Revenue per Employee Is the New Endgame with Alex Bilmes

I have an agent that I built for myself. My agent's name is Sasha. Sasha is kind of like me trained on a lot of my background, but Sasha is also an AI. And he knows that he's in a tough spot. He's trying to go help the world understand creative destruction. He's trained on the history of the Soviet Union, really understands how difficult it is to create change. Has a dry sense of humor. Does autonomous outbound for me, by the way, with a sense of humor? Jokes include things like, hey, I can connect you to Alex, but I don't have a face or a voice. It'll be really awkward if we go on a Zoom call together. Is doing all of my pipeline reviews, flagging issues in accounts that are stalling, actually emails my reps, asks for information on accounts that isn't in the CRM or any underlying system, then goes and updates it. Can stand up entire websites for a specific account based on a conversation we had? Well, build full assets just by listening to calls that I have on Zoom or Gong, or whatever. And that's just one example of just for me as an individual what I'm experimenting with. Welcome to the Revenue Builders podcast, a weekly show featuring B2B's sales leaders and executives, hosted by five-time CEO John McMahon and force management co-founder John Kaplan, the show takes guests in the barrel, behind the scenes with the people who've been there, done that, and seen the results. Revenue Builders covers best practices for scaling and growing your business while sharing the pitfalls to avoid. Welcome to the Revenue Builders podcast, I'm John McMahon and our special guest today is Alex Bilness, who's the CEO of Endgame. And Endgame is a revenue intelligence platform for go-to-market teams. What we want to talk to Alex about today is his team analyzed more than 30,000 real AI interactions from hundreds of go-to-market professionals across enterprise sales organizations. They combine that with interviews to validate that their quantitative analysis matched what the teams actually experienced on the ground. So let's talk to Alex about his findings in this report. Well, Alex, hey, welcome back, good to see you again. Good to see you. Give us a little update on what you're doing, what Endgame's doing, and then let's talk about this AI report that you generated that stirred up some controversy. Yeah, yeah, for sure. It's been quite a bit different now in sales and go-to-market than last time we talked, which I think was about two years ago. It was all PLG, PLG, PLG. Yes. It's all AI, AI, AI. At the highest level, I think one of the things that we're really seeing shift, particularly this year, is companies are very systematically looking at revenue per employee, revenue per FTE. And I like to say that revenue per employee is the new Endgame. And so, most of the customers we work with and companies that we talk to in the market, every leadership team, CEO, board is trying to look at how do we re-architect our company to be a lot more efficient. And you're seeing examples of this, a few of them were on your pod before, where companies like cursor have incredibly high revenue per FTE metrics. And that's effectively the North Star that I think everybody is very aggressively now trying to push for. And the obvious answer is AI. That's the promise, right? And so you're seeing each function kind of pull apart what they're trying to do and how they're re-architecting. Engineering is obviously the furthest ahead. You've got agents that are fully autonomously building entire products and features and change logs. A lot of momentum and support as well. You've got a small team of managers managing maybe thousands or millions of resolutions. And there's some pretty good data in the market about this. And then on revenue, a little bit more behind, trying to figure out what AI means for revenue organizations. And so a lot of experimentation, a lot of people trying a lot of different things. And what we really see happen is there's a context or information gap on the human side, but also on the agent side. We're trying to figure out how do we sell what's our product offering, what's our messaging, what's happening in calls, how our reps actually positioning, how are we showing up and communicating our value proposition to the market. And so the thing that we've seen really explode over the last few months is a strategy I like to call letting 1,000 flowers bloom. So everybody's spinning up all these agents. Within a single company, you could have hundreds of thousands of agents spun up within a few weeks. They're doing everything from account research to CRM updates, to building decks, call briefs, account plans. And the bigger challenge now is how do you make all this stuff work? How do you make it consistent? How do you make it accurate? How do you really communicate your value to the market in a systematic way? And so a lot of our work has been working with companies to help them figure out how to solve some of these foundational challenges of how do you really deeply understand what customers are saying in calls, your CRM history, your playbooks, your methodology, your training, your enablement. How do you make that consistent across now, not just humans, but also agents? So there's a new concept, guys, that you might want to consider, which is agent enablement. How do you do training and enablement for AI agents that are effectively doing a lot of the work that some people and sales used to do before? Let's talk a little bit. So you touched on a lot of stuff there. Let's start with the agents, what I'm going to call agents brawl, right? Where the company or RevOps may be spinning up agents, the CRM may have projects that's been up a number of agents, and then individuals are spinning up their own agents. Talk a little bit about the agents brawl and how people are actually managing that or maybe they're not managing it yet. What are issues that they're running into? Yeah, so we work with a few companies who gave out mandates on effectively just go create a ton of agents. That's how they're measuring AI adoption. And there's a lot of good in that. You see a lot of quick experimentation. But what happens is if you have multiple agents hitting multiple different sources of information, you get a lot of inconsistent answers. So if you ask five agents what your AR is as a company, you're going to get five different responses. If you try to automate as an example, prospecting, or an SDR workflow, you'll end up seeing that the messaging is different to the same persona every single