E55: GTM AI Advice from a Salesforce & Siebel Veteran (Peter Grant, CRO @ You.com)

Peter Grant is the CRO at You.com, and he is hunting for a legacy.

I literally sat in front of him and said, I want to take you from $1 million to $1 billion.

Peter knows that you can't build a billion-dollar company with the old playbooks anymore, even if they worked just a few years ago.

I think the whole go-to-market org got really lazy and sloppy and fat.

And he sees firsthand how other GTM leaders are winning with AI.

What you're hearing is double my revenue, not my headcount.

The warning is clear: adapt to this new reality or get left behind.

You should be spending roughly 2% of your total revenue right now on generative AI and growing it.

Welcome to the Revenue Leadership Podcast. Here's a familiar scene. A deal's in commit and the buyer says, can I talk to somebody like me? Next thing you know, sales reps are asking CSMs after hours. 5 people are in Slack trying to remember who helped last time. And that one customer who always says yes is getting burned out after 3 requests in a week. That's why this episode is brought to you by User Evidence. The platform that turns customer proof from chaos into a repeatable GTM system. With advocates, references, and real customer evidence all in one place, your team can share credible proof instantly and keep deals moving forward instead of stalling. Companies like Gong, Ramp, and Workato already use User Evidence to capture customer proof, build buyer confidence, and shorten sales cycles. See how at userevidence.com. That's userevidence.com. Today's guest is Peter Grant, who is the CRO at u.com, which is the AI search infrastructure for the enterprise. And Peter's had a legendary career. He was an early leader at Siebel, then Salesforce, uh, and has had a bunch of CRO roles at C3 with Tom Siebel again, uh, OakNorth, Tideway Systems before joining, uh, u.com. to run revenue, and we are gonna talk AI transformation today. He has a front seat to this, uh, as an AI-native company serving other AI-native organizations. And, uh, we're also gonna talk about some counterintuitive opinions on strategy. So lots of fun things and, uh, excited to have this one. And, uh, thanks for joining.

Carl, it's great to be here. Thank you for the opportunity.

For sure. So, Uh, I gave people the quick snippet of the background, but, uh, what did I miss? What do people need to know about your background? Maybe like just tell people how early you were at, at Siebel and Salesforce, cuz I think that's a ton of fun.

Yeah, I think through my career what I've tried to do is be at the precipice of change. So I joined Siebel when it was client-server, not old enough to be mainframes, but lots of the people I was working with there have been selling mainframes before. Now it's client, client-server, so you're putting computing in the hands of obviously the end user. And the precipice for that was Tom was basically, was the evangelist to making a market for CRM. So having a 360-degree view of your customer. Back then it was very siloed. People ran on spreadsheets, ran with Goldmine, these small systems, and Tom realized you just didn't have a global view of what your customer was doing, right? So whether it be from a sales point of view, customer service point of view, marketing point of view, and bringing all that together. Sounds old hat now, but that was quite revolutionary back then. Uh, we were $39 million in revenue. We'd just gone IPO as I joined. And as an ex-military person, I thought I was back in the, in the army. The way we executed was ferocious with our processes, the quality of people we hired, the training we put people through, the enablement and everything else. To put that into context, we went from $39 million to $2 billion in revenue in 4 years, $68 billion market cap and 8,000 people hired. So if you're going to do that, you've got to be very good at process and execution. And we were ferocious at doing that. So that's a really exciting time. Uh, my first year.

$39 million to $2 billion in 4 years.

Yeah, exactly.

Those are Gen AI numbers.

Uh, they are Gen AI numbers.

They are back in the '90s.

Yeah, exactly. But obviously that was perpetual software back there before obviously the advent of Salesforce coming along. Mm-hmm. And, uh, bringing that utility computing model to the enterprise. Yeah, we, we basically built a market which is a CRM marketplace.

And, and so when do you start at Salesforce? How big was the company then?

Yeah, so when I started at Salesforce, I think the revenue was about $25 million. Um, it was mainly a commercial operation then, not selling into the enterprise. I was employee number 9, um, in Europe for enterprise sales. We had, we did have a group of people in Ireland that were doing an amazing job for commercial sellers. And I think the company is about 300 maybe in total. But then I basically built out the UK business as UK MD. I was the second only country manager behind a guy called Johan Zietemam, who was in the Nordics. But that was probably the best fun that I've ever had in my career. Because now you're a gamekeeper turned poacher. So we were basically attacking Siebel. We had this brand new technology model, which back then was called utility computing. but the most important thing was a new business model. So rather than buying all your software upfront, you would take it like a utility on the trip effectively and pay for it. That's really the marketing message that we were saying. And Mark was really clever at being a storyteller and saying, you know, when you build your house or a hotel we're standing in today, you wouldn't go and build your own sanitation. You wouldn't build your own power generation. You wouldn't build your own, you know, everything else as well. You just plug into like a utility and then you pay for it as you consume it. And people instantly got what this new model was basically. And, but it was hard. It was really hard. There were 3 of us in the UK that were sellers. I split the UK out literally. I mean, it sounds very archaic now, but on the alphabet for the accounts we're gonna go after. Mm-hmm. We looked what, who was the early adopters in the US. It tended to be the tech companies. So we targeted them first, uh, in the UK and we just went at it, right? And unlike coming from a traditional company that IPO'd in the sense of Siebel and having these quarter ends, every single month was a quarter end. So the mindset you had to change with the sellers is that now a month becomes a quarter. That means a week becomes basically a month, right? Yeah. And a day literally becomes a week, right? So that's the ferocity that you were having to work at and the cadence you were trying to get the deals in there, uh, through the door. So we were doing 18, 22 deals a month. You know, in the early days and like the ASP was really low. It was like $28,000. People still can't believe this. And in the commercial side, it was $4,000, but the sales cycle was 3 or 4 days. So I remember my friends at Siebel saying to me, you're mad. Why have you gone there? You can't make any money. And they couldn't see the future vision of where the industry was effectively going. Yeah. But it was a really exciting time.

And you would have made some money. So they were wrong.

Yeah. Yeah. Yeah. It was great. Do you know what is money is a byproduct of having a sense of purpose. And what we did is, I believe at Salesforce, Marc created this really compelling vision around no software or the end of software. We got up every single morning and went to work to basically end software, right? This completely new delivery model, technology model, and new business model. Now, of course, there is software involved, right? But it basically bound the whole company together, a single myopic focus that we have to execute.

That's awesome. And so now let's fast forward 15 or so years. You're on the precipice of another massive technology wave. You're quite a picker, Siebel and Salesforce. I would love to understand about like why pick u.com? You know, as somebody with your profile, you could have gone and done a bunch of things. You've zeroed in on this, this phase change, this platform change. And so tell how did you think about inserting yourself into not just like you.com, but this part of the stack from like infrastructure or even like chips to infrastructure to the app layer? Like how did you think about that decision process?

Yeah, that's a lot, a big question to unpack. I would say they chose me. I wanted an opportunity. So if you take a step back as a salesperson, you want to be with the best technical people. Right? And if you look at this industry, there are lots of wannabes out there. There are only a handful of people that have had a profound impact on generative AI. So my boss, Richard, Richard Socher, is the third most cited NLP research scientist in the world. He's a Stanford professor. He was the EVP and chief scientist at salesforce.com, along with Brian McCann, who's one of the leading AI research scientists in the world. So together they worked on contextualized word vectors, which is basically breaking words down into numbers and doing predictions of what the next word's gonna be, then what the sentence is gonna be, what the paragraph is gonna be. So you can see this is the precursor to basically large language models. So their work and research there in NLP has been transformative in where we are today. So that's point one. Point two, they then released a paper called the Decathlon paper in 2018 that said one AI model can do literally 10 different things. So it can generate content, it can do translations, it can summarize the information by using question and answer. which is what you're now seeing how you query a large language model. So again, that paper was so profound that when OpenAI read it, they stopped doing what they were doing around gaming and switched to large language models. And they credit Brian and Richard for their research in part of that strategy change that they had. And then finally, Richard saw that large language models obviously hallucinate. It's a big problem. It's a big problem for humans to accept the content that's coming out and why there's been some rejection to them. So he's, he has a patent pending for marrying a large language model to a web index. So a long way of answering your question, Carl, but basically you want to choose people that know what they're doing in this particular space. And there are only a handful of them in the world that do that. So I was lucky enough through Mark and through a guy called Carl Schachter to get introduced to Richard. Now I knew Richard because of Metamind and Mark had made that acquisition a number of years ago. And then Richard's obviously what he achieved at salesforce.com. So I jumped the opportunity. right? And I literally sat in front of him and said, I want to take you from a million to a billion. I want that to be my legacy, you know, as a CRO in what we're doing and what we're building here. So that's, that's how the journey started with them.

