Why the AI Agent Era Will Change Everything with Bonfire Ventures' Brett Queener

Sean Lane 0:00 Hey, everyone, Sean, here, before we get to the show today, I have a bit of an exciting announcement that I want to share with you our operations podcast audience. I wrote a book. The book is called the revenue operations manual, and it hits shelves this month. So why a book? Why me? Why now? In January of 2023 LinkedIn released its list of the 25 fastest growing job titles in the United States. The number one job on that list was the head of revenue operations. This book unapologetically champions the operator. It's co authored by my former boss, Laura Aden, an amazing operator in her own right, and Laura and I believe that operators are at the center of what makes companies run, and it's time that we all embrace that truth. So this book is a guide for operations practitioners. We've interviewed over 50 world class operators, many of whom showed up and came directly on this show. We've got examples of these amazing real world role models from a bunch of different types of companies industries, as well as cautionary confession corner tales for when things didn't go quite according to plan. The book also introduces and codifies this idea of the revenue operations mindset, which is a series of beliefs and statements that we believe that, when taken together will change the way that you approach your job for the better. If you want to learn more about the book, you can always check out revopsmanuel.com I'm also going to include a promo code in the show notes for revops 20 to get 20% off the book when you order it through our publisher, Kogan page, you can find that link in the show notes. All right, on with the show. Hey everyone, welcome to operations, the show where we look under the hood of companies in hypergrowth. My name is Sean lane. Have you ever had a conversation that just sticks with you for days? It challenges everything you previously thought to be true. You sit up at night rerunning parts of it in your head. I had one of those recently with Brett queener, and luckily for all of us, you're about to hear it. Brett is a managing director at bonfire ventures, but he's not your typical VC. Brett, to me, is an operator at heart, getting his start at Siebel systems in the late 90s, early 2000s before joining salesforce.com for over a decade, running their first revenue operations team at the company, and later becoming the GM of multiple products and business units. And later, Brett was also the president and COO of smart recruiters, which ended up becoming a unicorn. Brett was also one of the earliest guests on this show, and I've been fortunate enough to know and learn from him ever since then, anything that Brett writes, I read, and anytime I'm lucky enough to chat with him live, I know the best thing I can do is just get out of the way. You'll see all of this is why someone runs on the implications that AI is going to have on the application software market caught my attention and kicked off the cycle that I mentioned at the top of the show that has caused me to basically question everything in Brett's substack, which is appropriately titled Tales from the bonfire. He wrote two separate pieces, one on the three ages of application software, and another on what he believes the Third Age, the AI age means for all of us in our conversation, he breaks down those three ages, we dive head first into an agentic future and why all of us need to question what we think it means to do our job well. To start though, I asked him to take me back to his days at Siebel at the beginning of his career, and what the first era of software, the on premise software era, was like, Brett Queener 3:47 if you go back to before Siebel, my first job I worked, it's like my last honest job, and I worked at a medical manufacturing company in Milpitas, California. Here's how long ago it was. There were no tech jobs in San Francisco, so I was a young Yapi living in the marina, commuting an hour and a half each day to lovely Milpitas, California. And I was that guy who brought them from DEC VAX minicomputer, where it was a dumb terminal, there was no flexible UX into client server. I remember trying to justify the expense for a 386 Pentium computer. It was like four grand. So like when I think about the on premise software era, well, let's just talk about software in general, which is, if you think about the software interface, like, Why does software have tabs, list, views, form, views, search, etc, they're basically dumb uxes into the underlying tables in your database schema. And if you think about it, if you were going to go do something at Siebel, we did salesforce automation, you would have accounts, contacts, opportunities, quote sales, orders, or whatever it might be. And you know, there were records in red. Etc, etc. And the value proposition for why people bought this was so that management could track what was going on, so that, in theory, they could do analytics around a process that they cared about, to understand what were the issues and where could they go, pinpoint what they could work, whether that was, Oh, I'm not creating enough pipeline, or of my pipeline, my close rate is poor. And the close rate is poor because it's top to middle of funnel, or middle funnel to bottom of funnel, or it feels like this competitor is showing up, etc. But the interesting thing about the data, the data only got in there. If your ICS and the key salespeople enter this data in the system. But to use these systems, if you think about it as a human, if you wanted to go do something, or what I call like a CRUD action, right? Create a record, read a record, update a record, delete a record, or, God forbid, look at a full set of records, and maybe create a dashboard, if you weren't stupid, so you knew which dashboard to go build. And hopefully the answer you're looking for wasn't across two dashboards, because then you needed, like Jimmy Neutron to figure out the trends across two or three to kind of decide, aha, we have an issue here. But as a user, you just wanted to do something. Hey, you