 minutes, and I can probably talk for a couple hours, so I'll try and keep it short and leave some time for Q&A at the end. Basically, you know, with flying, I'll try and convince you that you need to nail it. I'm going to talk about, you know, what do you do if you— how to figure out if you haven't nailed it, and then basically how to find it. So why should you listen to me? I spent 15 years in sales, mostly in a closing role, but you know, before Predictable Revenue, I was doing lots of account exec, you know, the full-stack sales rep, you know, that was pretty popular 6 to 10 years ago. I founded 3 companies, and this is the first one that didn't explode. My first one was a CRM company that had hockey stick growth, but it was facing the wrong way. I'm Canadian, so I had to get a hockey joke in there.

Yeah.

The second one, we went from $0 to $40K monthly recurring in about 6 weeks, and then it fucking exploded. Engineering literally caught on fire, and it was like a sales enablement platform, and the whole goal was sending emails, and we couldn't send emails. Like, sending a single email was taking like 2 days. And our customers were like, "Well, this sucks." But the interface is great.

But it sucked.

It failed at doing the one thing that it did. Our customers hired us. So I failed a whole bunch and I'm on my third and hopefully this one will take longer to fail and I'll learn new and different things.

So that was the "why we should listen to you" part.

I think maybe, yeah, let's go with that. I've made a lot of mistakes and so I've learned through those. But I think the moment that's relevant is the one that I didn't fuck up or I haven't fucked up yet and that's Predictable Revenue. And we've helped hundreds of companies build outbound sales teams. And I'm going to share some of the things that we've noticed. Just so many of them don't go through the— they'll go through the early customer development and product exercises. They'll map out and they'll find the sort of that gap in the market, right? And you'll do such a great job, or our customers have done such a great job communicating that from the product team to the engineering team, building the product, and then all of a sudden it's like, okay, why Marketing, sales, let's go. It's on you. And I think the real key is that there's so much of this research and so much of this learning that's already been done as a company progresses from the core team of 5 or 10 to 20, 30, 50, 100. There's just so much that doesn't get communicated, so much that doesn't get documented. And when you're at that scale, it can really cause your growth to slow down because nobody knows. If you have the best SDR team in the world, And hell, if they're focusing on the wrong target accounts, are they going to get anywhere? Probably not. And so it really comes down to market, message, mix, however way you want to slice it. The right message to the right people. I find a lot of the exercises when you get into, you know, marketing roles, and I started my life in marketing and then I moved into sales, so no slant against them, I'm guilty of it as well, but most of the— I don't understand any I did a Google search for Rain Man and I found this and I was like, yeah, that's better. But most, most of the total addressable market or ICP or persona exercises, they're theoretical, right? They're these like avatars and we've got this cool picture and there's all these things. And Wayne's like, yep, I've done 10 of those today, I know what I'm talking about. And I think the thing that from the sales side that I'm always struggle with is that you get these like great documents that describe what the person likes for dinner, how they like to comb their hair, their favorite sports teams, but you don't actually get a list of these people. It's like, all right, salespeople, go figure it out. Like, how do you do that? I don't know, use Google, use LinkedIn. You're like, well, fucking great. And so one of the things that I hope I can share with you that this workshop has helped us do, that I'm going to show you how to do, is help sort of take that step from the the theory into the practice. And so basically what we see is our customers kind of go, before they do this, you're going all over the place. You're, you know, you're booking a ton of meetings, but you're not closing anything. You're, you know, a lot of your probably initial revenue is possibly coming from the founder and the founder's network. You're great at a lot of things, but you're not sort of world-class at one thing. And it kind of turns into this, right? As we come back to the marketing, we say, okay, these are the things that we want to say. This is our language, and we're not really speaking our customer's language. And so I don't know if you like Jim, uh, I'm totally drawing a blank.

Larson.

Anyway, Gary Larson. Gary Larson.

Gary Larson.

Speaking to jeans, Jay. Um, yeah, one of my favorites. Um, and basically And this is what we end up saying when we just take the words that marketing gives us and we start putting those into sales emails, which is— I see happen so much, but it's really dangerous. And I think the way that I try and explain it to people on the marketing side is if you think about your SEO specialist or your conversion rate optimization people, they're not writing content that's brand cool. You obviously want them to communicate the right message, you want them to get the general idea across, across, but they're gonna tweak and test and tune things because their job is to start a conversation. And I want you to take the same approach and apply the same lens when you're thinking about sort of writing your core messaging when it comes to sales. Obviously, through Predictable Revenue, we built an outbound sales team, so a lot of what I'm gonna share is sort of through the lens of writing outbound sales messaging, not just for emails, emails, calls, LinkedIn, however you And one of the things we've noticed that's really helpful is instead of going super broad and trying to be small fish in a big pond, is make the pond small and be it, you know, you'd be the same size fish, but it's so much easier to just focus on a really small market and have a really intimate conversation with just one person as opposed to trying to have a very morata general conversation with 300 people. $300,000. One of the things that's helped me quite a bit sort of on my journey as an entrepreneur, as a salesperson, as a sales leader, is Jobs to Be Done theory and is learning actually more about how to be a product manager. And one of the, one of the interesting things that I'm not even close to, I don't, I hardly know anything about, um, but it's had a big impact on our team on how messaging, on how we talk to customers, is the job to be done theory. And I'm not going to go through and explain it because I wouldn't do a very good job, but if you think about it, the way I think about it is—

That's it.

There you go.

It's the job to be done.

That's job to be done.

Yeah.

It's the progress our customers want to make.

It's what's the job to be done. It's as simple as that. It's the end result. I need a quarter-inch hole. That's it.

Exactly. I knew if I put that out there I wouldn't have to say it and somebody would read it for me. And, you know, I remember hearing about this when I was like, or not, I remember hearing this quote when I first started off in sales and it really stuck with me and then it's been one of those ones that I think you read any blog post or book about the Jobs-to-Be-Done Theory and you'll see that quote. So anyway, that's, you don't have to learn no Jobs-To-Be-Done theory in order to get used to— for this workshop to be useful. If you do want to dig more and if you do want to sort of take it to the next level and do the longer version, I can recommend a couple of books. But reading about Jobs-To-Be-Done theory is super helpful. Google, or if you go to YouTube, it's Milkshake and Jobs-To-Be-Done. Clayton Christensen's got a great video. I'm sure most of you have seen it. So getting to the workshop. And Franco, how am I on time?

Good.

You got another 10 minutes?

10 minutes?

Perfect.

So basically, this is a workshop we did for— I can't remember the company. But basically, what we're going to do is we're going to lay out this matrix. And Franco's going to email out the slides, and I've got a workbook that I'll share. It's actually a workbook that we'll be running for our customers. So you can feel free to take pictures, but I've got an email address at the end that you can email me, and it'll automatically send it to you. Or you can email Franco. And he'll share it with you. But really the idea is you want to get the right people in the room and then you go through this workshop. And basically there's two different— you can invest any amount of time that you want in the workshop. The sort of bare minimum is an hour and a half just for the workshop itself. Not because it's arduous, but because people tend to talk. And I'm going to recommend actually bringing in customers and salespeople. And when you get salespeople in the room, we tend to like to talk, right? So the right people to have in the room for this is sort of the founding team, right? Guys like Gil, your product team, right? The people that actually found that sort of gap in the market. Your marketing team, because they're obviously translating sort of the product features and product, you know, the underserved needs into your ideal, your target accounts list. And then you also need your sales leader. And most importantly, maybe I'm biased because I came from sales, but you've got to have the salespeople in there, right? Because the founder understands, the founding team and that core product team, they understand the vision. They understand where they're taking the product. And then on the flip side, you've got the sales team that understands what the customers are actually buying, the words that they're actually saying. And I find in, in these exercises, those are too, those voices are too often under her. And so they can be really, they can bring really great insights. Obviously don't pack your room with salespeople because, you know, you don't want to just, this can easily take 3 or 4 hours and I've had these workshops stretch on and you're just ready, you're like, I'd rather be at the fucking dentist, I need to end this meeting right away. So I'll go through some of the categories because this is barely legible. Really there are 6 sort of keys key columns. And the way we're going to do this is we're actually going to start on the left side. We're going to start with column 1, and you're just going to list out all the use cases. So you're going to go from top to bottom, and each use case or each job to be done gets its own column.

Fair?