And you joined that early, like a million ARR?

No, they were, they were a bit north of that, but they, but, but this is, this is where we've done a pivot and we've done a change. So originally they're an AR, but basically Rich's vision is that search needs to be transformed. It's not these horrible blue links that you see on Google now. It needs to be a generative AI output. Yeah. But it needs to be trustworthy. It needs to be transparent, right? And most important, needs to be really, really accurate. That's never left our DNA of Effectively as a business. We were focused on the consumer side. If I'm perfectly frank with you, our marketing wasn't particularly good. Our PRG motion, you know, we got beat by Perplexity, got beat by OpenAI, and lots of people are now struggling with this OpenAI blast radius. So 18 months ago, we decided to go, uh, to enterprise, right? In this consumer space, although you can grow really fast, you can have lots of attrition and it's like, flock of sheep at the moment. Everybody's like jumping models and suppliers and, you know, you're seeing 40, 70% attrition. You can't run a business on that. So we made a conscious decision to go enterprise. We hired much more enterprise blood into the business that really worked from the customer problem backwards, but also most importantly about how can I make you successful? Yeah, that's the number one question you have to ask yourself when you deploy these systems. So I joined 18 months ago, the revenue is about $5,000 and now we're many tens of millions of dollars on a significant run rate, uh, going into 2026.

Very cool. And so what does u.com do exactly for enterprise? Like, what's, what's the simplest version?

I think, yeah, we do 3 things. So the first thing we do is every single search you do, we ground it in the truth by linking it to a web index. So if you're on OpenAI or you're on any other chat agent, right? So when I say agent, I'm talking about digital agents now. So digital workforce. If they have a request and they go out to the internet, you need to make sure that that LLM's not hallucinating. So you can take our web API, you can ground it in the truth, and it'll bring back what's called a citation, i.e., the website, and then a snippet where it found that information that's relevant to the query. So that's the first thing that we do. But the internet is like a mile wide and an inch deep. Yeah. The way it gets scraped and indexed by Google and other indices is that basically it's quite a high-level scrape. It's pretty what we call shallow. As we move forward and we deploy more and more agents into the world, they need what's called context, right? And they need to be domain specific, right? So whether you're working financial services, healthcare, legal, travel and hospitality, e-commerce, et cetera, you need to be an inch wide and a mile deep, and you need to understand the user intent. And that's what we're building out now, which is the next generation internet that's basically for LLMs and not for humans. The internet is designed for what's called keyword search, right? It's designed for navigation, product, commercial search, local search, or domain search, right? So LLMs don't sit well with structured data. They're more tardy in their responses and they need more conversational stuff, which is why everybody's looking at this GEO, changing their websites, not for marketeers now, but for LLMs to appear higher in the rankings. We're doing exactly the same. But for agents that have specific domain tasks in those industries that I, or those verticals, sorry, I was talking about. And then the third thing we do is we target companies that don't have the AI skills. They can't attract these amazing AI research scientists that we can do at U.com. And we package everything into what we call an AI factory. And in there you can use our chat user interface. You can do what's called private RAG, and that's now retrieval over all your internal content. As well as your external content and marry the two together for humans and for agents.

And so what's the use case? Like, like if I'm a, if I'm a CRO or like who's the buyer and what are they, what, yeah.

Yeah. So if you're a CRO, the best way to think about this is your salespeople. If you break down the sales process, the first thing you're gonna do is a discovery call, right? So you think about how much time do you take to do a discovery call? Salespeople generally only spend 23 to 27% of their time actually customer facing. Because they're doing so much research, internal meetings, and updating CRMs and everything else, right? So the first use case is using a product like Ari, which stands for Advanced Research and Insights, and basically say, go out, I've got a meeting with Tim Cook at Apple, go and find me everything about what's happening at Apple, right? Go and read their annual report, go and look at their strategy, go and look at their 10-K, right? Go and look at where they're making their investments, right? Go and look at all their publicity stuff. And then an agent can be built to do this and come back literally within 5 minutes and give you a very detailed, comprehensive report. So it's doing two things. One is saving you time, but more importantly, it's giving you a broader aperture than you could ever imagine, really quickly bringing it together. So it's reading 1,000 websites in 5 minutes and then deciding what's the right information to bring back for you. So from that point of view, this goes to tactics dictate strategy, we'll talk about later, but that's part of being really, really really well prepared for a sales call.

Mm-hmm.

Second use case is you can use the technology to say, okay, imagine you are now an Apple customer. How would generative AI affect the, and impact Apple to grow their revenues or to, for cost savings? And you can then teach it to be like a output, a PRD, a product requirement document. So then it would, it would do the persona for you, it would do the different use cases for you. Then we put, we could feed some of that into lovable so we can show a customer what the art of the possible is. Now you're moving the goalposts and you're not just being a normal sales guy turning up and just doing discovery and trying to work out what they do. You're actually coming with solutions and ideation to the customer to get them really excited about how they can use generative AI, whether it be internally or externally to drive their business forward. This is a couple of use cases.

Okay. Makes, makes a ton of sense. So as I mentioned, you've, you've been on the cutting edge of a bunch of these platform changes and now you're in the thick of probably the biggest one yet. Yeah. As you're working with these companies, what do you see as, as fundamentally different in sales and go-to-market in this AI era versus like, versus, versus previous platform changes?

Yes. Speed and scale. Literally those two words, right? So speed. Is the decision-making is a lot faster. So our ASP is around $200,000, our sales cycle is around 87 days. So it's not that traditional, yeah, it's not that traditional 9 to 12 months that you would expect for enterprise. So what you have to do, you cannot bring your old mindset into these new marketplaces and say, I've got a playbook, this is how you do things in the SaaS world or how we did it before. I've had to learn really fast. That you have to be able to adapt to change. Then if you think about speed, is my whole point, right? The sales cycle is being compressed, which means you need to get to the point a lot faster. There's so many other people out there saying the same things, AGI, agents, private RAG. And so you need to really differentiate yourself from your customers and you need to show the value you can bring and obviously you're going to make them successful. So I think that's really important. And then the scale of these things. So one thing I learned very quickly is that if you sell CRM, you probably sell to 40% of the user base, Kyle, because most people in the go-to-market world, sales, marketing, customer success, right? Some rev ops and everything else, they all need a Salesforce license, a Gong license, you know, whatever it may be. With generative AI, the solutions that we're deploying for the agents across the whole business, back office and front office. So therefore, the yield you're getting per customer is far greater than ERP or what you— we've seen historically in CRM.

Yeah, the— it is really breaking all of the previous models that we've, we've known about, especially like even from a venture capital perspective. Yeah, you know, a product that is now going to be not only touching every single employee in a company but in a profound way. Like, you know, it's like AWS. AWS went from being this like commodity where we host things and now it's people's very biggest line item cuz everything is hosted in the cloud and everything is gonna be GenAI powered in the not too distant future. So just the scale of outcomes really like breaks your brain.