know, I want to contact Jim and Marriott, or I want to know if we have similar which deals we were competing against this person, et cetera. You had to think about what you wanted to do and then translate that. Okay, how does this application work? That I would go do that, and what I call is you have to hunt and peck that translation of I'm just trying to do this. I mean, they're trying to do something. I'm trying to answer a question. I have to hunt and peck across the application. Understand how your application works to do that. Now, in the on premise, the challenge with the on premise world was, well, one, we shipped CDs. So you had a CD, you had to install it on a local device. If you wanted to change that software to meet your specific business process. You had to hire a Anderson Consulting or somebody like that. And they would, you'd have to pay them like 10x what you're paying for the software. You would pay for it all up front and hopefully see the value. The other problem with this model was, let's just say Siebel came out with Siebel 3.2 and you were on Siebel 3.1 well, you'd had to install the new CDs, and then you'd have to hire the consultants to make sure you kept the customizations you had before, and then any new customizations to take advantage of the new feature. So the deployment model was horrible. The user interface was the user interface of, oh, by the way, a lot of applications today. That was sort of that era. It was an expensive market in that. This is where you saw, like people that are trying to do all in one suites versus sort of standalone suites. But the interesting the economics of this business, because if you didn't make your number, because you didn't have much Perpetual Revenue, you're kind of screwed. And so sort of these very violent market dynamics in that in that era where, similar to the database market, where the number one player in a market would have north of 50% market share, and like two, three and four would split the other 20, and then it didn't matter. And so you had these people that emerged as winner within a category, Siebel for CRM, PeopleSoft for HR, business objects for analytics, Hyperion for fpna, right? So there was also this weird thing where, like, what you would do is you do these partnership with the SIS. My first job at Siebel was managing the Anderson Consulting relationship because they could make $10 for every dollar of software. In theory, they were supposed to be neutral, but what I would do is get them to commit a large number of certified people, and they had to get those people off the bench. And so even though they were supposed to be neutral, software companies, because these products were hard to use, buyers would look to the S size to make choices, because the end product wasn't what you bought. The end product was what the product you bought, what was deployed and what the consulting firm could do with it. So that's how I think about sort of the on premise era. But from a user interface perspective, look, you're hunting and pecking across tabs, fields, etc, and it's not helpful to you as an individual user, like if you're a salesperson, Siebel didn't help you sell. It may have unearthed things to managers aware where you were deficient, and then maybe they would determine, by looking at it what it was. And then hopefully they would translate that into how to make you a more effective seller. But the software itself, Salesforce automation, didn't make a seller more effective. The thing we always joked about, even at Salesforce or Siebel, which is, if the way you go about selling is dumb and you automate it, well, you're automating idiocy. So best of luck. Sean Lane 9:51 You're hopefully getting a sense for Brett and his perspective and why I like talking to him so much. The punchline at the end there bears repeating. Siebel. And other software companies from the on premise era that he's describing didn't help you sell. They didn't make sellers more effective. You could take crud actions, as he calls them, create, read, update, delete, but that doesn't mean you became more skilled as a result of using that particular tool. Starting to sound familiar, so if that's the on prem era, things must have gotten better when we moved to the cloud software era, right? Brett Queener 10:27 My thesis, it gets worse. Why? First and foremost, Yeah, completely. The innovation there was the deployment model, yep, right at the time, you didn't have to pay a lot of money, pay as you go. And then, of course, SaaS companies were like, Let's get you into annual subscriptions, right? And the big innovation that Salesforce rolled out was really metadata customization, such that you could make your customizations, and often you could do it in a declarative way. And we took customizations, which used to be a job, and set up to either consultants or people that stuck in it, sort of stuck between it and sales. And we made it declarative, right? And so like sales admins before Salesforce were not heroes, right? If you go to Dreamforce, you try to figure out, like, why is all this energy around this application seems kind of not that easy to use for an individual role? Is because you took a class of people that were sales admins who consider backaf support and good luck hiring a VP of Rev ops that isn't a chapter chair of pavilion in Des Moines who's going to work for less than 200k right? We made those people champions, right? And we did give business more control. So that was great. Where I'd say it got worse, was it still tabs, list views. You look at Salesforce tabs, list views, form views, etc. We make customization really easy, so now an admin can go in there and just start mucking with the screens. Here's some fields. Here's some more fields. Wait a minute. The hell's going on here? Oh, right. Second, it used to be really hard to integrate applications in the on premise world, right? I got an on premise software that's being customized, etc, and now there's another piece of on premise software. Oh, but those softwares go from three, one to three, two, and that one goes from four, one to four, two. Then