And this is just because we're gonna later on in the exercise, as we progress and we go through the other columns, you're going to want to tick off and say Oh, this one also does that. And then I've got a spreadsheet that you can translate afterwards. So it's typically helpful if you've got someone who's really detailed taking notes or doing so afterwards. So basically we're starting with the use case or the job to be done. Then we translate and then the next step. And so you go down the column, you list out all the use cases. And then what are you going to do? We're going to go horizontally. Yeah, I got that right. I get a little dyslexic horizontally sometimes. With the rest of these. So we start with use case here, and then we're gonna go across popular points of pain, etc., etc. So popular points of pain just means these are the common pains that our customers are talking about. And this is great to have the salespeople and the marketing people in the room because they're hearing these things. They've got the copy on the ads, they understand what's converting on the marketing side, on the sales side. If you bring your SDRs in, they understand what email templates are converting, the salespeople people understand the, you know, what their customers are saying when they're doing discovery and they're finding those hot button issues. And so, you know, I'm sure there's— typically when we do these, customers have lots of this documented. It's just a matter of getting it all fleshed out and putting it into the matrix. So popular points of pain, common pain, you know, the common pain points that we're hearing. The believable solution is the English version of what your product does. No buzzwords. This is— you have to turn around and explain it to a 4-year-old, and they have to understand, right? So we're not allowed to say synergy or anything like that. It's got to be a plain English description that our moms can understand. Tangible results. This is where I get to be Dr. Asshole and really push back, and I encourage you— this is, you know, this is the one place to really, really, really push back because we want to see metrics, but we don't want to just see and hear metrics, "Oh, we did this for this." We want to hear customer service. And this, especially if you have the right people in the room, can be one of the best columns because this is writing. This is where your best messages can come from. If you go back to the Far Side comic, it moves us from speaking sort of human into speaking dog. Not that our customers are dogs, but you get the analogy. So with tangible results, and this is a again where it's helpful to have somebody spreadsheeting. We're trying to frame it up in customer stories. So customer X experienced Y, or we did this for customer X, they experienced Y result, and it produced this, this much ROI over that period of time. It's going to vary based on sort of your customer base, the product, everything you do. Last piece is the persona— or not last piece, second last piece— persona ICP combination. And so there's a lot of different ways of sort of framing up ICP, persona. I'm a big fan of doing— to go that one level deeper and just not do a general, okay, for this one ICP, it's for this ICP, and then within that ICP it's this specific persona. And then I'll show— I'll give you a spreadsheet so you can map out all those different combinations and then two popular ones that are related. And the value of putting into a spreadsheet later is that you can now sort, you can look for groups, You can sort of— you can write things for specific personas and not necessarily reuse them, but it's not as much work to translate it to the next one if you arrange them properly. And then the last one is unique genius, which is my co-founder's fun way of saying you're a differentiator. He likes making up his own words, "nani nishi," genius, and I just sort of took the compliment. And so it's basically, you know, in addition to sort of the tangible results you've created, it's how you separate yourself from your competitors. This is a look at the spreadsheet. You can see we've got the pains and the how we do it and the proof and all that. And then you've got your personas. You know, back in the day we were doing personas, but I do PCP, ICP combinations now. And then you just tick them off here. And then this sheet enables you to sort by the combinations and see where they overlap and see where they gaps are. Because the best thing that you can do is have those specific conversations, and sometimes there's quite a bit of overlap and it does make sense. You might have one persona ICP combination that only has 10 customers, and it makes sense to group it and try and get up to 50. Yeah, actually, to that end, in that particular case, are those personas all part of the same buying group, or in this case are you actually targeting different individuals, like in a complex sale with multiple— Yeah, when we get into like building a sales strategy for this, it totally depends on like where you are on the ABM pyramid, right? On the top end, if it's a strategic account, then, you know, each, each account is going to get like a personalized marketing campaign for it. If it's the sort of next layer down, I think they call it ABM light, basically you're going to be first personalizing personalizing every single touch for each individual. And then below that is scaled, so it's a little bit less. So it kind of depends. There's no— like, it depends on the, on the cash and what it makes sense to, you know, invest. And I think that's Franco's next meetup, so definitely meet him for that. Sorry for jumping ahead.

How dare you! Don't let it happen again. My bad.

Yeah, and so that's the— this is the sheet we basically used to group We look for common trends. You know, sometimes there's overlap. Sometimes we got a bit too aggressive splitting up our personas and ICPs. Sometimes we didn't get aggressive enough. Helpful to have it here so you can basically pull it off the whiteboard, because if you don't do this, all the information is going to die and then you've lost it. It'll stay on a whiteboard or in a picture on your phone of the whiteboard and nothing's going to happen. So it needs to go into the spreadsheet. And then basically here's the— where the rubber meets the road. For each persona ICP combination that you decide is somebody we want to target, interview 3 customers. So this is the long version, interview 3 customers. You can do more, you can do fewer depending. And then you need to translate it into realistic contact and account targeting criteria. None of this left-handed shop owners nonsense. These need to be attributes that we can find in a database or things that our salesperson phone, so we've done research. The number of personas, and I'm sure you've never done this, or ideal customer profiles that we've received, and it's all these criteria about, you know, they must love clocks that are analog instead of digital, and they have to be a Mac user, and they have to have like X number of this, and you're like, how the fuck am I going to pull that out of the database? It's never going to happen.

I can't build that.

And so they have to be realistic targeting criteria, and this is another place where you have to be Dr. Asshole. Because if you get a little bit soft here, the exercise just stays in the theoretical, and it's this sort of piece that's bringing us into the practical. So the next step is translating that into targeting criteria and then using, you know, the typical sources, you know, ZoomInfo if you're general, DiscoverOrg if it's more engineering-focused, and there's a bunch of others. There's even some good niche ones. You know, we've used LeadIQ, we've used custom-built things. There are good and bad list providers. It's more about how you use them. I've heard great things and terrible things about every data provider I've worked with. It's just a matter of are you using the data correctly? And if you don't do, if you don't follow this process, you're gonna buy data and be like, "That data provider is shit." If you do use the process and you do it well, that same data provider can produce an excellent result. Then the last step is where you, randomly, are messing up. Come on, that's not some racist. Sorry, I'm from Vancouver and we're in a— not a rough area, but sort of like your version of the Tenderloin. We're like right on the edge, and so there's constantly fire trucks and ambulances and all sorts of what you would also see in the Tenderloin happening right up behind our office. And I have a podcast that's right on the line, so I'm used to it, let's just say. And so the last step here is just writing a core message, all right? And that doesn't have to be it doesn't have to be a cold call script, it just has to be a couple of bullet points, a summary of that persona ICP. And just take it out of the format, it can be freeform, but you just want to communicate the core idea to whoever it is that you need to share. So recap, I think there's a question.

Oh, sorry. So here persona and ICP, you're saying that the customer is the human. My confusion is, serious decisions, are they also sort of saying an ICP is an organization? So like a target— I target companies are ideal customers.

Yeah, so the way I, I'm— I guess the, the way I'm using the words ideal customer profile is sort of the how you categorize your customers, the, your customer accounts. So these are the companies and how you group them. And then the personas is how you, how you group the sort of people and the job to titles within the house.

Right.

So it'd be, you know, B2B SaaS companies, right?

500 to 5,000 employees.

Yeah, exactly. All your standard firmographic data.

Yeah.

And then the persona is the individual. So the product manager or the person that manages the product.

Okay.

The person that manages the sales team. Those are the, you know, and you can get as detailed or as not detailed as you want. I'm not an expert in personas, but we do, we do a bit. Does that help?

I think so, yeah.

Cool.

I'm just giving the recap. So basically, hard to scale. Our job is to translate the theoretical into the practical. You got to get everybody in a room, including your salespeople and ideally your customers. You got to spreadsheet it out. And then the last step is writing that core messaging, sharing it out, getting it out in the world. And that's how you nail your niche.

Woo!

Thank you, Colin.

Thanks, buddy.

Thank you so much. Um, I recommend, if you haven't read it yet, Reasonable Revenue and From Impossible to Inevitable— two must-read books.

One more slide. If anybody wants the workbook, I can email Franco, or you can shoot me an email to saskatoon@rainforestmedia.com, and I'll share that, uh, the workbooks that we use internally.

Awesome.

Thank you, Paul. So we have some Q&A time also at the end for all the 3 speakers. So now next speaker is Gil Morag from Mintigo.

Hi, it's always 6 PM somewhere. So my name is Gil, and just a couple of words about me. I'm heading our partner group in Mintigo working very closely with metadata, with Gil and his team. And I'm also a mother of 3 wonderful daughters, and I'm finishing tomorrow my yoga certificate, so I'm going to be a yoga instructor in addition to a data geek in Mintigo. And what we do is basically we kind of— our mission is to be the brain on top of your CRM or your marketing automation or any CDP or data set that you have that you're using for marketing and sales, right? And really the question when you have your CRM and you have your marketing automation is, do I really target the right accounts there, right? Do we have the right people from the right accounts? How do I go beyond what I have in my CRM? How do I figure out what are these accounts that I'm not aware of even and try to target them through later on metadata or through any other strategy? And we actually strive to answer all these questions. And we strive to answer all these questions through kind of two different angles, right? It's the angle of the data. We have tons of data today. So how do we take that data and make sense of it? Through AI. So what we do in Intego— and I'll just kind of drop a little around it— We take the data that you have in your CRM, your marketing automation, we add all the data that we collect on B2B companies across the globe, and we collect millions of data points on B2B companies. We run all that through machine learning to really identify in a granular way what is your ideal customer profile and how you can find these accounts that are higher propensity to buy, basically. And how you can make your SDR organization target the right accounts and go beyond what you have. And how you can personalize the message. So this is in a nutshell what we do. And we do that within your systems, within your Salesforce, within your Marketo. We actually help you with the data that you need to make better decisions and get greater ROI out of every marketing or every sales campaign that you're running. And I like to compare it basically, and this is where I finish about MintyGo and go to why we do that. And I like to compare it to dating. Like, how do I figure out what is my ideal partner profile, right? So I'm here, I want to meet a really cool partner. And the first thing I ask, kind of the first question I'm asking is, who am I looking for? To meet? What is a good fit partner for me? And it's not only about the height or the weight of that person, which is really the firmographics in our world with B2B. It's about a lot more. And for me, like, are they— do they like hiking? Do they live in Fossa City where I live? So many other attributes that would really make a good partner for me. So the first question that I'm trying to answer: who is a good fit? Right? And after I answer that, it's really how do I find my good fit? And it's not only about my network because, you know, I tried my network before and I couldn't find my partner. So I have to go beyond my network. So how do I figure out all these people out there and partners out there that are a good fit for me beyond my network, beyond what I have in my own systems, I would say. But after I know who's a good fit and I was able to identify this total partner market, right, my total available partner market, the next question I'm asking is, who is in market to partner with me? So out of all those awesome people out there that are good fits because they fit my criteria, right, they're doing yoga, they like spiritual things, I don't know, all these things, um, but they might not be in market. So I have to look for these signals. Are they looking as well to find a partner like me? So this is where I go to figure out that, and this is also something that we can help with, and you really want to figure out when you go out and do your own marketing campaigns. And after I figure that out, and I have this really awesome group of partners I start dating, right? Or I start sending emails and messages. What should I say? What should I talk about? The more I know about that person, the better my conversation will be. So this is where, again, data can help me. So my way to figure out my best partner is really taking all those data points and using them to have a more meaningful conversation. And the more I do that, the likelier it would be for me to succeed in my meeting or in my— whatever date that I have. And kind of the first question, and if I take it back to our world, right, we are trying to figure out who is a good fit for us, who is my target account. And how do you guys do that today? Like, what's your way to do it today to figure out who's your right target?

Salespeople tell us.