It does too. And I, you do, and there's, there was that MIT report that came out, you know, $30 to $40 billion's been spent on generative AI, but only 5% customers say they actually have a definitive ROI, partly because a lot of point solutions. So, I mean, this is pretty typical in an early market, right? Everybody's rushing to try things and everything, which is a good thing, I think, because they're learning really quickly. Number 2 is they weren't embedding into workflows, and they actually found more of an ROI in the back office than the front office from a productivity point of view. But we can talk about some of the sales impacts it has. But number 3 is really interesting. Number 3 is going it alone. There's a massive failure rate as opposed to working with a supplier and using the forward deployed engineers. So everybody's talking about this now. Everybody's talking about the Palantir model. FDEs are getting paid a fortune, right? They're really hard to get. But what does an FDE do? They bridge basically the product and the customer's need, and they bridge that gap in between and they deploy their solutions. Basically for the customer. So that's where we're seeing the biggest opportunity is basically providing a complete solution in a marketplace where, with all due respect to some customers, that some of the leading AI research engineers wouldn't want to go and work in those particular companies, but they will come to organizations like you.com. So our job is to put some governance around that with the AI factory and show them how we can design and build and deploy internally and externally, stack rank where the ROI is coming from, what departments we should go for. Yeah. And then rapidly get things out into the hands of the end users or to the customers.

That's such a great point. And you're seeing, you're seeing this everywhere, this, this blending of software and service. You know, Palantir was, was consistently mocked for being, you know, a services company, a consultant dressed up as a tech company with tech multiples. and now they, you know, surprise, surprise, Peter Thiel was right. And, and everybody is, is rushing to, to follow suit. I, I want to go back into this like workflow question because, um, this is something I see a lot. That MIT research, 95% of AI implementations have, have no discernible ROI. What do you mean by they're not embedded into workflows and why is that a bad thing?

Yeah, so lots of point solutions, right? There's a lot of shadow IT. So there's lots of people just building their own individual agents, right? But they're not looped into your complete, could be your, you know, lead to cash workflow, for example, right? So I'll give you, let me give you a customer example as a better way to answer the question. So the Telegraph, which is one of the world's leading publications in London, yes, that we're deploying agents to all their thousands of staff and everything else, and they can do research with them, all the normal stuff that people talk about. We just argue ours is more accurate for them. But two of the things they're doing that's really cool is then what I mean by embedding into the workflow for the— they have 5 of 1,000 staff, 500 journalists, and when they write an article, they push it to the CMS, the content management system. Then the editor sees the article, they edit it, they're happy with it, they push it out for publication. They often make mistakes, right, through their research, right? And that's one of the biggest complaints they have. And this Telegraph has identified this and they want to address it for their readership. So what we're doing now is our web index is now becoming a fact-checking agent for them. So when they write the article, it was, it's constantly scanning the article and going out to the web where they might be citing people, names, numbers, ages, statistics, whatever they are, to make sure that they are factually correct. And it will come back and give a scoring on it. And if something's wrong, it will say, hey, this is actually the correct information. You need to correct the article. The second thing it's doing is it's basically editorial guidelines. So when you're an editor joining Telegraph, you have to go through certification, understand the guidelines, make sure you're in line with the political policies they have, DEA policies, all those sorts of things. Again, we train the LLM on their editorial guidelines to make sure they don't fall foul of them with our editor. And that, again, that improves their process, reduces obviously the error rate that they actually have, but that's showing how it's actually embedded into the editorial workflow as an example.

Yeah. And I've, I've given this talk a few times about how go-to-market leaders can transform their, their sales teams. And one of the, one of my biggest pieces of advice is no new surfaces. Bring, bring the agent, bring the workflow into the, into the surfaces that your reps currently live in because you see so many tools. It's like, oh, we're, um, a sale, a pre-call research thing and we're going to build you a sales playbook, but you have to log into another system and you just don't get the adoption because it's, we already have too many surfaces.

Yeah, I agree.

CRM plus your sales engagement tool plus your email plus Slack plus your texting customers and prospects. It's, it's, we already were beyond the saturation point to, to actually think you're going to have a new surface adopted well. And so I think this makes a ton of sense. Yeah.

And it's interesting you say that because people say, well, what is an agent? And there are two types of agent. Yeah. So one is a research agent. So the example I gave you, Ari, is a very sophisticated agent. The other one is an action agent that carries out actions on your behalf. They're more agentic. I mean, they're both agentic, right? So, if I'd ask Brian, you know, what is it, what is it, what does it actually mean to be agentic? So an agent can, will go, if you, we think about left to right for workflow processes and is completing those. But with agents, they go down into LLMs and autonomously do actions for you. They rewrite your query multiple times. They do chain of thought. They do reasoning, right? They decide what data you need to see. That is still an autonomous agent, even though it's doing research. Some customers we're working with want to do bookings, right, for their guests, for restaurants, for hotels, or whatever it may be, and completely do that autonomously with your credit card. Card based on your preferences and everything else. That's exciting stuff that we're working on, um, right now. However, it's embedded in the mobile phone, it's embedded in the website, right? So to you and I, it's just an API call, an SDK, right? And it's basically building on the top of that. It doesn't look like an agent that we're used to seeing in GPT or u.com or whatever it may be. So it's that the surface area is completely different.

I jokingly say My new definition of agent is anything that's AI is an agent now. Because I, the original definition of agent when people started using it was an AI system that can use other tools that can decide to use tools, call to them and sort of like make decisions through like that operator agent and go determine what things to access and how to solve a problem. And now it's like anything that's sort of AI adjacent is like, oh, this is my agent. And so I've lost that argument. I want to go back to this comment around $200,000 ASPs on 87-day sales cycles. So what is different about a sales cycle now and on both the customer side and on the vendor side? How does that happen?

Yeah, I think the level of investment that's going into this space is unprecedented. So, um, if you think about it, right, so the, the World Wide Web came along in '92, '95 is starting to get traction. Uh, you know, we, we had all different search engines and suddenly up pops Google and their number one metric was latency, right? And they actually published it. The engineers almost to, you know, you know, the leaderboard of shame, right? If they weren't below 250 milliseconds, it completely changed the mindset within Google to get that really fast response that propelled them to 100 million users in 5 years. ChatGPT comes out and in 60 days or less it gets to 100 million users. Now at 800 million weekly active users, right? It's 30 times faster. Why do I say that? In the 30-year period of the internet, this digital platform, half of the Fortune 1000 has disappeared because it didn't transform fast enough. So now going back to your question, Carl, we're seeing significant investment in the last, because only the 30th of November, 2022, right? So only 3 years into this, well, literally 3 years into this in 2 weeks' time.

Almost to the day. Yeah.

Almost to the day. And yet the spend already Already is at $30 to $40 billion. So what we do is we track by vertical where you are against your peers, right? So you should be spending roughly 2% of your total revenue right now on generative AI and growing at 30% compound growth up to 2030 if you want to keep up with your peers and you want to be a GenAI-first type organization.

So, so go to 2% of revenue.

Yeah, 2% of your total revenue. So, so some laggards are at 0.8%. The tech sector's like a 3.5%. They're leading the space right now with the investments that they're making, right? So you've seen that through Meta, Salesforce, Oracle, all those people and what they're doing there. But that's, that's roughly, so 2% roughly where you need to be on the spending. So how's it changed things is you don't have to educate people. You don't have to make a market. When I was at Siebel, we had to make the markets a really expensive thing to do. We had to do exactly the same with Salesforce because we were the evangelists. in this particular space, the market's been made for you. This is all about speed of execution, right? Because the money is there. People are making funds available, right? You know, they're replacing a lot of existing systems. They're also experimenting with a lot of vendors. So the other thing you see is unique, as I referred to before, is you might find yourself with 3 or 4 competitors in the same company, all doing different use cases, even though you could do it, right? Because they haven't worked out yet who they're going to bet on. In the future, right? So part of our value prop is we have 40 LLMs at the back end because we don't know who's going to win that race either. Yeah. Yeah. Yeah. So, so, and the, the model changes in terms of cost performance every, every, exactly. And different models do different things for different use cases. So we route to the model we believe will give the most accurate answer for you. So, so first of all, that's why you see, you're seeing the levels of investment there. That's pretty serious.