the API breaks, like that's a mess with REST APIs. Oh, well, we can integrate these applications really easy. So now the average sales reps using like, nine applications. So wait a minute before I was hunting, pecking across one application. Now, to do my job, I'm logging into nine different applications, and often when I do something, no longer do I have to be like, I want to do something. I have to think about how to trans, navigate, navigate one app. Now I need to go from one app to another app to another app. And then people are like, Oh, it's okay. It's a screen within that. Or we've all put it in Slack, where it's just a bunch of busybodies chatting, and you just like, Stop the fucking madness. Like, what? Like, I'm just trying to sell something. I'm just trying to, like, maximize my w2 I got a mortgage to pay. And then it got even worse, because the nature of SaaS and then the just flooding of venture capital is, if you aren't a great company, well, you got some recurring revenue. You can live for a while, right? And so in a given category, you you're a buyer today, like, what do I buy? What do I buy for marketing automation? There's 1000 to look at. Oh my God. Now I've got to spend a ton of time trying to figure out which of these is better, and go to blogs and listen to the Brett queen or Sean Lane podcast, way too much complexity. Now, to be fair, we did see sort of the what I call product LED. Now, I don't want to call them a hack, but it's a little bit of a hack. What you do in product LED is you figure out a certain process that you do very well, and then you stitch together the screens and the workflow in sort of a declarative play that, you know, the user can just follow right now. The reality is, for many of these, these were products that sort of started out with individuals and grew where there wasn't something to sell, and so they had to do this from an acquisition perspective. But look, there are good examples of these. If you get it right, it's great, but they are more the exception, right? I remember when product LED? Where was I product LED. When was it product LED? And now everybody's like, we're sales LED. Most people that say they're product led, basically just have a much longer free trial where people, instead of trialing to see how the product would work in a situation, actually start to use it, and you monetize later. But I still think if we look at the majority of software applications today, in cloud, there are list views, tab views, etc. And I don't think for the individual user, we become more productive. And in fact, and look, I used to be the first head of Rev ops. And look, when you build these software companies, what do we do? We get champions. We make them feel good about themselves. They really understand I use our product. We might even write a book that the founder wrote, and we follow this approach, and then we put the person on stage, and they feel good, and they feel validated, and they're the master of how to use Marketo, or they're the master of how to use Salesforce CPQ, or they're the master of how to use clay, right? And then what happens is, I feel like we lost our heads a bit. Where we've become masters at using these tools or stitching these tools together, like, as opposed to like, better selling. I'm all for marketing attribution and marketing automation, the rest of it, but it feels like we spend more time trying to figure out what an MQL is, what the attribution is, as opposed to like. Does our marketing suck? Does our messaging work? Do we have any brand? Does anybody know about us? Every startup is like, okay, yeah, so I'm gonna hire an outbound SDR, because I don't have time to do marketing with my first 3 million, and they're just gonna dial. And I'm like, they're gonna dial and no one's ever heard about you, and you're gonna take a 22 lax bro, it probably wasn't a major in English. You're gonna put them in outreach. They're going to pound messages that. Who's going to write the LAX bro? Well, how's that working? You know, or the one I love in 1990 when was this in 2001 at sales for Siebel, we really had to kind of focus on creating real pipeline, because up until then, we had done what I call the partner jam scam, where, basically, hey, buy some of our shit. We'll buy some of yours. This is before, like, people realize that, like, Round Tripping is not a good thing. We also sold some stuff, but, but the economy started to dip, like, oh, we need to do real marketing, real demand gen. And I was like, wait a minute, we're creating all these leads and sales isn't following up. So it's the classic, what's an MQL, what's an SQL, etc. Remember, I'd worked in a factory. And I was like, this is like, a factory. I produce bread, but if I take the bread and I throw it out of the factory and Sony picks it up four days later, that's some rotten, crusty bread. So even in the on premise world, I created this thing where, like, here's a process, what's qualified, and then here's the criteria, and the reps have, like, three days to follow up, and if they don't follow up in disposition, I reassign. It little draconian, but it worked, but I'll go to Accounts. We're like, Okay, I got HubSpot, and then I got drift, and then I got chili Piper. And it's too much to ask the prospect who they are and what company they work at, so I've integrated this with this and Clearbit, and then, you know, they don't really want to talk to an SDR, so we just let them schedule meetings directly on the reps calendar. And I was like, so you use a seven piece of tools to run the most fucked up process ever. Because we all know, if you hand a prospect on your website to your reps. One, they're terrible at self prioritizing their time. And two, if you want them to tell you if that's qualified or not, it's a mess. And then, like, three months later, you're like, Hey, why are enterprise reps prospecting like, Well, it turns out their calendar is full all day with these SMB, looky, loser. Aren't we all qualified? And you just go, like, we have to ask ourselves. And you're chuckling a little, because you go into these accounts and