Exactly. So one of the best ways, like, oh, I have it in my Rolodex, I know that person, I can call it, right? But is it really the right way to figure out your target accounts? Another question is, in what level do you know your ICP, right? And that's kind of— it shows how old I am, basically. I was using these computers like, I don't know, 10 years ago. And how well do you know your ICP? Do you know it like, okay, every company above 100 employees is a good company for me? Or do you have this granular view of who is your ICP, right? And these are kind of the two questions that, you know, we strive to answer and help with data and AI. And, you know, we use AI— like, it's a scary word, um, AI, right? Today it's like robots and stuff like that. But actually, in our day-to-day, we use AI for everything. Like, Netflix recommends my best you know, movies for me to see based on what I saw before and based on probably other criteria, right? So how do we do that in our world of B2B and where we can apply data and AI to allow us to really figure out who are our target accounts and our ideal customer profile? And we use that approach kind of in an ABM strategy, and I'm assume most of you today are already doing somewhat of an ABM strategy. So we use AI for a B2B kind of context in 4 different ways. The first is for the account selection, and the account selection based on a very granular understanding of what is your ideal customer profile. Another place where we use AI— and I lost my 2 and my 4— is in personalizing the message. Because the more you know about the account you're trying to target, the better you can talk to their pains, the better you can say, "Hey, this is why you need to use our solution," right? Because you have that pain. And we do that a lot, even in Mintigo, like if we see that you're showing in-market behavior for ABM, I know how to address you. If I see that you show in-market behavior for AI, I can talk with you differently, right? I can really personalize the message to the tool that you're using, you're using Marketo, you're using Eloqua, you're using HubSpot, right? I can really talk your language and do a better conversation with you. But after you have that and you have your target account and you have your personalized messages, the next thing that you want to do is really go ahead and execute that strategy. And this is where Methodata comes in, where they can really run all these accounts through ads and LinkedIn and start to get the results for you and apply AI there as well to decide the best channel for you and the best campaign and the best combination of content so you will get the best results. So the more you use this automation, the better your conversion rates and your results would be and your cost to acquire your ASP on the revenue side and your cost to acquire a lead would improve on both ends. So how do we do that basically, and where all that smart world comes from? So your first approach is really like the regular approach that you would use when you don't have that type of data, is kind of going everywhere. I will try all the companies that are above 100 employees. I will call them all and try to get results, right? But really where you want to get more data-driven to get the results that we're talking about is when you start to apply more interesting data such as growth patterns, such as what type of technologies are they using, such as what type of hiring are they doing, such as what type of intent, what type of research these companies are doing online. Are they now researching for are they researching now for an HR tool, right? What are they doing online? What are these data signals that you have, buying signals that you have, that you can use to optimize your campaigns? And of course, what is their behavior? Did they come to my website? Are they downloading the whitepaper? Are they seeing my ads? So all this, when you take all this data and you run that through machine learning, the results that you have is your ideal customer profile.

Profile?

And this is one of our customers that when we, you know, ask them that question, what is your ideal customer profile, the answer that we got is, well, every company above 100 employees with revenue above $1 million. And yes, they're using, you know, CRM systems and this kind of, you know, generic notion of digital transformation initiatives. So how can you translate that? How do you know which accounts really fit that thing, right? This is easy. The first one is easy. You can get that from everyone. And when they applied that criteria, they got 100K accounts in their, you know, target account list. You can't target 100K accounts. One, it's expensive. Two, you don't have enough people to target 100K accounts, right? So what we did here, we actually went ahead, applied their data in our system, and this was the kind of ideal customer profile that you see when you use a database or a dataset that we have in Integill. So you know that they're using AdRoll. That's good for metadata that they know that, because now they can go ahead and say, "We're way better than AdRoll, right? We can do way better than them, and this is why we can do that. You're using a competitive tool." The pay-per-click monthly spend is between $5K and $50K a month. That's awesome. That means that I can really spend, you know, the amount of money that I need for metadata, right? They are using web analytics tools like Optimizely. It's a little more expensive tool. We know they're interested to see who is, you know, who is visiting their website. They're showing intent, as an example, on Marketo. It's my competitor. That's good. I can go ahead and explain why not use Marketo, but use another tool. So this is the kind of granularity that you can get in your ICP, and this also translates that notion that your salespeople know, oh, they're doing some digital transformation initiatives, right? That can really translate it into something very, very tangible, and that's the type of data that you can get when you use bigger datasets of data and third-party data as part of building your ICP. And all that is Automatic, so it's using AI, you actually identify based on the data that you pour in, what are these signals, what are these insights that are relevant for your business? And what do you do with that when you know all that data, right? Now you know this granular ICP and wonderful, so how do you use that? You use that to personalize, you use that to identify very specific audiences. And why do you do that? Again, to get better conversions. The more you talk the conversation that your customer is expecting to talk, the better your conversion rate is. And we see that with our customers, right? They started— so that was the original ICP was 100K accounts. And after they ran it through Mintigo, they were able to identify 2,500 accounts that are A's. And these 2,500 accounts converted 6 times better. Than the list of the $100K. So instead of a 2% conversion rate, they had a 16% conversion rate. It was just, that's mind-blowing to see that. Because that's the power of when you're using data to really take that decision of your target account list. And since you're spending so much money on campaigns, if it's display campaigns, if it's a sales campaign, an SDR campaign, and we have an SDR company out there, right? You spend so much money You really want to target the right accounts and target them with the right message and doing the right personalization that you need and talk their language to get the better conversions. So this is what you can do when you know your ICP. You can talk about— if you know they're using a competitive technology, the example of AdRoll for metadata, you can very much target around it. Why we can scale your operations further, right? If you know they're hiring a demand gen person, here you go, or whatever solution you have. If they're hiring and growing the business, you have actually your persona that you're targeting, super. You can say this is the way to scale your operations. And if they're showing intent signals, that's a very strong signal for an SDR to call already, because it means that they're in market. If I go back to my dating example, they're not only a good fit for me, but they're also in market to buy a solution that I'm selling. So this is when that type of research they're doing online is really, really interesting. And after you do all that and you have your target account list, right, what you really want to do is target smartly. And this is where my friend Metadata, Gil, helps out through display ads, through LinkedIn campaigns, through all these type of campaigns to execute on top of your target account list and your ICP. Any questions? Are you familiar with the terminologies of intents? Do you— yeah? Any translation needed?

Okay.

Any questions?

Yeah.

Where do you get all your data, the third-party data?

Yeah, so this is data that we collect. We actually— Mintigo actually We started as a data company. We are all coming from the Israeli intelligence, and our founders really know how to collect data. And we have like tons of crawlers out there that go to the various websites, to PR websites, and so on and so forth. And we collect topics. So we know what type of technologies these companies are using. We know what type of hiring they're doing. Are they growing? Are they shrinking? Are they changing their CEO? "Who is the CEO now?" All these little data points we actually collect from scraping. And also the intent data, we actually have awesome data sources of raw data that comes from ad campaigns and ad data, and we actually aggregate that into what we call topics or signals of intent. So we know if someone is now interested, for example, from Mintigo, if someone is now researching predictive analytics or if they're researching online— not someone, but a company, an account, he's researching around predictive analytics, which is relevant for us, or researching around ABM, so then we have a talk track about it. So all that data, this is how we started. We gather really millions of data points on more than 25 million accounts across the globe. So one of the things that you can actually do is figure out accounts that are not in your database today. So one thing we can score your database, very simple. Tell you are you really targeting the right accounts today in your CRM, your marketing automation, right? But after we do that, we can come ahead and say, hey, out of the database that you have, 20% is an A, 20% is really good accounts that would convert well. And we can go ahead and score our database, our 25 million companies, and give you additional A's. And we can do that in a fraction of, I don't know, in minutes, right? SaaS platform. Like, we do that in minutes, which is an exercise that would take months to do it manually, right? To grab all that data, match it, clean it, figure out really this target account list. It would take, you know, months to do that in a manual way. So this is where we can really help you scale your operation and hone in on the right accounts.

I guess the more clients you have, the better the data you have.

No, we don't use— we actually don't use the customer data. We don't kind of cross-pollinate customer data at all. We use our data. Two reasons for that. We would have loved to use customer data, but two reasons for that. First of all is privacy, right? We don't use customer data for that. But also our data is curated for machine learning, so we really, really make sure that we can use the data to figure out your ICP, right? And we, we many times we use first-party data as well of our customers because they want build more sophisticated predictive models on our platform. But it requires kind of a different level of understanding of the data because first-party data tends to be very scrappy, very all over the place, while our data is like really— we made sure that it could be used for machine learning. That's kind of the idea around it.

I have two questions.

One, in In your example that you gave, what was the average contract value of the company that you—

50K, I would say.

50K or 15K?

50K.

Got it.

Like 50K. Yeah, for them, I would say. That would be their ASP.

And then for your intent data, are you— there are some kind of B2B providers out there that are selling, in essence, intent data as an independent standalone Are you purchasing those? Do you go to any other outside sources to purchase that intent data, or your intent data is totally proprietary?

Our intent data is proprietary. So we do buy the raw data, but we are the ones that curate that raw data to be useful for machine learning again and for the needs of our customers. Yeah.

How are you grading— How do you classify your ideal customer? Is it like a letter grade or is it like a score?

It's both. We actually allow— so when you know the ICP, right, when we figure out what are these insights that are defining your ideal customer profile, right, then we go ahead and score your database based on that, right? That's kind of the next step. And we can give a numeric score that is between 0 to 100, and we actually allow you— you can easily kind of select how to rank it. So you can select score above 85 would be an A. But you can decide it. So it's really actually— I love it. I love it that you can actually decide because that's— because if you want to target the A's, for example, or the A's and B's, you can decide because we can show you that how many accounts you want in that based on the scores. And move it, you have a way to move it and to define the exact amount that you have. SDR team can really only work around 500 accounts, right? You can actually define the A's to be the number that is, uh, this 500 accounts.

And second question, have your customers ever asked you to draw something more human understandable than a number?