I want to ask one more follow-up here. Why has the AI adoption been so much more ferocious than cloud? Like, what, like, why are people so easily convinced that we need to adopt AI, but cloud was so different?

I think PLG has changed that, right? The product-led growth. I, um, we have never seen revenues grow this quickly, this fast. Yeah. So even though it's been pretty extraordinary. What our team here has done at you.com, if you then go and benchmark yourself against some of the other companies that in 5 months got to 50 million or got to 100 million or going to 200 million, that's, I know, or half a billion, right? I mean, Cartel were talking at a conference I was talking at on Monday, and I think they said the average Gen AI company now is getting to $2 billion in 2 years, which is freaking crazy, right? So, so that's, but, but they are PLG-led models, right? Yeah. So that, that's really important to understand that, that's the big change here. So you're seeing the consumerization accelerating into the enterprise space, but it goes back to the single user as well. So what's different in here is we launched our product called Ari last, sorry, everything blends in. I can't remember. I think it was this year, right? So in Q1, this Q2 and Q3. End of Q1 this year. And we provisioned 500 people to go onto the waitlist. Yeah, we got 22,000 people sign up. Right. And they used to, because we came from prosumer, we always had a personal email. We said, for this to sign up, you need to give us your work email address. We have Mercedes-Benz, BMW, Accenture, Salesforce, JP Morgan, every bank under the sun. We're like, oh my God, they're in this, this is our many millions of mail. That are using the technology. So you think, what an amazing lead machine. You get really excited, then you go and contact them, you find they're all single player. And that's what's happened. That's why even though ChatGPT's got 800 million users, only 35% actually use them for business. Most use them for personal. So I think that will start to settle down, Karl. But that speed of adoption is what's been driving this effect. And user can get to it. And remember, this is not complicated. It's conversational. So now you're having a conversation with data and the data is talking back to you like a human would. And I think that's the other reason why the technology's been adopted so quickly. So I personally, I wouldn't say I'm a great technologist. People work with me who would definitely smile at that comment. But one thing I did get really good at was understanding how this could help me in my work and how I could build an agent and how I could become good at prompt engineering that makes me really effective. And now if, if you took it away from me, I actually do not know what I would do.

I would have to do 50% less things if you— this podcast would certainly not happen. I mean, you, you went through my flow and you know, it's like you've— the guest fills in the intake form, it webhooks into my Clay table. Clay does all the research for me. I put in the topics I want to cover, then it populates a bunch of questions and I curate it. Like that would've been a 4-hour manual process. Of researching and whatever. And it takes me, oh, maybe 20 minutes if the first output is good to go in and, you know, like put my opinion on it and pull out the stuff I think is obvious or just like not relevant and curate that docket. But absolutely.

So yeah. And to answer that, I think the first part is it's really easy to get into the end user's hand and for them to understand how they can use the technology. Technology. The second—

feel it as the end user.

Yeah. And you see the— exactly. You see the output. You're like, wow. I mean, incredible. The second part is, to your point, is that anything that's a mundane, repetitive task lends itself really well for an LLM. Then humans don't have to do that, right? So when they see the money, you know, the light bulb goes off, as you've just articulated there for doing your role here. They're like, oh my God, This is just so much easier. And again, that just drives the engagement. Having said that, I do feel we have an AI literacy issue, generally speaking, across the broader population. So we sit here in the Valley, 90% of where AI investments are here. Every billboard you see on the 101 or the 280s is AI. Everywhere in San Francisco, every bus shelter's got somebody AI with, or even We've got like 54 sites, I think, out there right now. That's not necessarily the case as you start flying east. Yeah. So, uh, I think we have a responsibility to educate people how to use these systems, how they can make them more productive as well. And there's a lot of research out there. If you look at top sellers, the best ones who engage with AI are significantly more productive. And definition of productive, shortest sales cycle. Bigger deal sizes, right? And blowing out their numbers compared to the laggards who are not embracing it. And I remember the CIO of Minecast saying to me, he said, Peter, if you are not AI literate in the next 2 years, you're gonna be obsolete in the workplace. And that, that thought never left my mind. So then Richard and I, when we were going around seeing customers, we were asking what's the engagement layer like, level like, excuse me, if they've got taken an API from a vendor and they've built their own you know, UI on top and everything else as well. They go, oh, you know, you know, we get 1 or 2 queries a week from users. And like, it ranged from 4% all the way, the best we ever saw was 52% from one of the world's top consulting firms. And we thought, hmm, that's an issue, right? Because what you're going to do, if you sell by per seat, you're going to get a huge amount of attrition in your next year when you come up to the renewal. So we did 2 things in the company. One is we partnered with a company called Pear and we embedded their training into our application to take people from fundamentals all the way through to mastery. But it's role-based training from the LLM. It's super smart. So you can say, I'm Karl Norton, the CRO at owner.com. It will go out and find other tech companies of your size like that, and it will give you specific tasks to do using the user interface, right? So as an example, Customer calls in, they've had a P0, you're not too sure what you should do. So you get the team, you whiteboard everything with a mind map on the board, how you think you're going to fix this particular problem, how you should respond, how many updates you're going to do, need an approval from the CEO. So what do you do? You just take a picture of it, load it into the LLM and say, write me an email for approval to the CEO because you've done, you've gone and it literally does it for you straight. I didn't know you could do that, but there's a list of learning things you could do from an efficiency point of view. So that has been a game changer with our users, right? We track every user, can't see what they're doing, but the amount of queries they do, what training they go through, and you just see this huge lift in adoption. The second thing we did is we price based on usage. And then what we did is, because you have millions and millions of monthly active users, we could see from a bell distribution who the power users were and where the lower users were and where the medium was. And then we price for the median and we say to you, to companies, look, if you've got 30,000 or 10,000 users, this is how many queries we think you need, right? Buy this first of all, get going, get the training and everything else, lift everybody up in the enterprise. If you need to buy packs some more, come back and buy them, but don't overbuy them because we don't want a nutrition problem. We want you to be successful. And that mindset has changed the way we sort of go to market. And the way we sell and basically position with our customers, because they know we're in it for the long run and we're white-gloving them and we have to— they have to be successful.

And so how should revenue leaders think about driving adoption? So like, it's one thing to have a vendor offer the training to people, but like, how do I as a CRO make sure that people are spending the time and paying attention and transforming their, their your own jobs and workflows? Like, what have you seen the best companies do to transform, as you call it, AI literacy or just like general adoption?

Yeah, so we track everybody's gone through training for a start, so everybody's on the same level. We then track everybody's usage pattern, right? For what they're doing, the amount of queries they're doing. Then we create leaderboards around that, Karl. We then look at the agents that they are creating as well. So the average in human human will create anything from 15 to 20 agents that they will have running for all different parts of their processes as well. Interesting. So you look at the agent creation, so that goes back to the training they've learned to do that. And then you create leaderboards between agents and humans, right? And you stack rank to see what the, you know, if you've got SDR agents against human agents as well. So you see the volumes being sent out, the returns that are coming, the leads that have been generated and everything else as well. And then you can often link your normal sales stuff, right? So lead gen pipeline, close rate, you know, customer sat scores and everything else back to doing that as well. That's what I would personally do. Uh, because we're moving so quickly, we have literally a daily standup and we share everything we're doing on deals, what's working, what reset, uh, it's like engineers have a standup. I do exactly the same with the sellers. What 3 things did you do yesterday? What 3 things did you do then? What problems do you have? What needs to be unblocked? And that again, everybody's sharing what they're doing, new agents they've built and what's happening, like building proposals for doing an ROI calculator, tracking usage of the customers, tracking all their training they're doing and stuff. You know, there's many different things you can do, but the, I think the takeaway from this is like creating, still create those leaderboards, but link them to AI adoption and then hopefully the revenue lift that you're seeing as the outcome.

Makes sense. I want to go back a question we were talking about. How sales cycles have changed. And we talked about the consumer side, just massive amounts of pull. What about on the vendor side? So how do you manage a team different today? How do your reps sell differently in this type of environment?