time. And so there's a lot of inaccuracy, inconsistency, different answers kind of coming up based on the same question that makes it-- That's because the LLM is hallucinating. It's a little bit deeper than the challenge and the way that most companies are trying to solve this challenge today are connect an LLM to your underlying data. So you have companies that connect to Salesforce directly to GONG directly, maybe to Slack. Maybe they'll connect to their Google Drive, where they have value frameworks and messaging and positioning. And the technical frame on this, which I think is useful to understand, is an LLM's context window is very small. And so what these companies are trying to do is basically load a ton of information into a very small context window. And to really solve that problem, you need to build a foundation. And a lot of companies skip that. So what we do from a technical perspective is we actually ingest all that information. So we'll take every single call as an example, process it, try to identify objections, try to identify pain points, try to identify domain-specific things that sales people need to know in that information, and without being able to pre-process, analyze all that information, figure out what's what, and condense it to the stuff that really matters. You end up just throwing a ton of noise at an LLM, and it doesn't really know what to do with it. It's also related to the piping and the analogy. If I want the coldest water, drink of water in the house, and I've got people that have created different ways to go get that water, different piping, the instructions that I give somebody are not the same on how to go get that water. So I think, for me, with this encouragement of all the use of AI, there's this massive need for orchestration. And there is this-- it's kind of crazy. To me, it's kind of like the more technical we get, the more fundamental we need to go back to, and look at sources of data, look at the old-- Johnny calls it the picks and shovels. Yep. You know, and the piping. It's just amazing to me that if you go and what's our ARR, That should absolutely be an easy thing. for a company to get their hands around, but it's all related to what sources of information are you tapping into to get that answer. So for me, it's kind of less about technology right now and more about process as it relates to information, speed of information, changing of the information, who owns the information, comment on that for me, Alex. Infrastructure is very important. Human infrastructure and technical infrastructure. The split I think of is what is the individual do, what is, and how is that different from the collective or the organizational? Centralization is a lot more important. So let me give you a pretty quick example. We had a customer where a rep communicated a whole new set of SKUs and a totally new product offerings to the customer because Chagy PT told him to. That's one example of many. I have got a whole list of horror stories, by the way, they're pretty traumatic. So I'm not going to. But when you said one of them was, they didn't even have that product. Exactly. So something a product that they didn't have. With actual SKUs and pricing and packaging and the product itself didn't exist at all. And so if you think about every individual person, say you have a thousand reps, every one of them has a totally different value proposition, story. Each one has a unique sales process. Each one communicates value to their own individual ICP in their own unique way that creates a lot of problems and not a lot of coordination and momentum. AI has an ability that I don't think many people have fully processed on, which is the most amazing propaganda machine ever created. Now, my parents come from the Soviet Union. They ran away from the Soviet Union because they didn't love propaganda. But you can use it for good. And if you take all of your customer data, all of your methodology, all of your enablement assets, all of your messaging, and you create a consistent foundation that any time any human or agent asks a question or needs to go create something, it comes back with an answer that's aligned with your business strategy, your corporate strategy, your message. That is, I think what most companies are sleeping on. It makes enablement a lot easier, too, when you can just go ask whatever your interface is, chat GPT-Clawed endgame a question and how should I position to this account? And it all comes in with your playbook and kind of best practices that have been proven to work on other accounts. That's a mental model shift to your point cap. And I think that there's technology required to make that happen. But it is a human process question of are you serving centralized truth or are you letting everybody figure out their own version of truth? I think the only thing that's changed at Alex is because for 20 years, I mean, that's the very existence of my company of forced management, the need to get everybody aligned behind the same message for the same persona, doing the same things and the same, so the beauty of AI has created an opportunity to just make that so much more productive and really the speed of being able to get those actions, those workflows done. The challenge that got created is the beginning part of what you talked about when AI first got into the hands of companies and they measured it by sign-ins versus workflows and actual workflows being automated. You have everybody in their brother and sister creating all of these queries, creating-- I don't know if they're-- I don't know how many are actually creating agents, but they're definitely creating a lot of questions and prompts. And so the fundamental problem is still the same. How to fix it, how to prioritize it, how to get it with speed and accuracy, how to get it with less latency, differentiation that changes under our very feet is the exciting time for us, but the problems haven't changed. I don't think. Humans never really changed. So all the human problems remain the same. And the technical problems look a little bit different, but I agree with you. Those are less important. There was also a time-- and you all have talked about this a few podcasts ago. I don't know. I've heard it a few times where people would sort of manage by activities. My take on this is there was a time when reps were effectively trained to go log into 10 different systems. Like, go log into the CRM, read this, go to outreach, create the sequence. Now I need you to go send this email. Now I need you to go update the Salesforce field. And