you're like, what happened? Yeah, like, we have a process. What's the process for doing? Let's look at it from first principles perspective, and then the tool, or how we do it come secondary, but the tool we've kind of become like the tools master us, as opposed to like, we're trying to become better salespeople or better marketers, or better SDRs or better customer success people, or like, at Salesforce, I created the first CSM in the SAS industry. I'm not going to get into how I created it, but it's a funny story for another day. And you know, love Nick like, Salesforce did amazing job creating champions and gain site, but like, I'll go and talk to a CSM and be like, how many hours a week do you talk to a customer to what? Yeah, the rest of them running these plays. I'm like, no, no, no, no, no, no, no. You were created because we're early and our products kind of screwed up, and our salespeople don't do a good job of qualifying, and you just got to get them love so that they love you and you send them schwag and the rest of it to buy us time to get the other stuff fixed. That's like job one. You spending three hours a week with a customer, and they would look at you like, really? I thought this was all automated. So I'm going on a little bit of a rant. That's a little bit of what I think about, what I call the cloud era. Better deployment model, easier to customize, I don't think easier to use. I Sean Lane 18:41 could listen to Brett all day go off on everything that has gone wrong in SAS, and I'm not pretending to be above any of these errors. I've definitely spent long periods of my career focused on mastering how to put tools together, rather than figuring out whether a particular function maybe just as he puts it, sucks. I had a CRO tell me once Sean, no one wins an award for going out of business with the best systems, and that has stuck with me ever since all of this preamble that we've been talking about with these eras is leading us to this moment that we find ourselves in right now. Brett believes that the eras where software did not actually improve the productivity of end users are over, and what's possible is shifting before our eyes. Brett Queener 19:30 So I've been doing a ton of research on because, like, people will come to me about, hey, I need to do this for SaaS. I need to do this for SaaS. I need to do for SaaS. Like, I'm old. I've been at SAS longer than anybody. You ask me a product question, you ask me go to market question. I'm not always right, but I'll give you the 80% correct answer in like a minute, because anything you're considering doing, I have effed up 10 ways to Sunday, so I'd be like, don't do that. But I was starting to think about, like in this AI world, how much of that advice is obsolete. And so I started doing. A bunch of research, talk about 50 people and what I'm trying to do, and it's hard, because it's not my day job. My day job is investing in founders and trying to support them, but like, what is the new playbook for running a SaaS company in this, whatever you want to call it, agentic world, AI era and the rest of it. And when it hit me, I got to admit, I walked out of a meeting and I was kind of shaking and in my head, I was like, Oh, holy shit. I don't know what this means. I gotta write this down. I think it changes everything. And when it comes back to it comes back to this UX interface, which is, what if software was the equivalent of you hiring a super smart, expensive assistant they just worked for you. So let's pick like enablement as a category we could is the right thing, enablement that somehow there's answers and it's a doc, and we put it in high spot and some scar schema repository, and the rep leaves the application, and then searches, because the answer is in a doc that hopefully has been updated later, and maybe somebody new to be updated because somebody marked it up, or was attached to an opportunity, and then some Poor set of librarians were keeping this stuff up to date. That's not the right way to think about it. The right way to think about it is, if you have 100 reps in a company, what if I said to you, let's say there are 100 great sales enabled. There's the top 100 sales enablement people in the market. They're not cheap. Maybe costs you 150 200 grand. What if each of your reps could hire that person, and all they did was work for them. What would they do for them? What questions would they ask? What would they tell them? What would they guide them on people? Like, that's crazy. That is nuts. I mean, that would be like, $15 million we can't afford that. We gotta cut I mean, our VC says that our our magic number is too high, and the amount of money we spend to enable reps is too high, and we're still fat on the hog, and we're in the zurp era. We're too virtual, and we're not hard enough. I said, let's just say you're a software company cures cancer. And they got to work through pro bono for a year. Forget that. And they what would they do? What would they say? Would a rep ask them a question and they go, Oh, I went into the repository and I found a doc that was updated three No, they'd probably be like, Hey, Mr. Thing, you're running against this customer. You're competing against this thing. It seems like you haven't really done a good job of discovery, because we don't really know what the real pain is. So if we're going to compete like, I could say, Hey, give me a data sheet to compete against x. I'm like, I'm not going to give you a data sheet comparing a set of features. Let's go back like, on the Discovery. Do we understand their core pain is. And do we think that core pain is something that we are uniquely good at? Because if so, we can set the right competitive framework, and then I want to give you the information against that framework. That's what a really good person would do. We'd be like, Hey, I'm up against what content is really good to use in this situation. Well, you're selling into a financial services firm in this sub sector, and this is your competitor, and this is your persona. And look at the last four deals that we were in that same scenario. Here's what the