So I think that's the beauty about, about us, right? Because it's not— we made a very big effort to make machine learning explainable. That's kind of— because when you get, you know, let's say it's not even the marketeers, the marketeers are easy, but you go to your sales rep and you say it's a 30, it's a 68, and then you look at it and say, no, it's not a 68, it's a 78. And, you know, it's like this conversation that you don't want to have, right? So what we do, we actually give you the insights. So it's kind of— it's not, you know, 100% explainable because it's still machine learning, but it's like in The 20/80 rule of Pareto, it's a good enough explanation. So for every, if your sales reps are using it, and we have plenty of SDR organizations that are using the work of the marketeer and using our data, you know what are these attributes for that specific account. So you know what intent they're showing, you know what are their fit indicators, what type of tools they're using, what type of hiring, you know all that. And your sales rep can see it's a 78 because of all these attributes. So we made it very, very important for us not to be a black box. So it's not a 68 black box or a 98 black box. We try as much as we can to explain the results, basically.

One last question. Gail, you gave an example a few slides back of a company that had certain technologies that was an indicator of something, and you've also talked about search behavior as an indicator of intent. I'm wondering about maybe something in the middle. Maybe there's a case that you've been involved with that you're most proud of because it's the most creative way of thinking about what you guys do where someone—

I mean, in B2B SaaS, most companies are looking to like change the way companies are doing business. So in a sense, it's an un—

like open a new market.

Exactly. So if you're already using other stuff, then yeah, I get that's pretty easy to show how we're better than that. But what if the company hasn't even begun doing that and they're— other than searching Is there something other than like we're searching for that solution?

So what we see is that it's, um, the beauty of it is the fact that it's a combination of indicators. It's something that typically humans can't, can't, you know, do, can't assess more than 3, 4 signals. The machines can assess 50 signals, right? Can decide that 50 signals, 100 signals are part of your ICP. So it typically would be a combination combination of signals, right? So I'm talking about AdRoll as an example, right? But if you see that the company has a very specific unique business model, which we can identify as well, they're subscription-based, and you see that they're using cloud technologies more than any other company, and you're seeing that they're doing and that and that, then they become a really high-value customer. And there is very creative ways that our customers are trying to open new markets. There are tons of creative ways to build the ICP. So what you would do, you would build the success, the list of your successes, based on taking the list of one of your competitors, which is a new company, right? So you'd say, OK, I'll take their customers, I'll forward it as mine. Let's see what's the ICP for them. For that, what are the results, so I could target that. I've seen customers doing even more interesting models like churn models. Who's going to churn, right? And they were collecting using our platform. It's Getty Images. And Getty Images, like every company on Earth uses Getty, right? Like there is none— they're an imaging company, like, and you use it in every presentation. Everyone uses Getty Images. But there's also competitors. So they want to make sure how they keep their share of wallets, right? So what we built is actually, are these companies using other technologies, other images that are not Getty Images, right? So we have that data that was collected for Getty, and using also transactional data that they have, we were able to build a really, really interesting model for them. And they called it the win-back model. How do we win back these customers and we get back share of wallet here? So there is tons of creative ways. The moment you start to work with data, and again, I'm a data geek, so coming from that, the moment you start to use data, it's when it becomes really interesting, because you get this layer of understanding of your customer that you don't have without that. And the moment you have this layer of understanding, This is where Colleen can come in and say, "Oh my gosh, here, these are the audiences. This is the strategy. This is how we need to go after it." Now the execution, you bring in metadata, you do the execution. This is where the kind of the light bulbs start to— and every time you present the ICP to a customer, the light bulbs, you see the light bulbs going up and down. "Wow, that's really interesting. Wow, I can do something with it. I can really personalize that." now. And as you know, I had a conversation with Colleen, she asked me, "Do you use it yourself?" And I'm again, I'm just partners doing partnerships, right? I'm not supposed to use that type of data, but every time I run a webinar with a partner, I actually go ahead and score the people that came to the webinar. Was it a good webinar? Did I get my A's and my B's? Did I target the right people? Should I continue to work with that partner? Should I continue to invest? So for me, like that's my little light bulbs to make decisions, right? And again, I do that in fractions of minutes. It's not that I need like 3 data scientists behind me to do it. They just upload it to our own Mintigo models and I get the results in no time. So it's really fun to actually do these exercises. I really like it.

I think you're out of time for this.

Oh, I love talking.

Thank you very much.

We will have time at the end for more questions. And now next speaker is Gil from Metadata, and he's going to talk to us about how you can activate these accounts and execute on all this data after you have selected your ICP and you are ready to go. How can you personalize the message and really engage with them and basically generate opportunities from these accounts. So please welcome Gil. Thank you, Franco.

Gallic was awesome. We like to work with Mendigo for one reason, one major reason, is that their data works. If you operationalize data in an effective way, their data works. It's true, they come from the intelligence community, they've been collecting data for decades, different kind of data, but they apply the same models for B2B marketing and it works. It means that when we use the audiences that Mitico provides, the campaigns perform better. If you just apples to apples comparison. So that's the main reason. So we talked a lot about data, analysis to understand your ICP, analysis to understand your persona, to come up with your target accounts. There's a lot of data today, a lot of data. You guys probably have a lot of it in your lab as well. But what do you do with all that data? There's a lot of— there's kind of a graveyard of data where you're overwhelmed by the amount of data. You don't know how to use the data, how to synthesize it in a way that it's applicable and actionable. And you usually don't have enough people to actually use the data in an effective way. And so what MetaEdit— maybe I'll give you a little bit of a bit on myself before I start. I'm the founder of metadata.io, a 3-year-old startup right here in Silicon Valley. My experience— I'm personally a software engineer in my background, an AI person, software engineer. However, I spent the last 8 years prior to metadata running marketing. I was a CMO in 3 B2B companies, and my biggest kind of problem as an engineer in marketing is that B2B marketing is a shit show. Vast majority of marketers no idea what they're doing. They're just like trying everything and like, "I'm off because the CEO is putting a bunch of numbers on them," and they try just a million things, and sometimes they're lucky and within a year and a half they found something that works, and then they're not fired.

Vast majority of them don't find it within a year and a half and they get fired.

That's currently the status quo, I promise you. Data supports it. Now, what is data-driven marketing? I've been in the role for 8 years. What's been successful for me is that I'm self- I'm a software engineer and in our head, left-brained, kind of left-brained head, everything has to have a cause. Everything has to have something to support it. So when I go into the room with my salesperson and he tells me, "This is my target account," I don't try to argue with him. I tell him, "Okay, let's try it. Also, let's go back to the CRM, analyze it with a solution like Critical, and let's see what the data says. And let's not argue about it, let's try both. And if we try both, we'll see what's more successful. If you're right, then 1, you're right, 2, we close deal. If I'm right, then you're wrong, and then you close deal and you get your commission. So it's a win-win situation. That's what I've been doing as a marketer, and that's what we're trying to do with Meetigo in the market, to kind of crush that whole marketing sales tension, if I will, because there's no need for it. You know, sales get paid by commission because they close deals, and marketing gets paid if sales are active.

That's it.