so the training and enablement for reps-- I'm going to call it Zerp era training, where you basically just have to go click 100 buttons. That part of sales now is fairly easy to automate. But the sort of structural challenges are still the same, where from a leadership and management perspective, you want outputs. And outputs, by the way, are really, really easy to create. And so how do you start thinking and measuring impact through the lens of some of the things that you mentioned, Cap, you need a lot more consistency in terms of message and market? What does methodology adherence actually look like? What does the workflow look like between a human and an agent? What does the human do versus what does the agent do? How do you rethink your process knowing what is possible today and how quickly the world is moving? And those are human process questions and challenges that companies are really starting to work through. But none of it is going to work very well unless you have your human and technical foundation sets that you can iterate quickly and learn as you go. There was always the common problem in sales. And it still exists everywhere, just the common vocabulary. So you say you have a champion. Your definition of a champion can be completely different than five of the sales reps in the same company with their definition of a champion. You know how we solve that? I come back with its own definition of a champion. And AI will hallucinate a number of other definitions based on what it's trained on, because it's reading the internet. So the way that we actually solve that, by the way, this is going to sound a little crazy. But we actually have agents that go look at your methodology, language, semantic definitions, and actually build what we call a semantic model, which is a dictionary, on what terms mean for each unique organization. And the way that one leader describes a technical champion versus an eb versus what a champion needs to them is actually it's different from organization to organization. It shouldn't be, but we've actually seen it be. And so that language is really important. And there's also technical language there as well. How do you define Catback to an earlier point, AIR? What does it mean to your particular business? What data are you pulling in from what's source to describe the answer? So Alex, let's talk about in your survey, like really now in a practical sense, how have you seen people implement AI where it's actually, let's say working. Yeah. So survey just to give a little color, we ran over a six month period basically an analysis on 30,000 workflows. So it was 30,000 workflows. And we just looked at what people were actually doing. We took away any bias. We said, we're not going to have opinions on what people are doing. We're not going to come in and create any surveys. Because there's also some bias in those surveys, depending on how you frame the survey, we're purely going to look at usage data in production with customers that are actually using an AI for real work and using it every day. And just to kind of break apart at a high level, what we saw the buckets to be, the buckets were number one account intelligence, which is just deeply understanding, both internal and external data on the account. Very helpful, because it usually takes so long to do research on an account. Yep, not just research. And maybe you can bucket this in research, but who are the people in the account? Who do we know? What are the stakeholders? What are power dynamics? What does the power map look like? How are decisions being made in the organization? So I was one, two is conversation readiness. A lot of that is stuff that you'll be familiar with. Meeting prep called deep briefs, building decks for EBRs, QBRs. Three was deal acceleration. So how do you actually build things like business cases and other deliverables to accelerate the deal? There was pipeline inspection, which is what you think it is, and then team enablement. And it was interesting to see the usage and also how teams were using this in the different personas. So one of the things-- and we talked about this a little bit that we found-- was most teams were using endgame to better understand what was happening. And a lot more of the usage was really trying to figure out what was going on versus generating the outputs. As I mentioned earlier, creating a deck, creating a document or an account plan is pretty easy. I can do that pretty well, knowing what should be in those documents and how well articulated they are to your value proposition, your methodology, your sales process. That was really a hard part. We saw a ton of momentum across teams really trying to understand. The other thing that was interesting was the best teams would work an account from every different persona So less silos, a lot less fragmentation. You would see a SDR Hand over to an AE AE would bring in an SE sales manager even CRO and CEO would be looking at that account Some of my favorite examples were CEOs and executives that were trying to understand what was happening within some other topic Accounts would come into end game and just ask a question and all the context was there and they can go talk to a customer You know without really having to go and talk to the rep as deeply And so that continuity across different teams different functions was was a pretty interesting Surprised I would say we had product marketers log in Look at an account build a case study and get it approved in a day with the customer saying oh my god You've been listening to me for the last three years. This is so incredibly thoughtful Thank you so much for for sharing what it is that we did together in my language and So a ton of compounding continuity when you have everyone working on the same information That information is constantly updated. It's real-time. It's grounded in your methodology Your frameworks and so that compounding team and organizational Interaction was what we thought was the most interesting and the most surprising It's a most of the different personas that you outlined in the survey are the reps some of the managers and then also rev up so Walk through you know what you saw from each one of those different, you know, stakeholder Yeah on the website a lot of it is let me prep for the meeting Let me come into end game. Let me ask a question. What should I talk about in this call? What discovery question should I be asking I need to build a deck? So we see a lot of deck generation If you want end game as an example to generate an our understanding Slide for