rep did that worked, here's what didn't work, and here's what docs worked. Would you like that? And what's interesting is that answer I gave, I just made it up. It doesn't exist. It's not written anywhere. That's a real time answer that somebody gives because they're super smart. Now, the customer would say to me, but you know what, Brett, if I hire this sales enablement person, it's going to take them, like, six months to be really good at my company, because they're going to understand my process and my competitors and what matters. And I'd be like, what if we could do it in a day? What if in a day I could give you somebody who's really good at sales enablement and really knows your data, your processes, your docs, etc. Would that be interesting? And instead of it being $200,000 a year, I don't know, would you pay 50 bucks a month? No, at no point in there, did I say the word software? At no point in the world did I say AI. But the reason I jumped to this is, I think the new paradigm for software is it should be as helpful to an individual trying to do a job, at least in the assistant model, we can get to digital workers later, as you hiring somebody really super smart, who knows everything that's going on your company that is the user interface. And that is exciting as hell, because if you're running a company today, think about if you're running a software company today, think about how much friction you have in your awareness. Demand. Jet selling onboarding, supporting R and D, building screens, building layouts, doing usability testing. Go buying Pendo because no one knows how to use it if your software actually worked as well as an assistant, if they could afford to hire so much of that goes away. And so for me, I was like, Oh, my God, maybe I'm too old to start a company. But like, I'm like, Ooh, wait. So for me, that's the big epiphany, which is that users are going to gravitate to an agentic interface. They just are one and two. They're going to gravitate towards one because the second a functional user, like a salesperson or SDR or marketer, et cetera, has their buddy that they like that helps them that they kind of train. They're going to ask them, Hey buddy, can you go do this? Sean Lane 24:52 Brett believes that this new era, the AI agent era, will be defined by an agentic interface where users. Have a brilliant, perfectly trained assistant to help them no matter what their role is. And he believes that instead of having 10 different assistants with different specialties, you'll inevitably have one. My wheels were turning. So where are we right now in this transition, and where are we heading? Here's Brett I Brett Queener 25:19 gave the example of AppFolio in the three ages of application software. Will Boxley was at Siebel. He worked for me in product at Salesforce, $600 million property management software company. If you look at the product, look, I love the company. They're doing great. It's a doozy. It's CRM and ERP for property managers and like, if you just wanted to do something they need to do every time. Apparently, a property manager, you have to check for inclement weather and figure out when something's going to happen. And then, in the old world, you'd have to go into like the Query Builder now, to use the Query Builder and kind of put bully in to figure out which properties are in the path of this hurricane or tornado. Then hopefully there's a child object that has the list of contacts. Hopefully you can sub query to figure out who the primary points of contacts are. And depending on the quality of your mail merge or messaging system, you might have to export, import into a messaging and if two or three of them, their primary language is in English, you'd have to carve out differently what the English or other thing and go send, take you, like, a half hour or with their product. Today, you can be like, Hey, I think there's a hurricane coming. It is in these zip codes. Can you pull up my list of properties? Here they are, if I trust it. Do you want to see them? No. Can you draft an email in the native language and send it to those people? Done? And then the inverse is, Hey, you want me to do this for you every time, every time there's an inclement weather? Would you like me to send these messages to these people? Or would you like me to prompt you that there's a message and I'm and where you want me to tweak the message? Yes or no? Now, in the old world, what would you have to do? Oh, God, I've done this. The user wouldn't do this. Some admin would be like, seems like they're doing this a lot. Okay? They should prioritize. Go build this complex workflow that goes and does this, etc. If you were the product team trying to figure it out, you'd have to really understand what are the 15 things and workflows you're trying to do. And then you try to do six months to stitch together the screens and the workflow a little customizable. And then usability testing and data. But you'd have to guess, you don't have to do that now. So then what then are 300 people in that folio doing? Are they having to go build all of these screens and all of these layouts and all of this stuff, some of it? Yes, but No. And so if you look at AppFolio, they have a competitor that has three other categories they don't have that normally would take. Normally would take them four years to go build and go figure this stuff out. But like, if we're talking about solving for three or four other workflows, maybe there's some proprietary engines in them. But look, you just add to your repository, to your metadata, you train the agent on your schema. I don't know. Yeah, I can do more. I don't have to build as much and so one, a genetic interface is where we're going. Now, I'm not going to get into like, what's the right? Okay, is it a chat bot? Is this? It could be a bunch of things, fine, but my point is the product should be able to do what a user wants to do, and it should be able to prompt to a user something they should pay attention about, that they should go do that is the new thing. Two, a user is going to ask the first one, that's super helpful, to do