And so, you know, ideally you can work together. The second thing is you can have a reason for everything that you do. You know what's working, what's not working, and you don't know it in your brain. You don't know it because you're hiring a really good agency and they keep all the information, and you don't know it because you have the biggest ego and the loudest voice. You know it because you have data and it shows you— you made that decision and you generate a million dollar pipeline. That's a good cause and effect. And so this is kind of the main screen for Medley. I like to show it because, uh, it's kind of the, the dream for, for a CMO. The reason I mentioned why it's a dream, because every time you go to ask for more budget as a something of marketing— director, VP, CMO, marketing manager, an intern— every time you have a project, you want to ask for more money, they're going to ask you, what did you generate? You put in $1. How much, how much dollars in Python you can generate? If it's 1, you're fired. If it's 3, then you can stay. If it's 10, you're promoted. And so here you show that you put X and you got 10X back. That's a win. And with that, you can go and get more budget and try more experiments and basically build your career. And so that's the main thing that I always tell people, like, it doesn't matter what you do, dollar in, dollar out, it needs to be a very big multiplier. Otherwise, there's no interest. So let's talk a little bit about the current life we live in. You have all these marketing technologies. There are literally thousands. I don't know if you know the LUMA, the LUMA Space. There are about 5,000 technologies. It's not really realistic. There are not 5,000 technologies, but about 1,000, and about two-thirds of them are B2C. But there's still a lot of technologies out there for B2B that basically do one thing. And you as a human, you have to buy this technology, try it out, then log in every day and use it, then log into another solution, and then log into another solution, and then take the data from solution number 1 and then number 2 and put it in an Excel sheet and follow everything that you do because it's a cheat show. You have many, many things that you have to figure out, and you have humans operating those technologies. And so even if the technology is amazing and your person, your marketing manager or your agency, is not very sophisticated or is not enough time into it, then the value you're going to get out of the technology is low. And so what MetaData says is that the technologies out there are awesome, but the people using them are awesome, but they're tired. They're tired running technical, repetitive, mundane tasks all day long. Coming up with UTM tags, buying datasets, segmenting stuff, onboarding audiences, logging into Marketo and Facebook and LinkedIn and AdRoll and doing retargeting, no marketing manager ever signed up for that work. If you ever watch Mad Men, "No, I'm not gonna do that." All of us got into marketing because we wanted to be cool. This is not cool. So that's the problem because at the end of the day when you do those technical, repetitive, mundane tasks every day, you end up running about, if you're lucky, 5% of what you should have been doing. Meaning if you have all these case studies laying around your resources page and and these cool vendor reports from Gartner, and white paper, and ebooks, and all these amazing collateral that you have. And you have all these apps, and you have all the digital marketing mix, essentially. All of your assets that you're trying to push out there in the world so people come to you. You have an unlimited number of campaigns you can run. You can try this audience with this campaign, with this collateral, with this creative, with this copy on this channel. The options are unlimited, but you have a limited time. How many campaigns can you possibly run as a human? 10 a week if you're like aggressive, 100 if you're insane, but you can't run 10,000. But a computer can run 10,000. And every one of these technologies today is operatable by API, which means you don't actually have to log in to any of those systems. You don't have to manually run those campaigns. And that's essentially the place where MetaDay comes, comes in. We're in the middle. And we're not here to give you insights. We don't use KI for insights. Sure, we obviously build insights before we use KI, but we're not here to give you the insights so that you act. We're here to build the insights so that our engine reads the insights and says, "Oh, okay, this happened, this campaign is doing better because 3 months ago I set it up and it's generated 7 opportunities. I'm going to go back to that campaign and 10x its budget and I'm going to eliminate 8 other campaigns that didn't generate that much." Usually it will take you 6 months to understand what happened 6 months ago. Now you have the ability to essentially run all these things in tandem. Get the data from Facebook. Oh, this campaign ran for 10 days and it generated, I don't know, 50 leads. But none of the leads became a QL. So maybe this campaign is not great. Although if you were just to log into Facebook, it looks amazing. All these 50 leads. But sales didn't do anything with any of them because they're unqualified. Those kind of insights are hard to act on, and they're also very hard to, in real time, understand. And that's kind of where metadata comes through. So having the ability to synthesize all these insights and act on them in an automated fashion, that's the kind of competitive advantage that you have when you use experimentation. So let's take a quick kind of example. Let's say you're, you know, you're— we talked to many companies and they don't really realize how many potential campaigns they could be running. They don't understand that they're— they live in a very, very small— they kind of run everything in a small silo and they're really barely scratching the surface in terms of what they could be doing. So even if you just have 4 audiences, let's say you have 1,000 target accounts, that's your entire— you know, you work with Mintigo and you literally have 1,000 companies in the world you should talk to in the next month. Even if you just break down these accounts by industry, seniority, technology that you use, or maybe even geography, you now have 4 audiences. It's very, very easy to get multiple audiences because you never want to target all 1,000 accounts with exactly the same message. That's ridiculous. It's going to give you very bad results. So you're always going to segment and start doing some personalization. Same thing for creative. You put an image of a woman, a man, a dog, a cat, people are going to react to it differently. Nothing is better than the other. You just try them out and see. Blue works better than red. Same thing goes for copy. Same thing goes for channel. Do you use the exclamation mark or the question mark? Do you use Facebook or LinkedIn? Do you use the white paper or an ebook? All these questions are things that you are today, the vast majority of you, are in a qualitative manner, kind of hunch. You have hunch. I'm going to try this white paper because I think it's going to be awesome. Or I think LinkedIn is really good for B2B. All these claims that are sometimes true, sometimes not. No data behind it. And also, if there is data behind it, it's benchmark data. It's not particularly for your company. And so very easy to get into hundreds of experiments in which you as a person can execute 5-10% of them. What happens if you have a little bit more than 4 assets? This is impossible to execute for an agency or for— I mean, an agency is made out of humans. Internal marketers are are working for the company, neither one of those teams have the ability to run so many campaigns. Just because even if you had 100 people executing campaigns all day, you still have to pull the data and not optimize the campaign based on impressions and clicks. No one cares about it. You go to the CRO and tell them, "I generated a million impressions." That's awesome. How much pipeline did you generate? In order to understand that, you have to take data from your marketing automation, from your Google Analytics, from your CRM, synthesize it, and then understand this campaign out of the 1,000 generated this pipeline. This is impossible to do in a manual fashion. It's not like scale or not scale, it's impossible to do in an accurate fashion. And that's where agencies fail. Agencies go and provide those vanity metrics, and if the pipeline ends up being successful, they stay. If they don't, they're replaced. And so what do you do with this? There's a good image I really like because it's very overwhelming to do that. Oh wow, I really like this. Aggressively pushing your face. So here's an example of a question that's very, very hard to answer. And every time I show the slide, I remember that I actually don't remember the answer. I have to follow to the next slide. This is a very, very, very basic question. If you have those 5 creatives, could you guess what's better? And you can see small things that we change here, like the text, the call to action, the image, the location. Impossible to figure out. I mean, for me, it's impossible. It's impossible to figure out. I never know the answer to these questions. I honestly don't care. But this is the winner, and it's significantly better. If you look at some of the results, we have some customers who are like 1% click-through rate, which is insane. It's like more than double the industry average. How did we get there? We do not have a long duration of working for us. The way we got there is by running 10,000 experiments. You run 10,000 experiments, you find the best creative for this particular persona, for this particular segment. You find that on Facebook, maybe you get half the cost per lead, although it's exactly the same persona that you bring in. And you also figure out that putting the image here versus there, or using a download versus, you know, learn more, will get you a 20% lift in your click-through rate, or 20% lift in your MQL, because it's targeting better kind of quality people. And so the lesson from here is that testing one variable is very, very, very slow and very time-consuming. But testing 7 is the right way to go, but it's extremely overwhelming. And so we recommend that you do the multivariate testing, and we also recommend that you automate it if you, you know, if you want to stay sane. And so this is kind of the, you know, that's, that's— we're talking about dollar in, dollar out. This is how the two things that you get from experimentation and data-driven marketing. So you run AI because you want to be sophisticated about how you optimize, and you want to run You want to use automation because you want to spend time on value-added tasks. You don't want to do the UTM tags and the retargeting and setting up campaigns. You want to be sitting and thinking about strategy, what kind of new audience should we go after, what's the right copy to use, what's a new content we should come up with versus executing all day long. And so you get two things from it. One, you understand exactly your pipeline. You know your ratios. It's a data game at the end of the day. If you generate month after month X amount of impressions and a computer is running those for you, so it's kind of predictable. It's not like you have a really good marketing manager one day and you have a really bad marketing manager the next day. It's the same AI running campaigns for you. And so there is the standard deviation, the difference month to month is very, very small. So you can rely on it. And when I say rely on it, I mean the ratio. If you generate, if you put $1,000, you're gonna get 100 impressions. I'm just making it up. 100 impressions, they will turn into 10 clicks. The 10 clicks will turn into 4 leads. The 4 leads will turn into 2 MQLs and into 1 opportunity. That's what the historical data show you. And so you can rely on it and say, okay, if my goal next quarter is to generate $2 million in ARR or pipeline, you can just plug in the, you know, reverse engineer it from there and see, okay, this is what my budget should be this quarter. And you can really rely on it. So we work with some big companies that that's what they care about the most. They care about predictability because they have a stock in the market, and if they go above or below their projection, they get dinged. And so it's very important for them. They have $5 million in marketing and they have no idea how to spend it. So we tell them, you've been working with us for a year, these are your ratios. For Europe, spend this. For North America, spend this. For this product, spend that. And you're going to get exactly 20% plus minus the results you want. And it's proven, it's proven to work because because there are no humans involved. And unlike Gail, I'm not going to be nice, I'm going to say we are here to replace vast majority of the work that you do today. Not you. Demand generation managers are going to exist, marketing managers and copywriters and creative people are going to exist for a good while, until that goes away as well. But for the time being, that particular job exists. And it's much more interesting to come up with a new creative or a new idea than to sit down for another hour and come up with 10 more UTM tags. I guarantee your life will be much more interesting, much more fulfilling, and you can be much more successful when you let an execution platform run the campaigns for you versus you doing that technical mundane repetitive task. And so we're here to replace the people who are doing a job that they don't like, they make mistakes, and they churn. And so that's what we're here to replace. The second thing that you may want to think about is when you need to prove your ROI for marketing, it doesn't always look the same. It doesn't always look like, uh, I generated 50 leads or I generated 100 MQLs. It's always— it's also not necessarily I generated $100 million pipeline because attribution is very sketchy game depending what vendor you're using and who's telling the story. Usually just happens that that solution generates the most, you know, pipeline revenue for you, and the other solution did it. And so it's a very sketchy game to play the attribution game. Who did that? Marketing? Was it sales? Is it journalists? Is it demand base? Is it metadata? Like, I don't know. And so one of the ways you can kind of uncover the lies is to just look at the raw data. When we were working with one of the companies that Franco actually knows, and we were going against, uh, one of the account-based marketing solutions out there in the market And they gave us basically 4 months to run a campaign. They gave us 50 named accounts. They gave us one, 50 named accounts, and the competitor 50 named accounts. And we ran campaigns for about 4 months. And at the end, we asked, OK, are you guys going to renew? They're like, we haven't even done a presentation to us. I was like, OK, we're sorry. It was the first year of operation. I had no idea what I'm doing. I said, what have my competitors done? They're like, well, we're supposed to show you what Here's the slides from the competitor. And basically what they were showing is we brought 6 of the 50 accounts for 4 minutes on average on your website.

It's pretty cool.

They went to Google Analytics, copy-paste a bunch of screenshots, and this is the ROI. We said, okay, cool. We don't have a Google Analytics tag. We should have put it there, but we didn't. And so we can't tell you how many minutes the target accounts were there, but let's look at the leads and let's look at the engagement. And so we looked at our own information from Facebook, LinkedIn, and programmatic. And we found that yes, some accounts came to the— kind of went to the website, we don't know for how long, but we noticed what pages they went from, what kind of website they came, what kind of clicks, what kind of channels they came from, what kind of creative did they click on, what kind of assets this particular persona consumed. And we could show them the entire kind of journey that the person went from from not knowing the company at all to potentially being in their pipeline. The second thing that we show them is that out of the 50 companies, 5 already started opportunities. 4 were influenced, but we didn't start them. One of them was triggered. And that's a significantly stronger value proposition. That you don't just talk about, again, impression, clicks, time on site, but you move to, I activated this account. They didn't know about us before, now they know about us, and here are some leads, and by the way, we got some opportunities that are directly correlated to that particular campaign. That makes it so much easier to get ROI for your campaigns. I guess the last thing that— we're not showing the revenue here. One of the things that I try never to say when I'm in a sales call, even when I'm speaking, is to promise results because we have no idea. We work with some companies that have really shitty copy, terrible products, annoying salespeople, you know, like messaging positions that don't work, creative that is dead. There's so many bad things that could be in your marketing mix. Well, one thing I can guarantee, if you use experimentation at scale, you're going to do much better than you did before. Because even if you have terrible creative and no one told you about it, or you know it by heart but you don't have the numbers— like, by the way, companies just like you, same size, same price point, using a different creative from a different agency, 3x your benchmark. Then they listen. But until then, it's political. Ah, is this agency better? That agency better? But when you put some numbers and your objective, makes the whole conversation a lot easier. And so that's the story here. We have a few customers. We have— we recently were adding more and more because we're an AI product by nature. The more we run, the smarter the AI gets. And so recently we started having accounts within one month generate like a few million dollars in opportunity. That's awesome. I'm like really itching to to put it on the website, but I'm not because then I'm going to have a customer whose marketing sucks and then it's going to have no ROI whatsoever in the first 3 months. But the point here is that when you use experimentation, it doesn't matter where you are, what your base is, you're 100% going to see lift because the nature of experimentation is optimization. After you measure it over and over, you try all the different variables, you're guaranteed to do better. And at the very least, if you're doing doing something wrong, you know how wrong you're doing it, and you know what exactly you should replace. And then you can actually improve.