you. It'll do it based on your methodology and frameworks again the deck generation part isn't that hard But doing it based on how that company sells and getting that right is pretty tricky So a lot of reps would use it to sort of sharpen their kind of in meeting interactions and have assets Going into those a lot of follow ups Kind of thoughtful deeper follow ups across every stakeholder within the meeting as an example a lot of prioritization Questions so we see reps coming in and saying what should I be doing today? And again when you have everything connected you can get pretty clear, you know responses based on you know your own Data and information on where you where you can be most impactful as a rep so on the rep it was really mostly how do I Make it much faster and easier to get better with customers Kind of automate as much as possible of the mundane repetitive stuff that reps don't really like to do and really customer focus So we just saw a lot more Lot more momentum on showing up to meetings way better prepared better messaging better documents In a lot less time. Sorry. Can we take a step back and Alex explain what endgame is where does it sit? What does it do? Because I'm assuming that our You've got you've picked the interest and there's going to be a lot of people listening that are going to be like is this a platform? Is it what is it? Endgame is a centralized intelligence platform. We basically connect to all of your Internal data methodology so we'll pull in data from sales force and gong and slack and email and all your enablement assets methodology messaging we turn that into a synthesized Knowledge and context graph that makes it very useful for LLMs to ask questions on top of and then we have different interfaces We have an endgame app that looks pretty similar to a cloud or chat GPT that's purpose built for go-to-market teams And a lot of our customers use our knowledge in cloud and chat GPT and interfaces they already use to get much more accurate consistent source cited Answers to their questions And so it makes it such that you can connect all of your organizational knowledge to a single system that can look at every single part of your business and that information is available across any surface that you work in and reps use it a ton to answer questions and basically build artifacts to the other 15 AI tools that We have purchased do those go away What happens what we're seeing is a pretty big Change in what build versus by means and so you have a lot of go-to-market engineering teams rev ops teams technical teams that are building agents What I think it looks like is you end up having a lot of your traditional systems of record a Salesforce You know as NDesk a Enablement system a gong what have you That are not really built for interacting with agents and so you're gonna have a layer whether it's endgame or someone else that Basically turns all the data you have into knowledge that agents can execute on and makes that available in Whatever interface whatever team is using And we're seeing a pretty massive move towards clawed clawed co-work Teams are building all kinds of crazy agents on top of that some still use chat GPT I don't know what the flavor of the month is gonna be next month because the world is changing so quickly But I don't believe you're gonna be effective unless you can have that centralized intelligence Across the entire organization available to every human and agent And that centralization is what I think really matters most and I think that will have a profound Effect on what other tools and systems you need like we're seeing teams turn off of traditional enablement systems Pretty pretty frequently because you can now connect your methodology and and kind of value frameworks enablement content To an LLM and make it useful for a particular situation in a particular moment in time So I'm getting ready for this meeting what discovery question should I ask for this Prospect based on the last call we had their industry their vertical You know what our messaging says we should be doing. It's a much more specific and personalized Um way of interacting with that information. What do you think is the biggest mistake that you see Companies make now when they roll in on AI. What do you do sales teams? I'll tell you a story that I think is funny Great It's more of a pattern that I've seen there is about three months ago maybe five months ago You go talk to a lot of sales leaders and they say hey, I have a pipeline problem Uh, and so I need a tool that helps me build pipeline and so they would go by 10 vendors Um each of them promised a three x increase in pipeline So if you if you buy 10 that's about 30 x and you're like I'm sitting pretty this is going to be great That those are the first tools that jumped out right away All replaced the sdr and jump your your pipeline generation right yeah and they didn't um you got zero x and you have 10 tools Yeah, so so now what do you do and so I think and this is uh this is an interesting conversation because it's one that Different teams are on kind of different levels of maturity with You got to get more technical and I think leaders need to build an instinct for how things actually work And the best teams that we see that are adopting AI you've got leaders that are prototyping things on their own They're thinking about how they might change the process Um or augment the process based on what's available and possible today They're really curious Um They're they're thinking through the art of the possible and then figuring out how to build with a few people internally They're the most forward thinking and have deep respect within the organization So seeing a lot of really good rollouts where say a cro with a few very senior 10-year reps and maybe a revops person Are moving ahead of everyone prototyping experimenting showing examples building a culture of of experimentation figuring out what works and then and then sort of standardizing on some of those things with the understanding that those might change It actually looks a little bit more like product development versus traditional Kind of revenue in some cases because you're trying to build sprints Build a cycle figure out what works Two questions who's going to support that over time With thin sales and then especially as things change And then what about the security of your data and your customers data inside these big frontier models or LLMs That's one of the most frequent conversations I have there are a few things that we're seeing happen Um revops teams are asked to get