AI in Sales - ELRIG Presentation

All right, really excited to go through this presentation with you all. This is a presentation I did for ELRIG and the vendors who showed up earlier this week. Basically we did the same thing as last year where we had a little session with a couple of talks and a panel. So next year if you're around at elrig, you'll come on by. I'm sure we'll be doing it again. So this one is all about AI and sales and so I'm going to share a couple of specific workflows that you all can use and deploy your sort of immediately into your processes. But before we dive in, I want to do a quick background on who we are here at Succession. So if you don't know who we are, I'm Harrison, one of the co founders, another co founder is nicclair. And we basically work with life science companies in three sort of capacities. One is around top of funnel lead gen, so actually run lead gen as a service. And so some of the stuff that you're actually going to see in here and see us talking about stuff that we actually use with our clients as well. And we also do sales training, so whether that be in person, virtual or on demand, we also have sales training that we do for companies and then we'll also implement different AI and automation workflows into your commercial operations. And so the last couple of years we've worked with 75 or so, maybe even a little bit more than that. Now companies in the life science space, specifically around these three core areas. Cool. So now diving into it, before we go into the actual workflows, I want to give you a high level framework for how you can think about and how you should be leveraging AI in a day to day basis. So the first thing that you want to do is you want to recognize the goal, which is you have to know what you're trying to get out of AI and what is the outcome that you're trying to get. What is the goal? What are you trying to achieve? You need to understand what that is first and it's probably the most important thing that you need to do because that way you're going to be able to tell AI what good looks like in order to help it, you know, get the context that it needs in order to actually give you the outcome that you're trying to achieve. Right, so we'll talk about that in here in a bit. Then you need to articulate what good looks like, meaning you have to give it specific examples of what you want the output to Be I'm sure you've all gone into ChatGPT, asked a simple question, and then you get back something that looks really terrible. And I think one of the main reasons for this is all of these models are trained on like the entire Internet. And so it's trained on basically every average thing, right? So it's the average of everything that exists. And so if you want something that is above average, you have to specifically tell it what good looks like. You then need to provide the context. So the more context that you can provide, the better. So this is going to be all of the background, everything that you're trying to, you know, help it to understand where that's information about your company, people that you're prospecting, whatever it is, you need to make sure that you're providing an immense amount of context. And the more context the better. So let's say feed the beast that is AI. And then we don't want to just take the output and then send it or whatever we're going to do with it, you know, publish it, whatever it is. You don't want to do that. You want to make sure that you edit it and improve the actual output. So you're, you know, editing, making it feel human. You're removing those EM dashes, you're removing some of that language. It's like if this, then that and like all these things that you inherently start to understand, like this is AI written. That's the type of stuff we want to go ahead and remove. And then you also want to fact check everything. There are still hallucinations, even though it is now leveraging the Internet, going out and actually citing its sources. Sometimes it will cite a source and one the source isn't real, or the thing that it's sourcing or citing from the source is wrong. And so always make sure that you're fact checking things that are important and that need to be fact checked. And then lastly you're going to actually sort of deploy all of these things out and actually use them, document them and save them. So if you have prompts that you use on a regular basis, you could turn those into like custom agents, custom GPTs, all sorts of like ways that you can document save this stuff or you can just simply make document a PDF, you can turn some of these things into skills for Claude. So there's all sorts of really cool ways that you can document and save this, but then that way you can then share it amongst your team as well. So not just for you, but you can also then share these Prompts and these different agents that you build specifically with the rest of the team so everybody can benefit from it. Great. So the six workflows I'm going to talk about here are going to be around prospecting messaging. So this is a really common one, but I'll show you where a lot of people go wrong, kind of where you need to do things better to make the outputs better. Meeting follow up so I can write faster email follow ups, account research, data enrichment list building, inbound lead, follow up, and then some creative ways that you can stand out. So with that said, let's jump into the first one which is prospecting messaging. So I think the way I see a lot of people think about leveraging AI right now is they'll say, hey, I work for your X, Y and Z company and I want to write a prospecting message to this person and they'll put in that person's LinkedIn profile and then like assume that AI is going to do a good job of writing those messages. And if you've done this, you probably look at the message and you're like, wow, this is