more. They're not going to want to use three agents. You know, you ever have friends that like send you some messages on Facebook Messenger, which I still understand is like the most world's horrible messenger. Every time you have to reinstall it. I don't really understand what the hell is going on. And then some people send you stop on WhatsApp or something on text, just send me a text. I'm pretty good on text. It's pretty amazing. If you think about text, if you're married or you have kids, you're like, Hey, honey. When are you home for dinner? Three o'clock. Okay, hey, I want to go golf for an hour. Is that okay? I'll be back from here. Okay, good. Did I log into some interface? Go through a bunch of set of tabs to go? Do that seem to work out pretty well. But you know, when you get those three different ones, like, dude, just send me a text. I don't want to I don't want to do this here this year, this year. Just want to do what I need to do. So my bold bet there is, if you flash forward, then that agentic interfaces for the win. Users are not going to want to use more than one agent. You're going to have massive consolidation in terms of who ends up owning the interface for these assistants, for functional users, I think that's what's going to happen. And so what I think that means is there are a number of people today. It's interesting, right? You'll meet with a company that solves something that Salesforce doesn't do, or workway Doesn't do, and you'll ask them, Hey, so, like, they don't want a second screen. It's interesting, right? We'll ask that, like, when would they go into your application? Oh, we'll be in Slack, et cetera. It's interesting. They've already said, like, there's a second screen question. Was already basically saying, if we could make this invisible from a screen perspective, the user would prefer it, right? We've been saying this for like, the last five to 10 years, like, why would I force the user to go somewhere else? And so I think there are a number of applications that want to be customer facing that are that get moved down to the platform layer, and I think they're subservient to what I call a boss assistant. I think salespeople have one boss assistant. There are a number of apps today that I think move down the platform stack, and as long as they do something super meaningful, and then it'll integrate to the boss assistant and allow the boss assistant to do something they can't do. A day. But if that's the case, you really have to have some really interesting metadata. Sean Lane 30:04 Can you give an example of that? Because you make this point multiple times, right, about the importance of relevant metadata, and if you don't have that, like, good luck, right? You should go try to get acquired by somebody today and go home. Can you give me give folks an example of what you kind of mean by that, and you know, maybe a way they might be able to connect it to a company they know, or an example in their workflow that they know. Let's Brett Queener 30:28 look at the go to market tech stack. Is a bloodbath right now, right? I got outreach and sales loft doing SDR stuff, and there's some lavender, and there's some Apollo in there, and then you got reps kind of using that. And then I got Gong doing messaging and calling, and it's got its little AI agent trying to tell you kind of what to do. And then you got Clary calling itself a revenue execution platform, amazing product marketing for a product that, if you just use Salesforce correctly, you probably could have forecast it correctly, but God forbid. Kudos to those guys. Like, I know it's your favorite, no, but it's just like, well done, and they're gonna be a revenue execution platform. And then you got Salesforce that's now throwing its thing out. They released their agents this week, etc, etc. And then within that category, there are hundreds of, like, two to 10 million companies that are people are using for this use case. Oh, I use this for Oh. And then I've got sales enablement high spot, and those guys trying to tell me what to do in a rep. And then I've got, oh, I might use some like, sales room technology, because I've convinced the customer to agree to be in my thing. So we can agree on next steps, and I can see what they're doing. Oh, and then I've got this sales room demo thing where I've created it and I could share and collaborate. I've created, like, seven or eight different coal faces, and they're all trying to have agents. The funny thing, it's the same friggin rap, right? It's the same person trying to buy. They haven't changed. Why the hell would you have the rep, like, use six or seven things and force the customer through six or seven things. There's going to be one boss agent, and I'm sitting here by popcorn, wondering who it's going to be, right? Salesforce is going to fight to not be commoditized down to a database layer on top of a database layer. And the others are going to fight to try to be the engagement layer on top of Salesforce. But nobody's going to use for engagement in labor bots. They're gonna have one for the rep, one for the manager, and all of those small companies that raised venture that word, I call feature, not a companies raising 100x Arr, you're not big enough. You're never gonna be a boss agent. You never have enough spread. You don't solve enough problem that you're gonna be that boss assistant. So either you, I didn't get to metadata yet, you either, if you don't have enough metadata, you're not going to be a boss assistant, meaning you don't solve enough of the problem set that whoever solves the broad enough of the problem set for a functional user, if they get the gig interface right, I think wins. If you aren't that, then what you can be if what you solve has enough metadata around, it's not easy for that boss assistant to just go add that, and it takes time, etc, etc, that you could say, oh, you use this assistant. You want the system to answer this question. We integrate it in, and now you can use that assistant. You don't even worry about interface in and out. Can Do This fine, right? You are subservient, but if you don't have