Ah, I like this one.

That's it. Well, this is MetaData. It was a pleasure to present to you. If you have any questions, now is the time.

Or after, because we have also Q&A afterwards.

Right?

Thank you very much.

Thank you.

Yeah, as Gil was saying, Gil and I know each other since like over a year now. We were sharing a client over a year ago, and just one thing that got me hooked on MetaDela a year ago was just Facebook and LinkedIn on the same platform. So you just work on one platform and run campaigns on both. That for me was enough for it to, to just use MetaDela without even all the other AI and everything else. But if you have been doing both campaigns on LinkedIn and Facebook, you know how painful it is to create audience on both and creatives and all of that. So just so for that, just for that was very, very interesting.

I don't know if I'm— yeah, nice. Cool.

So now the last topic I want to just briefly touch on is tracking engagement on target accounts, meaning, um, how do you know the progress once you have defined your ICP, you have your target list, how are you keeping track of how you're doing against that target list? What else can you do? Like, are you being effective at driving, not just from an attribution perspective, but literally, like, are those accounts moving forward to the next stage and at what pace are they doing that? Because with account-based marketing, it gets really tough. It's not just about number of leads and opportunities and pipeline, because sometimes enterprise sales is very long, it's very complex, and opportunity is not a quick, straightforward conversion that you know is one channel, one person converted. There are multiple people, long sales cycle, sales and marketing acting together. So you want to track these accounts and track the engagement of these accounts to show that, you know, you're actually working with sales and everything is working. So I'm going to skip on my background. If you guys have attended previous meetup, you already know that. And we focus at SaaSQL in helping companies converting, generating leads from target accounts and converting them into opportunities. That takes different aspects. Some of the clients we are working with, and so I'm going to skip on that and I want to start the presentation with this, kind of thinking back on what Colin said, from the fish in the small pond, small fish or big pond, at the same time is don't try to boil the ocean. And there are many companies that do exactly that. One client I started working with, it was still last year, when I asked who are they targeting, and they said, kind of what Gil said, any company is above 100 employees, above $20 million revenue, Fortune 1000, and this was pretty sophisticated technology. And I saw in their target account list there was a manufacturing company, there was like agricultural companies, nothing to do with their technology. And I asked, why is that on your target account list? Because they have a lot of money. And one of them is because it was on the Rolodex of one of the sales guy. And so, you know, you can have the best campaign, best messaging, best everything— you are targeting the wrong people and wrong accounts, it's never gonna work. And also, when you're targeting 50,000 accounts, you can manage— you don't know how you're progressing against 50,000 accounts. You don't know if you are doing something good or not, and you just— you can't have a good message. So narrow down your focus and be super, super specific. Specific, which doesn't mean that you have to cut off opportunities because that's fear number one I get all the time. Like, okay, we're gonna— if we focus on just one vertical or just this target account, we are missing opportunities, which is not true because you want to just be programmatic about it. You start from the ICP, you decide what are the characteristics. Ideally, you want to use, you know, data and AI if you, if you have the opportunity to to do that. But start with a manageable list. What, what is a manageable list? You know, it can be 300 accounts, can be 500 accounts, 2,000 accounts, depending on what's your business model. Are you going enterprise or the small business, mid-market? When I was at Branch, uh, we were targeting 250 companies. That was our total market because we're targeting the top mobile apps in the US. If you look on the App Store, there are 250 the, like, at the top. And so we had to be very, very thorough in how we're targeting those 250 companies, but we could tell at any given point how we're doing against that list, how much of that list knows about us, where do they sit in the funnel. And then you want to tier those accounts because not all the accounts in your ICP is going to be the same. There are going to be accounts that are worth are worth maybe a million dollar ARR and others are worth $50K ARR. They're not the same. And then if you can, you can leverage intent data so you know, you can prioritize within that ICP even better. But sometimes you can ask yourself, if they're showing intent, you might be already too late. So you want to get even before they start showing intent, if the intent is really they are in a buying cycle. So you have nurtured them enough so that when they show intent, you act. But that's another topic. So intent, finding the right accounts, I think, you know, as explained enough. So if you can leverage AI and use data, that's awesome. But at the beginning, you want to set the goal. So if you have 500 accounts, you want to have goals, obviously, for number of opportunities you want to generate, how much pipeline, how many closable close one deal, but also set goals of overall engagement. Like, you need to understand how you segment those accounts. Are they cold? Are they aware of you but haven't engaged much? They're in QA, so you have already enough people engaged, you have an opportunity, or even more, you have customers. And set those goals at the beginning. I like to move every quarter, so I have a quarter time frame. We set a list of accounts, we We set a goal for the quarter on how many opportunities, how many we have to engage. And then we set a budget for that because you want to be realistic about it. If you say, I want to have 500 accounts and I want to have 50 SQL, 50 opportunities from those 500 accounts this quarter, and my budget is $10,000. I don't know, something like that. It starts to be very unrealistic. You're trying to generate like a million a million-dollar pipeline with $10,000. If you can, good for you, I want to talk to you about. But it's really, you know, you want to be realistic, understand your CAC, which we're going to talk next meetup, understand the value of each stage of your funnel, how much it takes to generate an opportunity from a target account. And you can see your average, how much you've spent before. You've spent usually, I don't know, $5,000 per opportunity, you know, that's going to be more or less your average that you're to go for, and so you set a budget that makes sense. And so this is what I was talking about segmenting your accounts. So we do that in different ways, and I'm going to show one of the ways we do that, but we like to pull aggregate data from different sources. You know, you have Marketo, you have HubSpot, you have Salesforce. They have all these sources where you can find where your leads came from. You know, they come from events, they come from from LinkedIn and Facebook ads, they come from all these different directions. But you're going to have a lot of those accounts that are cold, and so you know that. You have accounts that are aware of you, so you have 1 or 2 leads, but maybe they're not decision makers, and you have to still push on those accounts, and then those that are in QA. And you should have different campaigns for this audience. Accounts that are cold shouldn't get the same creating an offer and messaging as accounts that have previously engage with you. So you want to build this classic funnel that you're already probably doing for leads, but do it at the account level.

Quick question on that.

Yes, please.

If they're already a customer, how does this work for a company that sells multiple product lines?

Yeah, so you can have—

Are there sub-segments?

Yeah, so in that case, yes. You can be as sophisticated even within, you know, aware, you can probably build another stages there, when you have just 1 lead versus 2 leads or 3 leads from different departments. You can become pretty sophisticated. Also, your definition of MQL can be very sophisticated and say, I want to have, for example, at least 1 decision maker in this department to classify an account as MQL. Or sometimes you just say, if I have 3 leads from that account, that's an MQL. That's a pretty generic definition, but it's better than nothing. At least you have an idea of what accounts are warmer and what are cold. And then opportunity is pretty straightforward. Do you have an opportunity with them? Good, that's an opportunity account. True, if you have multiple products, multiple verticals, then you can separate that further. So different data that we pull to look at this, you know, obviously from CRM, from Salesforce, And then we have some clients that we pull data, for example, from Segment because they are a premium platform and they have, you know, 100,000 users on their platform and they don't like to mix that with their CRM because— or even less with their Marketo HubScore because you break the record. And that's a lot of them, there's a lot of noise. So they are not present in their marketing automation, they are in Segment, so we aggregate that data to to look where you have premium users already, in what accounts that we should target for the decision maker. Because usually the premium users are normally at lower level titles, but you know that there is activity in specific accounts that then you want to sell a license to the whole company, so you go and target the decision maker. So that's what we do. You know, these are other platforms that do ABM in some shape or form. Marketo is one where you can build smart lists based on accounts. Engager is another platform where you can build account scoring based on minutes of engagement. That's how they define your MQA versus, uh, like cold or warm account is based on how many minutes they engage with your content, with your campaigns, events. Anything is tracked to that, that level. And this is a way we do internally, as I think you have with some of the clients, we basically use Airtable to aggregate all of this data and then cross-check with the target account list. Example, segment data as one thing, or sometimes they have a list of events that they haven't pulled into the marketing automation, and we cross-check to see how this account sits, and then we create a chart basically. So we know how we're trending. And then we have a goal on how many from cold need to go to aware, how many from aware need to become NQA, etc., etc. We use Zapier a lot to automate. I recommend it very strongly because it saves a lot of time when you're trying to create these sophisticated flows that you can, you know, add to a specific spreadsheet and and then create a communication email and do a lot of— create a task on Trello or Asana if you use all of that. It's really— it saves a lot of time and it's good when you're getting started. Again, intent data, Guillaume has talked a lot about it. Mintigo is one platform. There are other intent data— maybe they don't use AI like Mintigo, but you can find that. But don't Don't just add accounts because they show intent. That's another thing we saw. Like saying, oh, this— they show a lot of behavior, they want to buy, so that should be our target account list. No, it shouldn't be. If it's a bad fit from a firmographic perspective, and it's gonna be—

from an ICP perspective, right? If it doesn't fit your ICP, it doesn't matter.

It's too small, for example, or it's too big, even if it shows intent, you're not gonna sell to that account. And even if you do, you're not gonna be successful.

You're gonna waste your time and effort and budget.