really bad. And that's because yes, it is bad out of the gate. And so you have to really train and hone AI to do the things that you want it to by showing it what good looks like. And so there's three pieces of context that you have to provide for, for like any sort of messaging, I would say to be successful. And one piece is everything about your company. So this is going to be not just like your link to your company website, but also like the different products and services you have, your main differentiators, any sort of case studies, data points, things that you have that you can share, the Personas you sell to but they care about. Like all of this is just like one piece of context. And when I say, you know, you want to have these things documented, we're not talking about like four bullet points. Like this should be like pages and pages, PDFs and web pages, all sorts, like tons of contacts is what you want to be providing here. So that way AI can leverage it and really understands what it is that your company does and who you help solve problems for. And then you have the prospects information. So this is again not just somebody's LinkedIn profile, but it might be the LinkedIn profile, the website, there might be news publications, press releases, maybe there's hiring going on within their company, all sorts of things that is just like context around the prospect, like basically like deep research of that individual person in their company to know what is going to be the best angle for our messaging to be written around. And then this is the last piece where I think a lot of people mess up or they don't do it at all. And that is you have to tell AI how to actually write messages well, right? So this comes back to our framework of what does good look like? Well, what good looks like is you tell it what to do, what not to do, give it specific rules, give it frameworks of how to write, give it real examples of real emails that you've written that you've seen be successful and to give you context around how like important this is is we have a 27 page document that we use that we feed into AI to write these messages, right? So I mean you don't do that every single time, but you have this saved and now you have this sort of like framework and sets of frameworks for how exactly you want AI to write all of its messaging and create those rules. So that way when it does write prospecting messages, they look human, they look real and it doesn't look like AI. And I think that's the piece that a lot of people mess up. So these are the three core things that you need to pretty much get any prospecting message to look good, right? If you want average messaging, then don't do any of this and just kind of, you know, just go into ChatGPT and you know, give it your company info and you know, tell it to write an email. If you want really good prospecting messaging to stand out, which is what we all want, this is how you go about doing it. All right, so now into the next workflow that we're going to do is meeting follow ups. So I'm sure you all have at some point in time had a 30 minute process to follow up from a meeting where you have to like then try and remember, go through all of your notes, write all of those things down, put it in a follow up or find all the content that you need and then add it into that, you know, email. And it could take like 30 minutes to make all this happen. So what I'm going to do is I'm going to show you a little workflow to speed that up. And there's a couple other benefits of doing this, but it all has to do with call reporting and you've probably been on calls or you use call reporting software already. Some of you don't. I would highly recommend that you do it though. And what you want to do is you basically report the calls. And what's going to happen is a couple things are going to be beneficial just from having them recorded. One is all of the notes are going to be taken. So when you're on the actual call, you don't have to be sitting there writing things down yourself or typing your notes yourself because AI is going to be doing that for you so you can be more present on calls. So that's a benefit. And then I would say the pushback we always get here is from sales reps thinking that clients aren't going to want their calls recorded. And I would say that probably 99% of them don't care and that if they do, you can always get an NDA in place or you can always just kick the call reporter out of the meeting. But in general, I think a lot of them are even using teams or recording software themselves internally. So I think everybody's pretty accustomed to this now. And so you do it and if people have an issue with it, you can always just remove it from the call. Right? So I wouldn't make sure that you're actually trying to do the this stuff and you're leveraging this technology and you're not, you're making up your own objections for what you think your customers or clients might have issues with. So record the calls, they'll save you notes and then you'll always have access to go back to it. Then what you could do with a lot of these tools is instead of spending 30 minutes writing that follow up is you can ask that tool to write the follow up message for you. And it won't be a good message on itself, but what it's going to do is it's going to give you all of the action points, all of the bullet points from the conversation and it'll do it for you very quickly to where you can just edit the message rather than having to do all of the writing yourself. So it really helps speed up that process. And then lastly, so two benefits here is next time you have a follow up call with somebody, it's going to be great where you can go back to actually listen to the call and. Or you can query the call and just say, hey, give me like a quick recap of the top five things we talked about. Because sometimes