but if the problem you're solving doesn't really have enough metadata, then either the boss the system could build it, or people can build it internally, right? Like I told you about my bot. So we I publish a lot of things, not because I like to, like, see my I don't mind writing. It's kind of fun to write, but I don't write because, like, I want a 1000s of followers and somehow, like, I got cred. It's not like, it's that's the path to super deal flow. It doesn't hurt. But I do it because when I talk to a founder and I say the same thing three times, I'm like, I should write this down and share it. So I'll publish these things. I'll go to our CMS on our website. I'll try to categorize it in fuel categories to go look at it, and then on top of it, we'll run workshops with experts in the field. Maybe you on go to market, maybe somebody on how to do executive hiring right. Maybe how to think, how to set up your ESOP correctly. And I'll get this expert, I'll get on the call, and it's great stuff. And like, one founder shows up, and it's friggin embarrassing. It's embarrassing for the person I bring on, but it's really ridiculous, because if these guys, I because these guys actually knew so I was going to give up on I was like, forget this stuff. It's worthless. Like, no, no. So in like four hours, no, in one hour without writing code. I mean, I could have written some code. I use chat base. It's 100 bucks a month. I connected it to all my content, and I had to make sure be very clear, your job is to help founders and et cetera, all of our content strip the recordings, et cetera, et cetera. And also I'll find content from other venture capitalists, for people who I think is really useful, and I love founders to know it should be in the knowledge corpus. I'm like, Wow, that's great. But then I would have to create a section called content we love. It's not content we publish. You. You have to go find it. You're never gonna go find it. And then I train the bot on it in like 30 minutes. You could ask it a question now, where I had to spend 10 to 15 hours on it, was training that bot like, hey, dumbass, when somebody asks a question and you actually have a link to the document, include the leak of the document, or hey, when somebody wants to know who knows this, and we have the contact information, this is all inside family, give the information, right? That's live. And then they write to slack. And I said, founders, if you have questions, go ask bonfire. And the reason I like it in Slack is everybody can see other questions I've been asked, and you can search for answers. That's it done. And then I go to all of my founders, like, Oh, what about AI? And I don't write for AI. I'm like, Really, you sell in this vertical, and you say, really hard problem for you is that you sell in this industry, and you can hire a good rep, but it takes them six months to know about the industry. I'm like, that's not a scalable problem. Have you written any of this shit down? Well, okay, what if you read the stuff down and, like, you just put it in thing I bought, and somebody just asked any question that gives it the answer, and I'm not even getting like, what you can also do is is have a personal model that understands that personal history and that personal history and that personal individual, and I'm not even including an empathy model understands how this like person talking. You could add that and they roll it out, and they're like, Oh, holy shit, because in that problem I just did, there's no metadata, right? There's no, like, defined objects in a process for you to have metadata, right? So that's where, how I think about it, if that kind of makes sense. So Sean Lane 36:19 Metadata is data that describes other data. It's usually standardized and structured, and it can help with organizing or managing data that you have. Let's take Gong, for example, if you've ever used gong or Gong Salesforce integration, you're familiar with their call duration data or talk to listen ratio, call sentiment, keywords, these are all examples of metadata that are unique to Gong's offering Brett's counter example of training an internal bot on all of his writings and writings that he finds helpful. Yeah, this bot is really helpful and timely, but it's not defensible, really. I did the same thing with our show. We've done over 130 episodes, and as much as I like to think people would want to listen to every single one, a world where someone can converse with the content from this show, instead of listening to every episode, is likely more useful to most people, so I trained a bot on every transcript in the show's history. Brett's broader point, though, is that no matter the example, once you've accessed expertise or knowledge or product information in this way, you're never going to go back ever. So what do we as operators do with this new knowledge of such a seismic shift? Because I can guarantee you that if you're listening to this right now, you do have people who are manually updating stale pieces of content in high spot or your library, and you do have folks that are taking seven clicks to send an email an outreach or sales loft. And worse, you've got reps that are just fumbling in the dark, looking for answers that they don't know in order to make them better sellers. So for us as operators, where do we place our bets? How do we plan ahead for the future that Brett tells us is coming? Brett Queener 38:05 We don't want to repeat the same problem replacing 17 pieces of software with 18 box they don't talk together. It takes us three months to figure out why, like, reps aren't prospecting like, oh, turns out we were stupid. I think the first thing I would say is we kind of have to forget everything we've learned a little bit about the process that we're experts on a little bit. It's like you got to go outside and go blue and just do a first principles perspective. So before picking the bot, because I don't think the one thing on these bots, et cetera, will be interesting thing around if you move from one bot to the other, what's exportable and what's not? The most interesting that's exportable