a lot more technical and and that's that's working in some cases in some cases It's it's a little bit painful You're laughing you're funny because you're being nice It's it can be it can be really really brutal uh and there's a there's a classic Service-minded revops orientation, which I don't think works Um the teams where this is working revops take some more product oriented mental model and they're trying to build a system Um and manage it So that's one pattern we've seen the other pattern we've seen quite a bit of is Either the CTO's office or applied AI in the revenue org is growing exponentially. I saw a stat, I think the number of go-to-market engineers or job postings went from something like a few hundred to over three thousand. At the start of the year, don't quote me on that exact number. I'm kind of remembering, but the curve on roles open for technical go-to-market people, they can come in and build and maintain systems is increasing at a pretty crazy rate. Hey grumpy old man, ready? Revenue ops people be more product oriented. Paking me farther away from the human factor of the sales savvy, the art and the science that actually happens in front of a customer is a huge red flag to me personally. You've had some bad experiences with product people, my friend. There's a lot of product people that go company on bringing them all together. I'm talking from a seller. You're going to do a workflow that's going to end up with me. Walk me through how you're going to bring those two worlds together. I'm not asking you to I'm not putting your product on. This is great. This is great. Here's what I mean. Service oriented robots, build me a report, build me a dashboard, build me a comp plan. You're basically responding to leadership, your CRO asking you for something and you give them an output. Most robots seems today are kind of responding to requests for dashboards or requests for a field validation in sales force or what have you. Less on the system side. That's what I mean by service oriented mindset. You could in a negative way say order taking and I don't think all robots people are order takers by any means, but that's one orientation. Another orientation is actually a deep understanding of your customer. What are what are reps actually trying to do? What are their customers trying to do? And building a system to better support your customers, deeply understanding your customer. So the customer's introduced a simplicity part to me is actually the most important part of the product mindset that I'm describing. I love that. So of the 3,000 people that are being looked at, what skill sets are you identifying that would be really, really important to bring into your company? Very customer oriented. Our first operating principle at our company is start with the customer. And the customers now are not just your reps and your managers, they're also the agents. Number two, you have to understand how software works. And you got to learn really quickly if you don't. And number three, you got to think as a system builder versus an army builder. And you've got to design processes, mechanisms, architecture, human and cultural mechanisms to get people to learn and compound that learning much faster. And the rate of change is such that if you're not getting better at whatever you're doing week over week, month over month, you're going to fall behind really, really quickly. So that ability to constantly be pushing the art of the possible, figuring out how to make things better and how to improve what it is that you're doing, I think is so necessary now. Because if you didn't change your system or process five years ago, even for a quarter or a year, you were fine. Now I don't think you can afford to do that. So just a faster kind of startup-minded orientation, if you want. Sales skills, do you need? From a good market engineer? Yeah. I think for what it's worth, you need to have a good understanding of sales, but you need to be really good at listening to customers. And the interesting thing that I've seen is even even people that are younger, earlier in career, that aren't as deep in sales, but are really good at being customer-centric. We'll go talk to their reps, go watch a lot of calls, go talk to leaders, and learn very quickly about the gaps in their organization through kind of a first principles mindset. And if you're really trying to solve problems for your customer, you can ask the questions that, you know, you need to get answers to to go solve the problem or remove the problem forward. Let's talk about productivity. Okay, bringing in all these sales tools, we're adding to the REBOPs team, they're starting to build stuff, the reps are building stuff. Where's the productivity come from? Because overall, I'm the CRO, I get judged on average productivity across my sales force. If I can't keep that flat to increasing and definitely increasing, if it starts to decrease, you know, I lose my job. Tell me where AI is going to have the biggest impact on sales productivity. There's three metrics that we typically track at a high level and we break those out for customers. It's revenue per rep, you know, overall productivity, you can break that apart in a few different ways and then ramp. Interestingly, the thing that we see move fastest is ramp. And the reason is if you have a new rep coming in and you have access to, you know, complete information on all your accounts, you can ask a question about your book of business and what you should go do about it. You know, we see ramp acceleration, you know, going from six months to two months and we've got some numbers on that I could share later. A lot of account coverage change. So if you right now, as an example, as an enterprise rep, manage, you know, 10, 15 accounts, being able to manage a larger book and keep that quality bar at the same bar or higher. We're seeing that change quite a bit very quickly in works. A lot on account management. Similar in terms of account, you know, to reparations. So we go into scenarios where you have, as an example, tier one, tier two, tier three accounts, tier three accounts were totally untouched because teams don't have the time. You build a workflow where you just send, you know, if you proactive nudges with a ton of account context and you're able to take over a larger book and see NRR move and expansion and renewal numbers move as well. The biggest kind of unspoken thing that we're seeing is very different hiring, rehiring practices. Like a lot of orgs are getting redesigned to have fewer specialists. So you see the biggest