And you know, there's a lot of data, like we said, so artificial intelligence, as we kind of— the subtopic of tonight— can help a lot. So in conclusion, start with a well-defined target. I give example of our own company right now. We're targeting 500 accounts. You know, we sell to SaaS companies that are venture-funded, so in theory we could target 10,000 probably, but we narrow down to 500 because obviously if it's too big, they probably have, you know, large internal teams. If it's too small, they cannot afford certain things. So you start becoming more specific into like what a good account looks like for you, and then you can track. How you're doing against those accounts. And if one account hasn't heard from you and you know they should, you can be very aggressive in terms of campaigns that you can set up. And you always have that in front of you. We define the KPIs. So a lot of companies understand marketing. It's more likely that they underspend versus overspend. Because they are— especially when they are just getting started, they are afraid of— scaling certain programs, and sometimes they don't have the right data to say what works and what doesn't, so they don't feel comfortable turning up the lever. We work with a lot of clients that way where they were kind of driving blind a little bit because they said, "Okay, we would like to spend more. We know something is working, but we don't know exactly what." And so they were just underspending so they were not you're not hitting your goals. So you want to start testing very programmatically different channels, like kind of a kill show— how much you spend, dollar in, dollar out— and then start scaling what works and start cutting what doesn't. And set a realistic budget. Don't try to generate a million-dollar revenue with $5,000. It's not gonna work. You know, there are benchmarks for that. Even the most efficient companies in Silicon Valley will never hit that level, so you know you're probably not going to be even close to that. And then every 1 quarter or 2 quarters, relook at your target list. Look at how many of those have you engaged, so you can revisit, is it still a good list? Are the filters still correct? You might go upmarket more, depending on your product. Maybe you might redefine that. And so you want to look at it. You might have not— that you might need to tweak your ICP because it was not correct to begin with. So always review that and look at the kind of response you get. And so the beauty of focusing on the 500 accounts, 1,000 accounts, like the— no boiling the ocean is that you can go very horizontal. So you can hit the same account from multiple channels at the same time. So if you're running LinkedIn ads, Facebook ads, email, direct mail, events, webinar, all with the same account list, I guarantee you they're going to see you at some point and they're going to start engaging. We have quite a few examples of that matter where the prospect, the company literally said, I started seeing you guys everywhere, so I knew I had to engage at some point. And it works. Instead, if they just see you one time, one hour on LinkedIn and that's it, not as much. And that's it from me. Thank you. So now we're going to switch to a small Q&A. I have a few questions, so let me grab a couple stools.

Boom! Yes, I got it.

Great. So I want to start opening up. Is there any question before I start with my questions for any of the speakers? At least this is the time, so I'll start. Okay, so I spoke on my side. I want to hear your opinion on this. I'll start from Colin. The size of an ideal target account list, what it should be, because like I said, a lot of time companies start saying, well, that's too small, we are cutting off opportunities, uh, we want to include more and more and more. Or sometimes they go too narrow, these are my 50 accounts that I want to sell to because I have the top of the world. What is the ideal size?

Yeah, I mean, sort of like you said, it depends on, I think, how much you want to invest in acquiring that customer. There's only, sort of like Gil talked about, there's only so many experiments, only so many campaigns, and there's only so much time on the team. And so for us, the way we like to look at it is sort of the top 50, the next 200, and then sort of the rest. And that way you've got your top 50 is getting that sort of dedicated time, get in focus from the team. And it's going to vary from customer to customer, right? If you have 100 SDRs or 2 SDRs, these numbers are going to vary quite a bit. It's going to vary based on your market. But I think in general, when we're working with sort of Series A companies, you definitely want to have your top 50, your top 200, and then the rest can be broken out into sort of sub-persona ICP combos.

What— so from your side, Bill, that if I come to you with 50 accounts to target on LinkedIn and Facebook, it's a little narrow audience. So what is a good size that you can effectively run that type of campaigns on display as well as LinkedIn and Facebook? Because maybe I have my 200, 300 accounts. Does it work, or you have to look in a different way?

So it depends what kind of company you are. So we do have companies come to us and say, hey, we only have 50 accounts and we're targeting like 5 people within each, you know, every one of those 50 accounts. Like, we don't need MetaData. You barely need marketing. It's sales-driven. Yeah, just on the phone, go to those, visit them, direct mail them, like do whatever crazy stuff you have to do. It's only 50, so go do it. But if you're not targeting only 5 to 10 people within those accounts, you are targeting, let's say, 100, because you're doing cross-sell across Nike and Pepsi and GE and large conglomerates, then if your job is to do a cross-sell, then you definitely can work. Like an audience of 5,000 people is effective. So when we get customers and they have some idea of who they want to target, I ask them, what's the purpose? Are you trying to generate top-of-the-funnel leads so that in 3 months they're going to be ready to buy? Then you need a large audience, you need like at least I don't know, 5,000 accounts to go after because otherwise every lead is gonna cost you $1,000. If you're doing account-based marketing and you're super, super surgical and every one of these accounts is worth like half a million dollars for you, then that's fine. Just identify those people and go after them everywhere, including ads, but not only ads. Direct mail and emails before, and events, and buy one and only kind of vendors that will get you meetings and all of the above. That's okay. So we always keep the balance between the goal of the campaign and the audience that we're going after.

And from your perspective, obviously, you know, you see the data and all of this that should match the list. What if the list is too big because still a lot of companies match that ideal customer profile? Is that a signal that the ideal customer profile should be changed, or it could still be a good way to go?

Typically, when you work, the more data you actually use, right, and this is, you know, the better and granular you can get. And it's not only about, right, the specific ICP, it's also how you can take that down into audiences. And you can really take, you know, a big list down to audiences using data. So do you integrate with a specific platform? You can target all those companies that integrate with that specific platform, right? It's like how I showed you the hyper-personalization. This is where you can go and trim down to specific audiences and really focus— this quarter I'm going to focus on that specific audience and not kind of my whole ICP. And it also depends what are the tactics that you're going to use, right? So if you're using marketing automation, tactics that are typically cheaper, you can go a little wider and try to widen the net. If you're going to use sales, you want to be very precise in who you're going after and the persona you're going after. So this is where these audiences come into play. You would say, let's have the cold accounts— you call them cold— let's have from our target accounts, all the cold accounts, let's target through metadata. Let's see, start to build awareness. Let's see who is raising their hands and saying, "Hey, we're interested in your solutions." Those that are raising their hands, send them to the SDR team that Colleen is running, right? And they will go ahead and call those. So you can really start to trim and start to— when you have the data and when you start looking at it through DataAntle, you can really decide what's your strategy for every matrix. Are they cold? Are they good fit but cold? Are they good fit but already showing intent? Are they, eh, fit and showing lots of intent? So you can actually, for each one of them, you can define what's your next best action, I would say. Right? And this is how the trimming goes down and how you're very, very focused, not only in terms of understanding the ICP and the insights, but actually where they are in the buying journey and what you should do with where they are. They are.

So on that note, I want to continue on the same level. This is a— and from you, Gail— this is a personal experience that I saw with different clients where, you know, all of their target accounts that they selected was part of their ICP were pretty large companies. Not that we're only the only companies that we're targeting, but, you know, they like obviously to get what we call the whales to engage. But what happens is, you know, those take time to close and a lot of money, so it takes— it's a long sales cycle. So they don't see the quick turnaround and quick results. So what is the good ratio of those whales that kind of makes 50% of the quota of someone, or even an entire company, versus that more high-volume, mid-market, small business companies?

So I think that's again the beauty of an ICP because they capture both audiences. An ICP can capture both, like, the mid-market companies that Mintigo sells. Basically, if I look at us as an example, right, we sell to mid-market companies that are growing, and even like companies of 50 employees that are like growing like crazy and want to grow and don't have their own capabilities of like 100 data scientists to build their target account list. So they like to use Mintigo. And we have the whales that have 500 data scientists, but at the end of the day, that doesn't scale neither. So they need a solution that would allow them to scale. So we sell to both, right? And the ICP actually captures both audiences, and we target both audiences. And it's just, you know, it's just the sales cycle. The sales cycle for the smaller ones would take 2 to 3 months, while for the whales would take between 6 months and a year and a half, right? You start, you start to build the relations, another meeting, another meeting, another meeting, and it can take a long, long time. So you actually want to target both, and you don't want to target typically in those cases only the whales. And even within the whales, you really want to target the whales that are relevant, because there's a lot of whales out there and everyone wants to target the whales. But, you know, it's the easiest is that the predictive model would figure out the fastest way to get revenue is to target the whales, right? If you don't have a good predictive model. So what you want to do if the whales are your target, you want to take the target audience as all the whales, and within all the whales, those that are successful, and only target those that are relevant, those that are showing intent, those that have the high scores within the whales. So I think these are the two things that I would do. I would do a mix. Of mid-market and whales, but also within the whales I would target those that are, you know, relevant for your business, because not all the whales are relevant. Like what you mentioned, we can't sell to an agriculture company, that won't work, right? Although whales are really good for us, right, because they do have the budgets, the people, they understand why they need AI, blah blah blah, but agriculture companies would typically not buy from Intego, right? So we need to be smart about the whales as well.

Even though it has minted.

What?

Even though it has minted.

Exactly, it has tons of mints.

Yeah. So for you, you know, again, this whale versus mid-market when you build the audiences, or even for you guys on Metadata, how you think about, you know, those labs enterprise versus mid-market as part of your ICP or as part of the main account for the labs?

I can give you my own experience. Because every company is different. For metadata, we always give it a very healthy distribution of the startups that are going to teach us about the product, how to build the product the best, because they will take the most time, the most attention, pay you the least, and be very, very demanding. And so those really teach us what to do because they give us all the feedback in the world, also the feedback we didn't really want. And then there's the mid-market companies. That's where— basically, yeah, that's a band diagram that almost everyone is in there. So you want to get a lot of them because some of them are going to— you know, the startups are probably— the startups, the vast majority of them die. And so you have churn just because the nature of startups. But then you have the mid-market, they mostly survive. Some of them, they won't grow too much, but you have a good opportunity there. And the enterprises, those are the big bucks, you know. You get $300,000 more account, the VC gives you $5 million more. And so you want it. We always keep a nice healthy distribution between them. My biggest worry always is not to have more than 15% of my revenue come from the elephants and not have more than 40-30% of my accounts coming from startups, because both of these are super risky. I lose one account, I just lost like 10% of my revenue. There's going to be a big conversation to have with the investor. Yeah, so that's how we usually keep the kind of— keep the balance.

And you mentioned at the beginning something about, you know, the, the different tiers of accounts, how you guys distribute that. Because I assume even for you guys, if you start generating all these opportunities but then, you know, 9 months have passed, none of them have closed, it still doesn't look as good as— is that something that generates revenue anyway?