there's weeks or months in between different calls and conversations. And so by having them recorded makes it really easy to go back and get that intel so you're not asking the same questions over and over on calls. And then lastly you can even go back and listen to your own calls to see how you can improve. So you think about any professional athlete is always going to be watching film of themselves to try and get better. Same thing that you have here, right? This is your film just going back, listening to calls and figuring out how you can improve. And so if you do this over and over right, you're just constantly one, you're going to be improving your skills, you're going to be speeding up all of the follow up time from all of your meetings and you're just going to have a much better, I would say, flow of a work process to where you're going to constantly be on top of things at all times. So definitely record your calls. And if you're curious about different tools, there's all sorts of them out there. Zoom teams, Fathom, Otter, Fireflies, there's countless numbers of these. So it's just pick one and use whichever one you know you feel most comfortable with. This is an example of Fathom where there's this button up here at the top where it says ask Fathom, where you can basically pull up any call and you can ask it questions about that call and it'll answer them based off of the transcript. So really useful stuff, and I highly recommend it. All right, the third workflow is going to be around account research. And the way I like to think about account research and how you can leverage AI for this is to think about what are all of the things that you look for when you're doing research on an account manually. So think about an account, go on the website, go do all of the things that you would normally do to do that research, and then write down a document, all of the steps that you took, what were all of the things that you were looking for? What was your logic behind doing each of those things? Right? And so what you're going to be probably trying to uncover is like a brief overview of the account, what workflows your company can solve for them, latest news in the last three months, if you don't prompt it around this type of stuff, you may get news from a year and a half ago, which isn't going to be relevant. So latest news, leadership changes or announcements, what are the therapeutic areas, modalities, drugs in the pipeline, et cetera. And then what you're ultimately going to do is you can even ask AI to do this for you, but you're going to basically craft a prompt. And so what you might do is, you know, say to AI, hey, here are all of the things that I look for when I doing account research, create me a prompt that I can use over and over that's going to give me all of this information for any account that I give you. And one of the things I like to do with these as well is have like the input, like be just something that changes each time. So it might be a company domain, a company name, a lab, website, whatever it is. And then that way all you have to do with your prompt is simply change the input, which is going to be that, you know, company or domain that you're trying to reach out to. So in this particular case, this was for a client we put together. This is the prompt over here on the left, and then down here in the bottom left is where all, you know, we put Moderna in as an example, and Moderna is the input. That's it. So the prompt can stay the same. And then in this case the output over here is all of these different potential workflows that a company can or how this company could potentially leverage the different products and reagents and services that this customer has. And so this was a way that they like to do, you know, account research is these are the types of things that they'd be looking into. But you would again, customize this for yourself. And then what you could do is you can turn this into, if you're using Copilot, you could turn it into an agent. If you're using Claude, you can turn this into a skill. If you're using ChatGPT, you could turn this into a custom GPT where then it's basically trained on this prompt. And all you have to do is go in there, put in the input of the company, and it's going to spit out the exact research that you want and all of that intelligence. So then all you have to do is anytime you want to go do research, instead of you manually doing that 10 minutes, 15 minutes of research, you're able to just get this done pretty much instantaneously. So great, simple way that you can do account research with a prompt. And I think a subtle thing there around account research is use AI to give you the prompt. So tell it what you want it to do and then say, give me the prompt that's going to give me what this output should be, and then that way, you know, it's going to generate a really good prompt there for you. Cool. So now the next one I want to talk about is data enrichment and list building. So when it comes to just building lists or enriching data, AI is really good at doing A lot of this stuff at scale, repeatedly, very quickly. So this is an example of a tool, a screenshot from a tool called clay. Clay.com, you may have seen this on LinkedIn, but there's a lot of people out there that you are posting about Clay. Clay is a really cool tool, really powerful. And essentially what we're looking at here is each of these columns is like, or excuse me, each of these rows is a person. And then each of these columns can be different, sort of what we call enrichments, where you can pull in data points from across the Internet, from scraping websites, from people's LinkedIn profiles, all sorts of stuff. So this example of data enrichment was all we had was first name, last name and the account. This is basically an event list where we didn't have that much information, but