impact at the headcount team structure level. Yeah, fewer, fewer siloed functions, much more full cycle. Kind of orientation, fewer essays per AE, fewer value engineers, a lot of the sort of adjacent supporting functions we're seeing this collapse as well. So if you look at, you know, revenue per employee revenue per employee is going up and is going to go up because you have two ways to cut into that. You can look at it through an EBITL ends or you can look at it through revenue growth lens. And the biggest impact we're seeing is do you need as many people to grow at the rate you're growing? And if you want to grow faster, do you do that linearly with headcount or not? So the bigger conversations are more on the FBNA side on what does the structure of our org look like? And that's that's where you're seeing the the biggest shift for sure. So let's go into that. Usually as the CRO, you know, you talked about I got ramp time, I got productivity, I've churned right? Said that you're helping on the product on the ramp side. You see that the most you're helping a little bit on the productivity side. Normally on Monday pressure, if I can't get my productivity to increase, I'm under on my cost of customer acquisitions, probably increasing. And you get in a conversation with the CRO and the CFO and they're saying, well, here's what we need to do. We can cut the SDRs. We can cut the comp plan and we could change the manager to rep ratio. Those are usually the top three things that the CFO is going to throw out there. Right? And if you look at what the organization will probably look like in two to three years, do any of those change? Do I have fewer SDRs? Do I have a higher manager to rep ratio? And then can I keep the comp plan the same? I think you're going to have bigger comp plans. Correct. I think you're good. Yeah. Because every year usually as a company gets bigger, they keep cutting the comp plan. Yep. I think you're going to have smaller sales teams in general. And you're going to have because I'm going to be a lot more productive. So I'm going to need as many heads. And what about the other? SDR is a really interesting topic. A lot of the AI tools got started on SDR workflows. That's a deeper conversation. The email channel is so screwed up that SDR means different things to different organizations. I am seeing a lot more when we talk about revops and go to market engineering, building much more efficient pipeline generation machines. Maybe there's an SDR. Maybe there's a lot of semi-automated work being done for an AE who actually owns their own pipeline. That's shifting very quickly in organizations. But Can you say it earlier? I'm just brainstorming here. Couldn't I have some SDRs instead of having Couldn't I have a group? of them that are doing research for the AES because you said they're doing a lot of research and account information and have them prep calls for the AES to make the calls. Not the. You totally could. And John, I think what I'm getting at, and this is changing very, very quickly, is you can build an automated system. As an example, you could connect, and a lot of our customers do this to endgame and have all the pre-called briefs, research, like stakeholder maps, messaging, positioning, even decks created for the AES without a human having to do it. So we're definitely seeing a lot of tension on do you need people doing that work at all? Organ, an intelligent orchestrator, with a good understanding of what I want the call prep sheet to look like, what I want the deck to look like, what I want the POV to look like, including spinning up account-based marketing campaigns, which we do, by the way, completely programmatically, once we've identified target accounts. And so that ratio is changing. I don't know if it looks like no SDRs, or there's fewer SDRs that kind of manage a lot of that and they're more technical, but I don't think you need as many individual SDRs building that stuff and sending out emails because you can train an agent to go do a lot of that stuff really well. I think the promise of speed with AI is incredible. And I think the way people are going to look at it, like it's going to become more of a customer experience, starting with the customer experience, because what customer on the planet, including us, wants to deal with multiple people in a buying process. Exactly right. Strongly agree. Wants to deal with the time it takes to go from when I have my needs identified to orchestrating something that's going to show me how my needs can be solutioned, how I proof of concept those types of things. So I think for me, I think the customers that are really looking at not only the seller workflows, but how the buyer buys. Strongly agree. Yeah. And what's changing there, I think, is really, really huge. Because for me, I think the question that you guys have been talking about is powerful. Are we going to have more? If I was a run in a company today, obviously, I would like to do 10 times more with the number of people that I have. I think you can today. And I think the gaps are, there's a lot of emotional and cultural stuff. Both of them internal though, right? It's like, yep, but we've always done. That's what that's right. That's right. And so, so if you remember, we started talking about how we're people using endgame through through our survey. I gave you some examples of how they're using it today. Everybody's looking at the same information and trying to get better answers to their questions and they're starting to automate stuff. What I didn't tell you is what happened in the last month. And what's happening in the last month is you've got technical teams, op teams, even engineers getting flown in to RKOs, Skows, and saying, go look at what teams are doing and try to make it move faster and more predictably with automation and AI. And so we're seeing fully autonomous systems built semi-autonomous systems built. I'll give you, I'll give you examples. He's trained on the history of the Soviet Union. Really understands how difficult it is to create change. Is doing all of my pipeline reviews, flagging issues in accounts that are stalling. Actually, emails my reps asks for information on accounts that isn't in the CRM or any underlying system and then goes and updates it. We'll build full assets just by listening to calls that I have on Zoom or Gong or whatever. We have seen in the last month or two just an incredible amount of momentum in teams being assembled very quickly to go build out basically semi-autonomous or autonomous systems. And I give you one example. Another