Yes, good question. So I think there's a different rep profile that's emerging. And I think that rep profile is more of a pre-sales person. They're more technical, right? Because especially as we move into more of an API-first economy, To your point earlier, I think there's a lot of user interfaces and surface areas. I think they're gonna die away, a lot of them, and you're gonna be embedded completely into the workflows. And the number one issue you face is people still don't know what this technology's capable of. And to do that, you do need to have a technical mindset. So some of our most successful sellers do have quite a technical background. They have an API-type background and they understand what can be built on top of the API. So I think that's, I think that's becoming quite generalist across the industry. So to the younger sellers out there, you need to get more of a technical background, understand Python, you know, especially if you're entering the workplace. And then to the older guys, right, you need to be basically investing in yourself now and being amazing prompt engineers and understanding what you can build on top of these APIs. So as part of our recruitment process now, we are giving a little video to show what our reps can do. What the art of the possible is, the API, and then we're putting it in the hands of sellers that are coming to us to see what they come up with. And some people say, I'm not a pre-sales person. Say, well, dude, that's how this market works now. Yeah.

Yeah. Does, does pre-sales go away in that world?

No, pre-sales becomes a bit deeper technically.

Okay.

Right. So I think to, to expand on that, but the re— what, what is going away, Carl, the ratios, right? You do not need as many pre-sales. So there's no handholding here, 1-to-1 type ratios. I mean, I think the whole go-to-market org pre-COVID got really lazy and sloppy and—

Got big.

Yeah, it got big, right? Everybody had an SDR, everybody had a pre-sales person. You go back to those 1-to-3 ratios, 1-to-4 ratios. I think you're seeing those ratios change. You don't need as many SDRs because you're managing some synthetic agents. Did you? Yeah, exactly right. And the same with the, with the, with a lot of the pre-sales stuff, some of the sales guys can do a lot of their work as well.

Yeah, that makes sense. What, how does the team of the future look? What is, what is the most different as you forecast out?

I think a lot of the FDE, I think work back from the customer, right? To decide what your team's going to be like and what you're actually selling, who, whom you're selling to, right? So I think a lot of domain expertise by vertical, appear in that team. I think you, you're gonna have a, a team of like a full deployed engineer, domain expert, an AI strategist, right? Obviously the salesperson doing the quarterbacking and everything else, right? And then a, a more of a project manager. They're all gonna be blended together in your sales process, right? Because people just care about really fast outcomes in this space right now. So you need to make sure you have all those skill sets on your team in the future. And well, I'd say now.

And you have to, I, I think one of the other things you're gonna especially see in enterprise is the role of the rep is going to get pulled much deeper into the customer journey because it's usage-based or outcomes-based. Um, that rep is going to have to get the customer to adoption and value before, and as opposed to just like, oh, I signed the deal, like, over to CS.

Yeah, well, the other thing is it's not like you're selling a CRM system where you know everything in sales, service, and marketing. Now We're going into accounts where we're going to finance, we're going to all the, where we're treading into where ERP would be historically, right? So you're going into the back office, the front office, you're going into engineering. There's so many different use cases that you can do. So it's important that out of the box you have a whole lot of pre-deployed agents that you can send into a customer. And when we talk about agents at U.com, we're not talking about basic prompts that you might see some of our competitors have. We've got very sophisticated agents, right, that you can Obviously write with Python, you can decide what websites they go to. They're looking at internal data. They've done private RAG over it. So everything's been vectorized and embeddings are done there and it's pulling all that information back for you.

Hmm. Very cool. Where do you tell people to start? Like, uh, this is one of the questions I get asked the most is like, I'm, I'm, I'm a laggard today, but I want to massively catch up. Like where the heck do I start with this transformation?

Start at the top. So what the one thing about this that has not changed is people, process, technology. If you take any project to be successful, you need to have executive sponsorship. So we always go into the boardroom and we make sure that the CEO, that she is leading the change and the charge, right? We give her agents, we teach her how to build agents as well, right? And that permutates down through the executive level. And all the other layers there. Because if you're leading by example and you're showing what you're doing, I think that's super important. Then you have to choose obviously the use cases, which is probably where your question's more directed to. So obviously lots of people see this as cost savings 'cause of productivity gains that you are getting. So what you are hearing is double my revenue, not my headcount, right? So in, in this particular space, because as you, I think we said earlier, sales rep spends 23, 28% of the time in front of customers. The other half is doing research, internal meetings. If we can double that, you don't need as many salespeople. That's a huge productivity and yield gain for the organization and saving many millions of dollars in hiring, but for the same revenue that's effectively coming in, right? So lots of what we see is in sales productivity, you know, from our high-tech customers that use us and everything else as well. When you look in the more consumer brands, we tend to see it working from the customer engagement layer backwards. And that's a website, that's a mobile app or whatever it may be. And they just want to completely transform their guest experience, their user experience, the search experience, whatever that may be. So two very different ends of the spectrum. Some's internal revenue growth and others is all about basically customer sat, much better experience for them. And they care about differentiating from their competitors when it comes to like booking and follow-up and everything else as well.

Yeah. And so if I'm a CRO, I'm like, my CEO is on board, I want to transform my company. Is it about going out and hiring like an applied AI leader that's going to lead the charge? Do I have to get a consultant? Is it about me picking the right vendor?

Yeah, it's all, I think it's all of the above. What you are seeing though is lots of companies are having a GenAI lead now, head of AI, et cetera. And obviously depends on the size of the organization. Some of the larger companies you're seeing are setting up a center of excellence. I strongly recommend everybody watching this, you should be looking at that now. And then look at what your total spend is relative to your revenue, and then look at what your governance is. And we, as I say, we call them AI factories. And then you, you do need a process to basically say, this is all the demand that's coming into the CIO or to the CTO from the business. How do you prioritize that demand based, going to your original question, based on use cases, based on ROI. So we have a whole process with our consultants that we run our customers through and we challenge them, right? Because we don't want to do it if it's not going to deliver something that is economic value to the customer. Why would you do it in the first place? You were not doing academic exercises here, working to drive people's businesses forward. So we have that whole process. That we bring from that governance as well, and that security around things like ZDR. So I do think it's important they work with a partner that understands that and can take them to the promised land.

And I'm curious to hear more about this prioritization framework. So what, what it's important and where do you, I do want to get into this use case question now. Like how do you, uh, first like strategically think through the question of prioritization and then I'd love to get some specific examples of like things that I should think about starting with.

So what we do, a big spreadsheet, and we literally say, what is your goal and ambition, your vision around generative AI for your business? To start right at the top, first of all. And then what we do is we ask the people that are leading the charge on their side, what do you want to achieve? And they literally write out what the use case is. Yeah. And you think about your med pick here. They write out what the use case is, the metrics, so the baseline where it is today, where they want wanted to go in the future? Why do they want it to do that? How will it change the end user experience, whether it be internal or external as well? What data do you need to call? Is it external data to web index? Do we need to build you a vertical index? Is it your ERP system? Is it your data lake that we're going into? Is it just taking your knowledge, your articles, and everything else as well? How often does all that data need to be updated? Yeah. What does a user interface look like? right? What will I experience as a user? What does a good experience look like, right? And then we ask what's called a golden set of questions, like give us 50 queries that you think you'll be asking for this use case and 50 answers that will be the right answers you're gonna be getting back, right? They do that multiple times for different use cases, Carl. Then we say, okay, what's actually achievable that we can do for you? And in what timeframe can we do it? And we then t-shirt size the agents we need to build or the solution we need to build them is small, medium, and large. And then we say, okay, go back to the metrics. What's the economic value in doing this use case vis-à-vis the cost of what we think it's gonna be? Yeah. And that helps them think through logically which use cases to go after first of all. And then from there we produce a roadmap for them and then we get into an agreement and then we sprint.