we wanted to figure out the email addresses and contact details and research for all of these people. And so we could use Clay to basically have the inputs of stuff that we know about them. But then let's go find out everything else, such as let's go find the company domain, let's go find the email for these people, let's validate if those emails are valid. Let's then go try and find their LinkedIn profile, pull in their title, headline, summary, et cetera. We can even find out when their start date was at that role. Lots of really cool things that you can do here. You can even go to a website, you can scrape the website, you can find out their modalities, therapeutic areas, and you can do that all on a row, by row basis. And so you can imagine now if you wanted to go and write messaging to these people, you would have another enrichment where then you would put in your, your prompts and all of the information there. So that way you could do that outreach at scale for all of these people, right? So you can write really good personalized, you know, one to one messages at scale to everybody in a list. So another really cool tool that you can use for list building as well and ways that you can leverage AI is this tool called Exa E XA AI and this is their website's capabilities. And the way that this works is you can make a really hyper targeted list of people based off of specific criteria. So in this case we were looking for people that were likely to use Oregon on a chip. And so what we could do is give it a set of criteria such as basically are they working in biotech or pharma as a scientist and then how likely they would be to use Organ on a chip. So maybe it's referenced in their profile, a paper, a publication on their website, whatever it is. And then we also wanted to make sure that their therapeutic area was identifiable. And what EXA will do is literally just go onto the Internet and find as many of these people as possible. So in this case it analyzed almost 5,000 people and came back with 717 matches of people that are like this ideal customer profile for us. Now there's probably going to be a little bit more of cleanup that has to be done. But this is a really cool tool that you can use to get hyper specific lists. If you want a big list, I would say this isn't the best tool. You're probably better off using like a LinkedIn sales nav, a side leads, something like that. Whereas this is going to be like hyper specific. And, and you can create a number of different criteria. So you could even add a criterion here where the company has had funding within the last three months, where maybe you could add in, oh, they're hiring for a specific role and all of that stuff would be pulled in and added to this criteria. And it's only going to match the people for you where all of those things are true. So it's a really great way to kind of custom build an ideal customer profile of like a very specific subset of people. And so if you combine these sort of two capabilities where, hey, I've gotten all of this data enrichment, I can clean up all this stuff, I can write messages at scale, I can uncover all of this data that previously would take me, you know, 10 minutes to try and figure out for each row. Pair that up with creating really hyper focused, personalized ideal customer profile list of people you want to reach out to. Now you can figure out how you can write really good messaging to the right people. And I think that's, you know, really powerful when it comes to how you can leverage AI today. And the next one I want to talk about is inbound lead follow up. And this is a source subject amongst marketing and salespeople. And so you've probably, if you've been at any company ever who's had a marketing team and a sales team, this is what happens when it comes to inbound leads, right? It's this story, right? Marketing will generate lead, quote unquote leads. Sales decides not to follow up on those leads, right? The pipeline is down, right? So there's not enough pipeline to cover. And so what ends up happening is marketing is going to yell at sales saying, hey, you're not following up on our leads. And then sales is saying, well, they're not qualified and they're a waste of our time. So then marketing does is marketing will generate more leads to try and fill the pipeline gap. And now you end up in a situation where it's just like heads butting. You cannot agree on like what should happen with follow up sales is like marketing is generating shit leads and then marketing is saying that sales is not falling up. And it's a tale as old as time. And this happens every company forever. I think every company's had this problem. I've never, I think seen one where this isn't actually coming up. And so how do we solve this? Because this happens all the time. Well, there's a lot of really cool tech now and tools that allow you to follow up with every single lead within five minutes. So I'll show you how it looks. So this is a general flow. So a lead may come into the website and if they fill out a form, we might have something on that form that says, yes, I want to talk to sales. Great, send that one to the sales rep. They can't complain when somebody says they want to talk to sales as not being a good lead. But what if they don't want to talk to sales? What if they just attended a webinar? What if they downloaded a white paper? What if, whatever, just just not really a hot lead right out of the gate right now. But what you can do is then you could take that lead's information, enrich it with something like clay, write that message for them specifically, and then automate the follow up. So then you could put them automatically into a inbound lead, follow up, workflow or sequence. And then only when people respond positively do you then funnel those back up to the sales team. And so the power here is that somebody like