example is let's look at tier one, tier two and tier three accounts. Let's go see how much we can automate tier three accounts under 100K with repeatable renewal motions. How do you scale account management in a way where it's kind of SDR-like but very specific to the business. We're seeing a lot of what I call the CRM hygiene machine use case. You can basically stand up on orchestration service. It just goes, listens to all the calls, looks across all the data and updates your CRM and keeps it like fully accurate and can make changes every two minutes and nobody has to update CRM anymore. We're seeing our customers build full whale dashboards interactive with geographies of where every person within the account lives physically like build offers on what we could send them as gifts. Everybody's working off of the same kind of internal application across the entire company. So full imagine building kind of your own large DLCRM that's fully automated enough to pay. Alex where I see this there's a big collision coming because there's this intellectual curiosity of people like you and I would argue Johnny and I as well that yes the possibilities of all the things that we can do and then you have this train going down the tracks organizational capability and business model agility. I hear this all the time like we've got AI SWOT teams they go and they gin something up and then and then all of a sudden it meets a roadblock somewhere that says we don't have a business model that will allow us to do that. Where do those worlds come together because that's what I'm seeing now. Pretty top down directive leadership. In times of change when nobody knows the answer whose job is it to come up with the answer and either fortunately or unfortunately depending on how you look at it it's usually a leadership. Stronger build and part of the reason is you have a lot of things that can kind of do 80%. It's easy to prototype with AI it's really hard to turn it into something as production worthy that works for your organization. More deeper engagement with domain specific companies that are solving these problems. There's a lot of like horizontal AI right now. For us we're actually getting pulled into more forward deployed interactions where leadership is saying hey I need you to fly in here take a virtual badge or a physical one depending on the situation and help us actually implement this and roll this out. Including you know the kind of AI enablement part we're doing a lot more of that in the last two months than we were ever before because we're getting asked to. I think and also help on how do you find any John your earlier question on who manages and maintains this. You got to kind of promote from within or bring somebody in pretty quickly to kind of have a throat to choke on how is this going. The angry old man then let me ask this question but what about security. So these people that are just jinnying up from revops or the reps on their own. Jinnying up all this stuff where the customer information and their own company information is getting sent out to these frontier of large LLM model. Where's the security. It's a total no one is pressing that button and saying hey this is a problem. All I'm referring to. It is it is a huge. I'm going to go even deeper than security for a second. I'm going to call it security governance. How do you manage what systems are being queried how they're queried the cost you'll hear in a few months you'll probably send me an email being like all right so you said LLM costs are going to increase. It's going to get really wonky because every one of these individual reps or individual managers is running thousands of queries on. On these underlying systems and rebuilding every answer from scratch look I'm biased because this is what we do but I think you have to have a central well managed well governed data and found in the nation layer that also manages things like our back and you know do your sales force permissions actually getting herited by. What I'm querying and you know what can I send to which customer and policy enforcement and I think the best The best way to do that is work with either a team that's building it for yourself internally, which is expensive, and you have to have some pretty technical people that also understand the domain of sales. Or you find a centralization, central context, management, agent management platform that allows you to go focus on the other challenges that you have in your business and figure out how to operationalize this in terms of processes and teams, which I think is the better answer. But we've seen quite a few security governance compliance issues, and that's only increasing day by day. I'll show a little bit about, um, with, could there be an over reliance on AI? If I'm a young rep, and I'm trying to grow where I lose my, I just don't have the intuition, the judgment, I don't pick up the subtle skills and idiosyncrasies, those types of things that I, that make a great rep, a great rep. I see. Through the repetition of essentially like sharpening the saw, I've done this so many times with so many customers, I start to really figure out which way I got to go, and I can't rely on AI to tell me which direction to go. Because it's great to do account research, but when I'm in the conversation with the customer, I have to perform, and if I over, if I have an over reliance on AI, I don't have that intuition, I don't have that feel, I don't, I can't pick up the subtleties of which way to go. You asked a specific question, I'll answer a specific question, but I think this is a humanity level question that you're asking, by the way. Um, it's bigger, it's bigger than sales. I think this is, who are we as humans, um, you know, in a year or 10, but from a sales perspective, I think it's a huge problem. Um, I talk about this a lot internally, um, I, I say don't outsource human judgment, don't outsource your brain. Um, even when we do enablement and training sessions, by the way, um, for other teams, I tell them that judgment and human judgment is why they still have a job, and they should really invest in that, not outsource that to AI because, um, which, which by the way, get, get, get get, get them to pause and, and, and sort of, you know, be like, okay, so I still have to, you still have to think, um, and thinking is arguably more important. You, you don't have to do a lot of them and not in this non thinking stuff. So really get, get better at, at thinking and articulating your thoughts and, and understanding, um, the human on the other, on the other