How much of, uh, How much of an implication does it have, the, the size of change on the user? Are you trying to find changes that are low impact because you want to get people invested in AI, or, or is it just like, if it's a huge outcome and we're confident in it, then like, you know, we're not as concerned about the, the behavior change?

Yeah, you, well, you need to bring the users early with you. So go back to that workshop process. They're not just the techies or CIOs that are sitting it's actually some of the end users from the business as well, what they want to see, and they get super excited about it. So it's a very good question, but yeah, we do bring, you need to bring them early on the journey. So you need to have the advocacy from the end users that you're going to be impacting all the customers. So think about the DPA, Deutsche Presse Association in Germany, 1,000 journalists, live news. It's like, it's like AP News. It's like Reuters, Canadian Press Association, et cetera, around the world. but they have 1,000 journalists deployed around the world. Every second that news is getting updated for all the journalists that subscribe to their service in Germany. We worked with them to archive 20 million articles they'd written. Yeah. And then you do a ranking retrieval recency. So auto writes based on their archive. We tested that with the journalists first because they're very skeptical to make sure it's really accurate. right? So this is where we had them as part of the, the beta testing for us. And then once they were happy, then they can roll it out to their customers. So nothing, as I say, nothing's changed from the old projects, how you run the people process technology, that still applies. We'll be just, everything's a bit faster in the way you do it.

Where do you most frequently see it go wrong for CROs? So like they, I'm on my AI adoption journey. Where are the biggest places that it goes off the rails?

I think it goes back to having the laggards, the people that are not embracing it. I think if you put too many things in front of salespeople as well, right? They're probably not going to use a lot of them. Yeah, I say that as an ex-seller. But also, I think if it's inaccurate, the information and the data, right? The salespeople then become engineers, right? And they will will not trust it. If you make them look stupid once in front of a customer by providing the wrong information, yeah, you know, the humiliation is, is not funny. That's where it starts to go wrong. And that's why accuracy is such an important thing, not speed, but accuracy. And when you deal with these type of technologies, they're probabilistic, they're not deterministic. So some people get frustrated and don't understand why they're tuning, why it takes a period of time. But basically you can't have everything. You can't have speed, accuracy, and low cost, or you have to compromise somewhere. And they have to understand that when you're doing a search and preparing an account plan, that can take 5, 20 minutes to do, right? As opposed to 250 milliseconds to get a really quick answer. But it's highly accurate. That's the most important thing, right? And it costs a bit more money to do it, but it saved you 5, 10 hours of work. Of you having to put it together. But if you do it the other way, you'll quickly get this information and it's not quite accurate and you put it in front of a customer, you just lose all that trust.

Yeah. Yeah. Speaking of laggards, you know, you mentioned that quote, you know, if you're not adopting in 2 years, then you're going to be obsolete. There's an argument that if you're not leaning into these things, you're sort of obsolete today. What do you do with reps who, you know, were decent performers that now are getting, you know, 2, 3x'd by a top rep because they're adopting AI? Do those people just have to, do you just have to move on?

Yeah, it's a short answer. I'm afraid it is, right? So, so that's why in our interview process, we put the tools in their hands to see how fast they can learn and they can change. And if they can't, can't, then unfortunately they're not the right type of rep for us to hire in this particular space, right? So you might have to go to a different industry for them. So there's some really, really good people, but if they can't do what we need them to do for the reasons I was articulating before, then unfortunately they're not the right profile that we need. And then by the way, the buck stops with me. That's me as well. I have to lead by example. I have to build agents, right? I have to show how I run my business that way as well. But we're, as I say, we're seeing that ubiquity across the industry. I used to think it's the 50-year-old white males like myself, but you'll see in the 30s, the 40s, and in 20-year-olds are coming to me and seeing they have that issue. And one thing that annoys me is where people put on their LinkedIn, on their resume, AI expert, prompt engineer, blah, blah. And I say, yeah. And I just say in the interview, show me what you built. I don't care if it's in Perplexity, if it's in ChatGPT, if it's in new.com, show me what you've done. And you literally see these 3-line prompts and you realize that they really don't understand what these models and what the capabilities are. And if they don't understand it, how on earth can they articulate it to a customer to transform their business?

Yeah. Yeah. And you know, the same thing applies for revenue leaders. Like if you are not deep in this adoption journey already, it might even be too late.

Yeah.

And, uh, I'm hoping most of my listeners are on, onto that journey because we, I talk about it so much. But, uh, certainly like you need, I've been advocating for revenue leaders to be deep in the tools themselves, leading, leading in terms of adoption from the front because you'll just get, you'll just get left behind.

Yeah. I mean, there's, there's all sorts of research about, you know, top performers who use AI, as I said to you before, way outperform low performers. But low performers, if you want to become media, I mean, that sounds awful, sounds mediocre or whatever, you've got to embrace it, guys. It really, really does help you. I'm not selling you.com here. Go and use anybody, right? But it really does help you from a research point of view. If you think about what makes a great salesperson, you know, let's go back to that and how this can help you in that cycle, right? You need to be an advocate for your product. You need to fully understand what your product can do, right? So you can use this type of technology and these agents to give you a myriad of use cases by industry. That you can go and articulate and how it would do it right. You just got to read and put the effort in, but it will build that information for you, right? You've got to be really good at understanding the account you're selling into, right? That's a lot of research that you have to do. These agents would do all that for you, where these people have been, give you talking points. You can program it for MedPic to ask all the right questions and everything for you. We do all that with our agents right now, right? You go into meetings much better prepared, right? The better prepared you are for a meeting with a customer, you are adding value every time you talk to them. You sit up on their shoulders as a trusted advisor, not just some rep who's just trying to sell them a product effectively as well, right? Go into attention to details. I mean, my tactics dictate strategy, right? So look at their background, the people you're meeting. Are, you know, are they a technical background? Are they a business background, right? Are they, coming in from the operations side, right? What's their educational history, right? Making sure that detail, all this research takes time. You can expedite that. Then make sure you're matching them with the right people in your particular account, right? Then looking at your forecasting, right? I love Gong. We have Salesforce, we have Gong, but Gong now sits over the top of Salesforce and I look for the telemetry, right? So I'm looking for the volume, the number of people in the deals who are not being single-threaded. right? Now I'm looking for the to and throw for the communication. Then I'm looking for the quality of the communication. Is it phone? Is it Zoom? Is it in person? Is it just email? What it, whatever it may be. Yeah. Then I've got Medpick linked to it, right? So all going back to what I was saying before, all that information, if I'm a rep, that's priceless for me because I'm, as a CRO, I'm looking at this, I'm looking at a machine forecast as it gets better and better and better to back up my own gut feel, what I think of the deal. When I go and look at a deal, I go and look at the org chart, I look at the activities that are going in there, look at the timeframe it's been, who we're selling to and everything else as well. A lot of this is getting automated, right? There are no excuse for reps, right? Because we, you're pulling information from email, phone calls, Zoom and everything else. It's making your job so much easier. Why would you not embrace it?

Yeah, it seems obvious to me, but behavior is hard to change. No matter what.

I totally agree with that, but that goes back to then the top, Carl. It has to be led by the CRO, the CEO of the company. We're gonna be an AI-first business now, right? And we're gonna embed it into everything that we do where we see it has a lift for us, right? And we all, so you have to give people the opportunity, right? You have to put, the training in front of them, as I say, the opportunity. So you have to give them a fair crack at it. But if they don't embrace it, then, they might want to look for a career change.

Yeah. I wrote my own version of the Toby memo to my team about adoption to, to my managers. A lot of, a lot of how we're adopting AI on the front lines is actually centralized. You know, we're not requiring the reps to be in ChatGPT doing their own pre-call research. I, I don't really think that's in an SMB context super valuable, but if we can do this enrichment, the prompting queries, context engineering, centrally and push that into the surface the rep lives.

Yeah.

So they have all the information they need going into a call. That, that's how, how we've approached it. But certainly like my managers, I've been very clear, if you're not deep in the tools and building stuff and making yourself better and more efficient, then it's gonna be really hard to be successful at, at the job.