somebody on the marketing team can own this and build this. I think a lot of salespeople are great with technology a lot of the times, but marketing tends to use tools like marketing automation. They're more familiar with like workflows and setting these things up. So I think it's a perfect fit to sit in the marketing team. And when this happens, marketing knows all of their leads are going to be getting followed up with very quickly, going to be getting high quality personalized outreach. And then when those people show interest, they'll funnel those to sales. And then sales is getting the hot leads that they can focus on. Marketing knows that their leads are getting followed up on. And again, this is a really powerful system where you can set this up to ensure that you're creating this better environment, right. Where all the leads are getting the touches that they deserve and then sales is only working the leads that are the most interested. And so you create a much better environment around lead follow up. I think there was a Harvard Business Review article. I mean, this was years ago, but I'm sure it's even more, more common now, which is basically the amount of time it takes you to follow up with a lead. As soon as it goes past five minutes, like it basically just drops off a curve. Your chance of converting them, like, excuse me, drops off a cliff. Your chances of converting them. I think the reasoning behind that is if somebody's coming to you, filling out a form, giving you information, like if you don't get to them very quickly, they're going to go off, do the next thing in their day and they'll forget all about you. Right. There's so many other things going on in their life, in their day to day that, you know, they're focused on. And you know, if you're not top of mind right when they're thinking about that problem, that's going to, you know, obviously lower your chances of converting them. And so we can't realistically think that a sales team is going to follow up with every lead in less than five minutes. You know, what if that person is on a call? What if they're out in the field, what if they're at an event? And so you build these systems to automate these processes. You can ensure every lead gets followed up with as soon as possible. And this is probably the lowest hanging fruit that you have in your company right now is simply building out this inbound lead follow up process. And it will save you. You know, actually I should say it will probably just naturally convert way more leads and build way more pipeline. And you're probably honestly better at doing something like this than you are trying to like spin up a whole bunch of outbound prospecting campaigns. Because these are hot leads and they know who you are. And prospecting is typically done to people who have no idea who you are. And so optimize your inbound processes first. And then that way when you do outbound and people come inbound, you are then able to, you know, handle those leads appropriately. Cool. And then, you know, again, I think kind of hit on all of these different things. But the idea here is you're going to be able to handle more leads with less effort and better. Like you're going to have better conversions and you'll be able to do it with more leads, and it's going to take less effort from everybody across the team. So big win for everybody here. All right, the last one here, some different creative ways that you can stand out and leverage AI to do that. The first one, this is actually funny. Nick and I were supposed to be on a call talking about something. I can't even remember what it was at this point, but we got stuck on this tool called Suno, which is an AI music generation app. 1. It's a ton of fun, so I'd highly just recommend you do this for fun yourself. But what you can actually do with this is create custom songs for prospects. So you could write a song about how your product or service or whatever can solve a problem that they might be facing. And I guarantee you nobody else is doing this. So if you were to do this and send it to somebody, it probably would be really funny and really interesting. You'll probably catch their attention. And what's cool with tools like clay, you could probably even do this at scale where you could, like, write a song for every prospect on a list and, you know, have it sent to them. You basically write the the script or write the lyrics, you know, choose the genre of the song and basically you create it for everybody. But you could also just do this for your top prospects that you really want to get a meeting with. Maybe you haven't been able to get through to them. This would be a really cool way and unique way that you could stand out. And if people get, you know, annoyed that you send them an AI song, I think those aren't very fun people. So have some fun with it. Prospecting's hard enough as it is. Another fun thing that you can do is you can also then create AI videos or AI avatars of yourself, you know, actually doing things out in the world or whatever you kind of want to do. You can get really creative with how you make these videos nowadays and then use that as a way to share information with clients. So if you don't feel comfortable making videos yourself, this is a great alternative to where you can make, you know, an AI video and I think lean into the fact that it is AI. Don't try and make it look like it is. You, like, make it very obvious that it's AI generated, because I think that's just going to be more fun and more interesting. Whereas if you try and make it look real and you don't like, and you don't address the fact that it's an AI, Video people might get kind of turned off by that, thinking you're trying to bait and switch them. But if you made something specifically for them using AI, I think it's a little more interesting. And then the last piece here is