Yeah, I mean, I agree with Gil on the two points. One you made in your talk, that not all customers are created equally. Some you work your ass off for and you generate tons of pipeline and they don't close it because because of a number of reasons. All the more reason to get smarter with your targeting ahead of time. And then just to sort of double-click on what you were saying there about making sure that you're not trying to bring in more than 15% of your revenue from those whales. I had an early advisor that basically told me to— the kind way to put it is stay the fuck away from any company that has more than 10x your revenue. And that actually steered us away from a couple of really important deals that would have taken us in the wrong direction. And so we've slowly grown, we've grown our deal size, but basically there's two things that happen when you chase those deals. One, it's Feasterman, right? The person that goes out elephant hunting and forgets to collect the sort of deer and squirrels along the way, that person starves out wherever you're hunting for elephants. And the second piece is if you actually do manage to land one, if they own you, if they are 30% of your revenue, they own you. You're basically their subsidiary, but you're not getting the same financial benefits. And so they end up steering your product/service in the direction that serves them the best as opposed to what's going to be best for you.

Awesome.

So I have two more questions, and if there isn't any question, this is your time. But I have two more questions for— I'm going to start from Gil this time. So the first one is about personalization. And when we talk ABM, immediately a lot of companies think about, oh, we need to personalize absolutely everything from the landing page, the content, the white paper, and all of this has to be super specific. And obviously that takes time. Time, efforts, people, resources, money. What is the right level of personalization when we look at the ICP? When is it too much and when is it too little?

As usual, I don't have a silver bullet answer.

It's just like— it's really an office panel.

I know, unfortunately I don't have one. But this is what we see from customers. Some of them are over-engineering the hell out of it. They just, like you said, they build very particular journeys for one company and you just spend a week. I know a company that spends legitimately a week per company that they're targeting. Again, if you have 50 companies that you're targeting, makes sense. If you have 500, it's not gonna happen. So the balance is to do the bare minimum to get that persona from that company to click and have a conversation with you. And so that's why we do experimentation, because we don't know and we don't really have the time to run 10,000 different variations. Of creating collateral, etc. So we put a bunch out there and we constantly experiment. You didn't like this message, you're gonna get another one. You didn't like this message, you're gonna get another one. You didn't like this color, you're gonna try a different color. To what scale? To what scale?

Do what scales.

To what scales, exactly. And what scales is not always obvious. You may do it for one month and get leads that will never turn into opportunities. So 3 months later, the longer you let the experimentation dive deeper into your into your funnel, the further down the funnel you're going to optimize towards. So if you let it— if your sales cycle is 3 months and you let AI optimize your campaigns for 6 months, now you can optimize towards booking. You know what campaigns— forget about impressions, clicks, leads, MQLs— what campaigns lead to bookings. And it's again fairly predictable. So that, that balance should be kept very tight. The only time that we personalize very, very deep is when we have a a small number of accounts that look exactly like our best customers. The small is usually a few hundred, and we do everything against them, like what we did with the, with the direct mail. So very personalized email that are followed by direct mail. In the meantime, they also get very personalized ads. Let's see, you go to a landing page with one thing is personalized there, the name of the company. Every company will see exactly the same landing page, and they'll think they're special because their name is there. That combination 20/80 rule works.

I have personally that experience in one company where I worked, not gonna mention which one. We tried to personalize for 5 accounts and we were building specific landing page where we went on their site and their app to look at what they were doing and take screenshots and put it on the page and do all of that work. It took 3 people almost a month for 5 accounts. We didn't see much more engagement than the other 100 that were running one campaign, one message, one everything. So obviously if you're selling to IBM and you have a team dedicated to sell to IBM, sure, do that. Build a white paper just for IBM. But in the majority of the cases, that's not worth it if you are a startup. So what do you think, Paul, on the personalization side in terms of the email or the campaigns?

Yeah, I mean, I can't comment too much on the landing pages and that content on the marketing side of things, but certainly on the sales side of things, if we're talking they're in your top 1,000, you know, you're probably going to want to be investing a little bit in personalizing some of that. Especially just in the last— I don't know if anybody else has noticed, but in the last 2 years it's become probably 10 times harder to send cold email at scale. I think you get a lot of people, "No, I don't." So I think it's, for sales development people, we have to be a lot smarter. We have to engage more with the marketing side. That's one of the things that we're starting to see as a really popular trend is taking a lot of the data that marketing has generated and then passing it off to the sales development team so that we can execute on that. And using these sort of warm signals to start the conversation as opposed to just cold list. So if you're doing— if I'm doing pure cold, you know, I'm going to invest quite a bit. If it's a, you know, top 50, I'll invest a little bit less. Probably on the top 50, I'm probably going to spend, you know, half an hour to a couple hours depending on how much it's worth to me. If I'm working on my top sort of, I don't know, beyond outside that top 50, I generally have a 3x3 model, which is you want to learn 3 interesting things in about them in the first 3 minutes that you're researching that account, you save that in your Salesforce so you can bring them back in the day. And I say 3 minutes, but really 3 to 5 minutes. And when I'm training reps, I like to get a Pomodoro timer next to them because it's so easy to just get lost on the internet. You're gonna, you know, you're following them on LinkedIn, you're engaging with some content, and then all of a sudden it's like, "Oh, that's really interesting," and now you're looking at bikes on Wikipedia and you're like, "What the fuck did I get here?" And so having the timer is super important because is you absolutely— I have 100% done, you know, custom huge, you know, email that's, you know, this long and it's written to this customer, 0% response rate. The equivalent of spending a month of 3 people's time on one account and got nothing. Such a waste of time.

Yeah.

So as we think—

don't you, don't you template, uh, I mean, when you go in and you identify and you work work with these, with your clients to refine their sales process. Aren't you helping them template emails that their salespeople use, or that are automatically triggered anyway by an action, so that they're not wasting their time doing clerical work, doing data entry, doing all that stuff?

I mean, they're supposed to be selling.

For sure.

And I think what, in the context of a sales development rep's first time, having the target accounts and contacts laid out for the sales development team sort of reduces a lot of that clerical time so they can focus on the research and finding the really interesting things and then personalizing it. And I'm going to just bastardize something that you've probably heard a couple of times tonight, but we call it the 20/80 rule. The 20/80 rule, just to piss people off, because it's normally the 80/20. Basically, that's you've got your 80% is templatized and the first 20% is personalized. And generally, if I'm looking at an email, the first— they don't have to be separate.

It can be templated and it can be personalized.

For sure.

Yeah, I'd say the first line is personalized, and then usually the PS line is personalized, and maybe the body might be depending on the data we have access to, might vary based on the account.

I'd like to give an example there. So at MetaData, for example, we use G2 Crowd data. G2 Crowd, I don't know if you're familiar, it's basically a Yelp for SaaS. You go and check if a marketer is good or is hospitable, better, et cetera. And we buy their data. So they sell their data direct. You want to know exactly who's visiting. For us, the account is marketing, account execution, this company is that, we want to know about it because we're a small startup, no one even knows our name, but majority of the companies know us. And so we want to be in the decision, kind of in the table of consideration when they're considering other players. So what we do is we have two tactics that work very well with GitOps Cloud, which is an example of personalization that is completely automatic. One, every time we get an— essentially we get every day from GitOps Cloud companies who are visiting this page and we're looking for this keyword. They provide those two, those two columns. This company from this city, from this town, we're looking for this keyword and on this page. Now it's useless if you don't know who you're looking for within that company. Good luck finding within 100 people who visit because it's anonymized. They can't give you the data for privacy reasons. But what we do in an automatic fashion is we go, we take your Salesforce data, we understand who your persona is historically, so we take that We find that person within that company, within that city, within that state. So we match and we guess, we don't know. There might be 50 people that could be your persona, it could be 10. If it's 10, send an email to all 10 people. And that's what we do. We take the exact keywords that they put when they were looking for H2Crowd, we plug it in the email, we say, "Hey, first name, looks like you just visited this page looking for keyword." We do exactly that. It doesn't matter what it is. Whatever it is, that keyword, we do it. And so very basic, nothing sophisticated. Reply back, oh really? Of course I put all of the information there to creep them out because I'm saying to marketers, marketers actually see it because it's a competitive advantage. So I show them all the information I know about them. Many times I get a reply and it just puts us, you know, some kind of engagement, zero manual work. The second one is we run advertising, right? This was email. There's a simple tool we built internally, but we also run ads. And so we get the list of companies. We know more or less, we guesstimate again, who is visiting the pages. And what we do is we use the keywords, especially when it's non-branded. They were not looking for metadata. They have no idea who we are. They were looking for account-based execution. So I tell them, check out the 5 top account-based execution. We're top 3. But let's say we were not top 3. Let's say we were number 19. Still worth it. Because now you've given me your email. I give you the report, which I'm getting from G2 Cloud as part of my package. I'm in decision-making, but now I know that you guys just give me my email, you're looking for this, I give you value, and in return I get a chance to sell to you. Those personalized, zero manual work.

Awesome.

So for me, it's just, it's exactly that. I think technology helps you personalize. Um, doing that in a manual fashion is really, really time-consuming, daunting, and quite impossible. So the idea is, is is actually using the data that you know about the specific company and the data that you know about your ICP, really define these 3 more strategic audiences out of your ICP, and automatically, if a lead hits your Marketo, and you know they're showing intent on ABM, right? Hit them with the right message around ABM. If you know they're showing intent, for us, let's say on predictive analytics, we can get more technical, talk about AI, we can talk really differently based on the interest, based on the ICP, and we can do that automatically basically because the data comes in an automatic fashion. So it's without human, we just route them to the right personalized campaign, and yes, we don't have 200 of those, we still try to limit ourselves to 3 or 4 types of campaigns or sometimes a little more, but if we have a really good story, and you have the data, you can actually automate the personalization. So that's kind of—

yeah.

So I had more questions, but I think we are out of time. Thank you. Thank you. I hope you are gonna attend the next meetup. It's gonna probably be in July. We talk about CAC analysis and funnel metrics and one of the funnel stuff. So thanks again and have a good night.

Thank you.