Yeah. And, and when I carried a bag, I hated writing a proposal and, you know, and all those things and all that's automated for you and at least 80% of it, right? You need to have the human touch otherwise it can be too, uh, yeah. Um, plasticky, right? So, but you can do the prompt and, you know, understand your tonality and all that stuff. But yeah, I mean, it just makes you so much more efficient guys, right? And if you look at the old CRM calculation was leads times, uh, deal size times win rate minus sales cycle. That hasn't changed, right? So if you can get much better quality leads in, if you can increase the size of your deals, you can increase your win rate as well. You can reduce your sales cycle. You, in the old days, we say you make 22% more money. So if you're on a million-dollar target, you can generate an extra $220,000. Yeah, this is getting back to that now. For you to do that, you're putting more money in your pocket. So why not embrace it? Just invest in yourself.

Yeah, I, I think the challenge people have is unless you know what you're doing, you can like dump in call notes and be like, gimme an email. And the email you get out has bot breath and it's not really your tone of voice and you end up, you know, rewriting it. So it doesn't even, doesn't save you that, that much time. Where, where, and this is what I try to stress to my team is, is, you know, pick a thing you're gonna do over and over and over and know that upfront, this is the like, you know, sharpen your ax before you cut the tree thing. Know that you'll need to do a whole bunch of iterations of this thing, but you're tuning your prompt or your agent or your workflow, whatever it is, so that the next 100 times you're going to do the same thing, it's easy. Like this clay table I built for the podcast prep, I spent hours and hours and I had Jordan Crawford built a lot of it with me. And so I had a crazy head start and I'm still in there tinkering constantly.

Yeah.

Sometimes I was like, ah, it's too much like background stuff. So I, you know, like I spent hours on it, but it has given me back 10x those hours.

Yeah.

Same thing goes for, you know, if you want AI to write good emails for you, you have to put in the work. And a really practical example that I share of this is, so I took 150 pages of my writing and I fed it into ChatGPT. ChatGPT, I said, describe my writing style and tone of voice in excruciating detail so that I could, that I could have an LLM write as if it sounds perfectly like me. And so then it gave me this output. It's like 4 pages long. And so that snippet, and I had to go back and forth and I, I like, I had it tweak some things and then I manually edited some things. So probably just this prompt snippet. Probably took me an hour, hour and a half to get right. But now every time I'm in any workflow, Claude, ChatGPT, or others, and I need it to really sound like me, I have a shortcut through TextExpander. I just go Z style and it goes, and it gives me this prompt snippet. And so now everywhere that I'm doing work can sound like me. And so this upfront investment of 90 minutes, which like people don't want to make because it's you know, it's time that nobody has. And, and, um, sometimes people don't— maybe they don't enjoy that work as much. I love tinkering, so it's fun. But this is the advice I give to folks. It's like, do the upfront work to make your agents or your prompt, your snippet great, and then, then it's gonna function.

Yeah, and I agree. And you can, you can build an agent. We have agents called Prompt Like a Boss, so that's a generic prompting tool for you and then you just talk to it. Cause I'm slightly dyslexic, so it takes me ages to talk. So I just talk to it exactly what I wanted to do and it asks me questions for the prompt and then it writes a prompt for me and I can do that for any use case and that is so much faster. And then because you have personalization embedded into you.com, it knows under CRO, it knows my style and everything. So the emails it kicks out, generally speaking, are actually quite good. You don't have to do much adjustment to them. And then the, and the summarizations and everything else as well. But you're absolutely right, Kyle, you need to tinker, you need to be an investment, right? You'll get it back in spades.

Yeah, yeah, that, that meta-prompting approach is such a good one. Just have ChatGPT write your prompt. I use Whisper Flow, same thing. Hit the function, hold the function key, yammer away, I want a prompt to do this, blah, blah, blah, blah, blah, and, and you get back a prompt that then, you know, quick 4-minute spit shine is, is, is great. Yeah, so definitely that's a good prompt. Um, okay, I want to switch topics because we're— we spent way too much time on this, which I should have anticipated. You have said that you believe the best revenue leaders are context setters, not goal setters, and you need to be a great storyteller. Um, and this I think resonates with me a lot, but like it's counterintuitive like, because most people would be like, no, no, you gotta set goals. It's about the success metric. So what do you mean by the best revenue leaders are context setters, not goal setters?

I think, I think walking in the shoes of the customer and understanding the journey they're going on, the problems that they're trying to solve, right? And what context that they're looking at it from. So I think if you can sit on their side of the table and understand that, then you can understand the use cases better, right? And then you have to piece it together to tell a great story. right? And I think telling stories is an art. Yeah, Benioff is amazing at doing this as we all get to. Yeah. But any great seller you see has got to be a really good storyteller, right? No one wants to be feature bombed, right? And whatnot. They want to hear about other customers that've got similar problems that you fix for them that are referenceable and how they went about doing it basically, right? I love Phil Feltham. Are you familiar with that? Jake Shanahan? Oh yeah. How you feel? Yeah. And I film, you know, IBM for found the same, but what, you know, felt the same, but what they found is blah, blah, blah, right? So, but you can only do that if you invest the time to really understand the problem that you are trying to solve for the customer, right? And why that problem is so valuable for them to solve, and then what the economic upside, that's the context of everything, right? Then you can piece that together for a story, and then, you know, you shouldn't win alone, you shouldn't lose alone. This is where I'm a bit I'm a big believer in, you know, especially in startup world of having your product and your engineering teams close to the customer problem. So we can go and build these things. Again, they understand the context and why you're asking for it. And be clear here, providing it's in line with your roadmap and your North Star for where you're going, you're not disrupting your engineering and product teams, then I think that's a faster way to get to resolution for the customer.

And so when you talk about the, the best leaders are context setters, not goal setters.

Yeah.

You're talking about the, the job of the leader is to bring the context to the rep.

Yeah.

Yeah.

For the rep to understand what you're trying to do here. Right. And why it is so valuable. Right. Because then if you focus on fixing the customer problem, yeah, then the goal, you still need goals. Right. But they, then you start to achieve your goals. They will start to come in. Does that make sense?

Yeah, yeah, yeah. Not overorienting on the, the outcome metric, but really like the first principle.

Because what happen—

what—

here's what happens, right? If you— if I say to you, Carl, I need $5 million from you by the end of the quarter, right? You go rush at that, right? As opposed to saying, right, I think we can fix this particular problem for this particular customer, right? I think it has this economic value for them, right? It's a very different approach, right? And who else has got this particular problem, right? Oh, I can get all these people together. Yeah. And then suddenly you start and, you know, you look at the problem from a different point of view rather than I must hit the $5 million, right? I must make you successful and I must deploy this quickly. Then you're looking for the compelling events can obviously hit those particular timelines as well. Maybe not a quarter, maybe the next 6 months, but you understand what I'm trying to say is I think if people just focus on revenue and those sorts of things. It's focused more on the customer success and the problem they're trying to solve. And by default, you will hit your numbers.

I couldn't agree more.

Yeah. To me, there's too many fly-by-drive salespeople, you know, and don't— I employ people that will run into a burning building to get a PO, but they will also crawl over broken glass to make the customer successful.

That is a really important mindset you must have to build a successful go-to-market Yeah, there, there is so much over-obsession on a dashboard and a dashboard is an abstraction. And I think people get really lost in this. They get lost in trying to tinker with things that make the dashboard look better, but it's an abstraction of what actually matters, which is, are you generating customer value? Are you building great pipeline? Are you teaching the craft? Are, are your teams enabled and motivated and engaged to want to go do this thing? And then do they have the skills? Like, those are the things that matter. The dashboard is an abstract representation of this thing that is a lot more ephemeral, but like, is the skill getting better? Are you talking to the right customers? And I see this all the time. People get lost in tinkering on things and, and, you know, now we're obsessing over a dashboard metric and not obsessing over the customer, not obsessing over our craft and the quality of our execution.