this tool called Replib, which is you can also use another tool called Lovable. Basically, the idea here is a concept called Vibe coding, where you can essentially use natural language to create software. So we've made a couple of these tools ourselves that we've shared, like the email grader Insights tool. We built a few internal apps that we use within our team to help speed up different workflows as well. So really cool way that you can just build software simply by using natural language. And so you can imagine maybe doing this for, again, some of your top prospects, where you make like a lead magnet or you make some sort of app that would be really cool that they would benefit from. So it's another kind of unique way that you can stand out and create content, create engaging things that are different from what everybody else is doing. And then you can also do all of this with images as well. So Nano Banana is a model from Google Gemini. It's probably one of the better image models right now. So you can kind of make custom images as well that you share. You know, there's another way to stand out. Rather than just sending the same data slides that everybody does, you could create some interesting images, kind of, you know, comma. Again, we're just trying to find ways to stand out and get in front of people, get them to respond. And these are just some different ways that you can do that. And here's some examples that we've built. So here's that Insights tool I mentioned, where you can basically just plug in your company domain. And what it'll do is it'll spit out overview the different products or services that you offer, how you can position them better, what problems you solve. Some market analysis let you pull in companies that you can potentially sell to prospects and people at those companies. We'll write some prospecting emails for you and even draft up some sales discovery questions. And so we put this together as a lead magnet that we could give out to other people that are like in commercial. Because for us, this is a way to showcase these are the types of things that we can do. And so it kind of makes sense. So you would want to think through what that looks like for your customers. Another one that we put out is one where you can find email addresses. So you simply just put in somebody's name and domain and it'll, you know, find their email. If it exists out there, it will go and try and find it and validate it, give it back to you. And then we also have our email grader which is trained up on that 27 page document I mentioned on how to write good emails. And then you can simply, you know, put in your email here, paste it, and then actually have AI analyze it, give you feedback, excuse me, feedback and coaching on how you can improve your messaging and then even have AI rewrite it if you want as well. So these are examples of some of the cool tools that you can use and we've used these as lead magnets to offer out to the community. They're free to use. So if you want to use them, go do it. I'll share the links for this stuff as well. All right, so just to quickly recap everything here, the Rapid framework is make sure that you recognize the goal. So what is it that you're trying to achieve when you're using AI? Figure out what that is and then make sure you are writing all of your prompts in a way that are going to drive you to that goal. Then you need to articulate what good looks like. Give lots of examples. Remember that 27 page document we use around messaging? If you want to articulate what good looks like from a account research perspective, again you could do that. If you want to tell it what a good meeting, follow up email looks like, you can do that. But make sure you articulate what good looks like and document it so AI doesn't just give you the average response that doesn't look good. Third, you want to make sure you provide context. The more context that you provide, the better. Feed the beast, improve the output. Don't just send or share or publish the exact thing that AI came up with. Make sure you edit it, you tighten it, you fact check it, make it look human and then deploy and document. So put these things out into the world, use them, but then document what you did. So document the prompt, write them down, save them, or turn them into GPTs, agents or skills within Claude. So lots of different ways that you can do this, but just make sure you do it so you're not doing the same thing over and over. We want to do this once, do it well and then that'll save you time every time going forward. And so the last piece here, these are the different workflows we covered. Prospecting, messaging, meeting, follow ups, account research, data enrichment and list building, inbound lead, follow up and some creative ways to stand out. Each of these are, I would say, great things that you could deploy right now into your workflows and the things that you're doing. But what I would say is pick one. Don't try and do all of this all at once. Pick one that you think you are best fit to be able to do with the skill set you have using AI, and then by doing that, you're going to get more comfortable with using all of these different tools. And then you'll be able to then pick off more and more of these over time. So with that, here are a couple of ways that you can get some more stuff from us. So if you want to subscribe to our newsletter, you can. And then also we have a podcast as well, Sales DNA podcast where we interview life science, sales and marketing professionals, founders, et cetera, and we talk through different sales strategies, techniques that you can use. And that is it. Thank you all. Hope you enjoyed it. This again was a presentation that I gave the other day at ELRIG out here in London, or actually it was in Liverpool. But yeah, it was a good time and hope to see you all in the next one.