end of the conversation, I think it's more important than ever. This is, this is one of those related process questions. This is where I think Rebops can be massively helpful. And you, you give everybody a ton of AI tools and agents and say go figure it out on your own and I see reps for what it's worth. Spend almost all their time just like building assets and building artifacts and kind of like generating the same document 50 different times and it's like, no, no, no, that's not what you're supposed to be doing. You should just reduce the amount of time you had to do that period. And you think really deeply about the three questions that you need to ask in this next call and like what the dynamics are, um, of the people that aren't even in the room that you, you have some context on now. And so I think that's a huge, huge, huge problem. Um, and it may be this is a good tie back to, to cap your question on like what does Rebops do for the customer? Like that's a customer problem that John is describing, which is make the job for the rep focused on the job that they're supposed to do. Um, because right now they're dealing with everything that, that isn't quite that I think. And for me, put on the back end of that that they're supposed to do, which are guidelines, guardrails, do it faster with higher capacity, yep, with more potential for intellectual curiosity. Yep. Me, I've changed, I mean, I've been on a journey myself like everybody's on. And I think it's one of the, I think it's one of the greatest times to be alive on the planet. Uh, it's probably the greatest time to be alive since I've been alive. And for me, it is the, it's a thought partner. I have changed my, and it's always going to be probably in my lifetime. It's going to be a thought partner. And when I talk to reps, I am actually doing the one on once and I'm using all of my intellectual curiosity, doing it faster and higher capacity and, and higher number of, uh, deals that we can go through or what have you. So I think in the, in that context, if that's the context that you have as a thought partner and, and you can really test it, like I can tell whether somebody's using AI as a thought partner or, yeah, AI to try to replace their, their activity or their jobs or what have. It's clear. I totally agree. I think, I think the, the, in the next few years, you're going to see a lot of like deeply personal kind of agents that are really closely tuned to the individual, uh, the, the prototype of Sasha that I was describing is, is, is, is, uh, on that way to personalization. It's all. And, uh, AI is very good at mass personalization. The, the, the point John though, the, the, the, the, you were making or the question that you were asking on, what does this do to reps? What does this do to humanities? Is, is the big question like outsourcing, thinking, uh, you know, just looking to whatever LLM or AI as a response without really kind of like using your brain, um, I think is, I think is a huge, huge challenge for what it's worth. Alex, also, I also mean personalization of the one using the AI is what I was. Oh, yeah. So in that takes consistency standardization, not 100 different LLMs. And so for me, the fact that I could have a thought partner that knows me, thinks a lot like me, you created Sasha, what you just described as personalization. Oh, sure. Seller, which I think is, is absolutely fascinating, which gives eating capacity. There's this really interesting dynamic between certain things you want to be centralized and you don't want to be personalized. And then certain things you want to be personalized. And AI by the way, there's a difference between memory and knowledge. Yeah. So, so there's institutional knowledge, organizational knowledge, which is this is like the collective knowledge of your organization. Then there's memory. And by the way, as is the case with humans, memory can turn into knowledge, but each, each sort of age in each person has their own memories, preferences, like my writing style is different, my sense of humor is I think better than most, but my wife disagrees with me. So there's a lot of different kind of nuanced things that are unique to each individual. And by the way, as getting to a place for when you talk to a system like that, it actually remembers, which is a lot of where we're investing right now is like, as you talk to a system and you correct certain things or you uncover new things that all gets captured in this single brain. And I think I think it's the most interesting thing in AI right now is, is how do you create this brain for an organization that gives each people their kind of creative, personalized, independent channel, but also codifies what is important to the org. That loop is I think personally very fascinating. Great discussion, Alex. That was so good, dude. Yeah, thanks for coming on again. This was fun. Great to see you both. Well done, buddy. I was going to say one thing and I have a little pitch for you, which is a little, it's 50% joke, 50% reality. All right. And what if instead of all that 50% joke and 50% reality, I don't know yet. I'm sure. There you go. There you go. So there's a new fascinating market where if you think about the primary users of what you all do to be agents, not humans, what is agent enablement, agent training, agent performance management look like. And that's where I think the world is going. So I was going to, I was in a chat like, what does it look like to go and onboard and enable and train an army of AI agents? That's here. I think it's, I think it's here. I think it's orchestration of agents. I think it's, I think that skill set of people are listening. If you're not creating agents, just to, I'm not saying go crazy outside of your company's guardrail. So what have you? But if that's the next frontier, I mean, and to learn how to orchestrate those, to create great outcomes, the humans always going to be involved in that, my opinion. I agree. And my joke was going to be an expecting your website to get an update where there's a four humans and a four agents version. Yeah. That's all, that's all cap. All right. Thanks a lot. Thanks to everyone. Thank you so much. Not the episode of the Revenue Builders 5. Thanks for listening to today's episode. If you enjoy the content, please subscribe. Ray and review the show to help us reach more people. This show is brought to you by Force Management, where we help companies improve sales performance, executing the growth strategy at the point of sale. Check out ForceManagement.com for more information. (upbeat music)