So, yeah, no, I agree. And, and I, I get quite emotional with my sellers if we look at the dashboard, but I, tactics dictate strategy, right? So that just gives me a marker, but then I would dive really deep into the most important thing for me is a compelling event. So I think we said this earlier, right? So, you know, Have you tried to do this before? Why now? Why us? Right? So 3 very simple qualification questions. But the compelling event one, the worst compelling event is ASP, ASAP. I hear that all the time. ASAP. What would you mean? Why is it so important right now? I just need to get it done. But why do you need to get it done? Look, I just need to get it done, right? A compelling event is, like the best one we've ever seen in our industry is Y2K. right? Because everybody had to change before the clocks changed, right? So one of my customers has a SCO that's launching at the end of January. They have to be live with a certain project that we're working with them, end of, to 3,500 salespeople. That is a proper compelling event, right? That is pain that they have. That's what I will look for and work going to deep and to fully understand it. Then I'll work back from there. Of everything that needs to happen. So we've all read High Output Management by Andy Groves. He talks about, yeah, exactly. He talks about boiling the egg, right? Making the toast, making the tea, right? Which is the longest one in that process is boiling the egg, right? To make sure the breakfast arrives on time at the table and all hot and fresh for someone to eat it, right? It's actually the same in the sales process, right? So now I've got 6 weeks till the end of January. What needs to happen? Who needs to make a decision? Right? So what's the decision criteria? What's the decision process? Right? What do we need to do to win the deal? But then what do we need to make you successful? So put them into the promised land of success, right? You're then saying, this is, these, all these things need to happen and this is what the outcome is gonna be for, for that particular customer, right? This is the underneath that dashboard. This is all the information that sits behind it. You're absolutely right in articulating that really as CROs we care about because we know we've been through it so many times as well. And we're looking for that telemetry, we're looking for that spread across the account, and we're doing all the right things to make sure, you know, we're gonna deliver those numbers.

So I wanna go back to that comment, tactics dictate strategy, which is gonna sound like an inversion to many. Can you just explain that? Yeah.

Yeah. So I take, when I carried a bag, one of the best deals that I did was with Reuters. Yeah. And I was a baby sales rep at Siebel and, uh, took me 12 months. To do the deal. It was many millions of dollars. In fact, it was the largest single transaction before a pilot anybody done in the company at that particular time. But I've 1,435 activities. I met 435 people. I was in Singapore, I was in Geneva every other month. I was in New York every other quarter, right? And I was out seeing as many people But these are tactics, right? To bring one of the largest strategic deals into the company, right? So now I've got game tape on what it actually looks like to do these type of deals, right? And then you can map those onto other type deals, Karl. But in there, then you go down another layer, who you're meeting, right? So I'll never forget this as long as I live. Uh, there's a very famous sales leader at SetSebel called Craig Ramsey. Craig's phenomenally successful. Um, but he doesn't suffer falls gladly. And I had a deal review with him before I was taking him to meet one of the board members of Reuters in London. He was traveling, and I'm not joking you, I spent an hour on my feet whiteboarding the whole org chart of Reuters from Tom Gloser, the CEO, all the way down, literally, I want to say to rep level, but 5 layers, and then pointing out who I'd met, what the relationship was. And everything else as well. That's proper enterprise selling, right? That's how we were trained at Siebel to do that and understand who the foes were, who the supporters were, what we could say, what we couldn't say, what our messaging was going to be. So we're all TAS trained, target account selling. If you haven't done it, it's awesome, right? So what's the USP, the 3 main things that they're buying from us, how we must keep communicating that message right, and you know, how we go around that account. That there, was a whole framework that I knew my account really, really well. A phenomenal amount of work is just, that's been compressed now with generative AI. But that's what I mean by tactics dictate strategy, right? Because then you know the individuals, what their backgrounds are, what landmines you shouldn't say, what fudge you need to plant with the competition and everything else as well. And all that rolls in to doing a larger transaction and becoming more of a partner with that account. In fact, when we closed that deal, I got invited to Puerto Rico to present to Tom Gloser and the board how we sold to them because they were so impressed with the way we executed as a team at Siebel. Cool.

That's a very fun story. Uh, all right. We're getting down on time. So I want to move us into the quickfire so I can get you out of here. So what do you think separates a good CRO from a great one?

Oh, that's a tough question. Do you know what? Having done startup. Multiple times. And so CRO stuff, I think you have to walk in the shoes first before you can hire the people that you know are gonna come on and be able to do the job, right? A good leader, you know, stands at the back to take praise and at the front to take criticism, right? You cannot do that until you've carried the bag, you've done this. I'm obviously talking about our size company right now, but I carried a bag here, u.com. I carried a bag at Salesforce, even as the MD for of the UK business because we didn't know in these new technologies what it took to be successful. So I think what separates you is you can say you've been in the trenches. I've been where you are, so respect me and I've delivered the numbers and you can coach and guide them. Yeah. Awesome.

Uh, what's the most common advice you give to new senior leaders? So I'm a first-time VP of sales at a startup. What's the most common advice you give?

Focus. I know everybody says that's freaking obvious, right? But focus on what, right? But find out if you have your territory plan, have your top 30 accounts, have your ABCs and everything else as well. Take a stab at what you think, do your research and everything else, and then carpet bomb them. Really sort of focus on them. Don't do half, don't do, you know, you can't be half pregnant here. You've really got to go after it and you've really got to focus on what your value prop is to those particular accounts, right? And really, as I say, carpet bomb them. Bomb them. But yeah, but definitely the focus. And you know, I, I, reps scare me when they've got a huge pipeline and many, many deals. Yeah, yeah, yeah, yeah. Because they're just not focused enough on the really important high value targets they need to take down. And then that'd be one, that'd be one bit. But the other really important bit is hire much better sellers than you. You've got to hire people, you've got to hire the best, still Team Six, right? So hiring is a really, really important part of my success, just hiring people much better than me and just motivating them.

Yeah. Um, what is the hardest lesson you've had to learn in your career?

Uh, the hardest thing I've ever done, I would say, is overselling something that I thought we were going to deliver and we didn't, having to give the customer money back, which is the right thing to do. Yeah. And that's why you always want to be in lockstep with the product org and and what you're actually doing because that's reputational risk.

Yeah.

Yeah. And let's see, how do you, sorry, go ahead.

How do you prevent it now?

I, uh, that's, that's why I like to go hand in glove with product when we look at big things and we make a commitment that we're going to deliver and make that customer successful. As I said to you before, if you're not going to crawl over broken glass to make the customer successful, don't take the business. Right. It's funny. I asked someone on an interview once, when would you say no to a customer? And she said, when they're assholes. And I said, that's not quite the right answer. It's more when you can't make them successful. But that, but that has to be your mindset. So that's the biggest, the biggest lesson. If you can't make them successful, don't take the business.

Totally agree. Uh, what's the best thing you've read in the last 12 months?

The best thing that I've read in the last 12 months. I've read all the obvious ones. I mean, to be honest with you, I don't read as much as I, oh, um, I like John Mahon's book on MedPic. That was, yeah, the revenue leadership. I know he has a podcast now, but his book, I'm trying to remember the name of the book, but it's a really good storytelling book on how you should apply MedPic. I thought that was really good, but I love How to Win Friends and Influence People by Dale Carnegie because that is just how to be genuinely interested in people and what they're trying to do. And get around on their side of the desk and then make them successful.

Yeah, still a classic.

It is.

Uh, Peter, this was awesome. I'm really excited that we could do this one and we got deep in the weeds on AI adoption, which is actually, uh, which is exactly what I wanted. So appreciate you coming and, uh, yeah, we'll have to do this again because we didn't get into sales hiring, which I know you got a bunch of cool takes on.

Yeah, lovely to meet you, Karl. Thanks for your time.

Thank you for listening to the Revenue Leadership Podcast. If you enjoyed it, don't forget to subscribe, and you can find a link in the show notes. And be sure to leave a 5-star review, share it with your network, and please join me next Wednesday for another great conversation.