is any training that you do, et cetera, is on rev ops and sales. I would really just go look. And then my next piece is starting with sales. About how do you think about rethinking, leveraging AI within your sales organization? Right? So the first thing that used to require a lot of really interesting intelligence was, which accounts should I go after which people? What's the hierarchy? What account should people pick? Etc, whether it's clay or the other stuff. Like, you don't need to be some super mad genius to figure that out. Like, I get on calls, you're like, well, the SDR and the rep picked the wrong accounts to go after. I was like, What the What? What? Them Sean Lane 39:23 filtering and LinkedIn Sales nav did not lead us to success in Brett Queener 39:26 2003 I uploaded the entire DBN database into Salesforce, and I remember having to get security to do that. And I was like, I don't know it's CRM. Here are the industries. How many employees do they have? How many people are customer facing roles? Here's how many are they? Early Adopters or not, within these industries. And I came up in aptitude to spend, and I created a set of territories so each rep had an equal opportunity to succeed. People have forgotten territories. They're like, ah, Round Robin and blah, blah, blah, reps pick. I'm like, Hmm, I want reps to sell. So first year, I would very quickly look at there are a set of. Common tools to go out there and look at and basically understand, what accounts do we want to go after? Who are the contacts? How do we score them so that we have the right set of accounts per rep that we're going after, and how we choose to engage them, whether, if we're going to go and outbound with a rep, etc, etc, like that, shouldn't require a lot of intelligence. I think the second thing to really go look at is what is the right inbound experience talking to Craig at qualifying some of the others, this whole like MQL thing was this agreement between like Marketo users and Salesforce users that we shouldn't have 20 trail lax bros waste their valuable time working from home with their mouse jigglers. I'm kidding. Many of them still work in an office like waste their time calling bad leads, really. And so this notion of there's a thing and there's a signal, and then, therefore somebody follows up, and it's a human that follows up, and then they're doing some basic criteria, question asking, but they can't really sell, and then it goes to a rep, and God forbid, if your product is demoable, some SE has to show up, like, it's like seven people. It's like an IBM sales meeting, like nine people. And then you have to hire an outside coordinator, doesn't work at IBM, to make sure the meeting goes well. So I would really dig into that, because there is something very interesting, which is that if somebody comes to your website, and if you can have an AI assistant or agent engage with them, and it knows enough about your process and your product and the questions, etc, and it comes across as actually helpful, the hard thing to overcome for consumers on that is we're so used to the first generation of bots or chat For the last 10 years of basically call deflection, and it's stupid and it's just some like, Did it meet the call tree? And, if not decision tree? Yeah, right. And so you just like, I want to talk to somebody. But the reality is, if that interaction could be better than talking to an SDR, because it should be able to be really looking at that and then understanding it's not just this is a leader. This is not a lead, because, based on that conversation, hey, looks like you're looking here something be a little early for selling. Should I, can I sign you up for this newsletter? The rest we can keep you the loop. And can we, can I schedule a follow up call 60 days? You can actually carve more than just ano diode in the past. Like, is this a lead accepted? Oh, it's not. We rejected it. Like, who's nurturing it? Oh, I don't know. Okay, well, that was stupid. I'd really look at the inbound side on the outbound SDR side. So much of an outbound SDRs job is busy work, and I think there's a lot of automation that now occurs such that they should be able to, sort of, like, move up a level in terms of, like, their engagement, their ability to have conversations, almost like Junior AES, like, if I look at the dialers and the rest of it, a lot of that's automated, who to contact and the rest of it, etc. And there I'd focus more on the quality of the messaging and the outreach and what it actually says and how we tweak it, as opposed to, like, what activities you're doing, sales engineering. Go look at that. Really. Go look at at what point do we allow people to really get and understand what our product is, and how do we go about highlighting to them if we've done the right discovery, where our product shines in that case? And how do you do that today? What does it take? Because one, oh, by the way, ses don't need to do RFPs anymore. That's automated. They don't need to do security reviews, plenty of tools that automate that stuff. They don't even have to configure custom demo or POC environments. They could use test box that creates synthetic data and creates that thing. Wow. So now, what's the sales engineer? If I don't make them do that grunt work, right? They're pretty good solution experts. So how do we think of what our solution engineer does? Can they do more? Et cetera? Really, I'd have them kind of go look at that. And then the final part is, what can we expect of a rep? I think we could expect. So what that tells you, to some extent. And then the other thing is, for managers, what's a manager of a rep? How's a manager, managing rep, performance, sales manager in this world of multiple screens, etc, software companies, friggin horrible. What do I gotta do? I gotta hire people. I gotta fire people. I gotta use Clary calls this Gong, etc, and figure out where my deals are. I better make my deals. I better make my number. I'm supposed to scale make my team better and call enable and figure out where they struggle. And I'm not supposed to be super rep. But damn it, if I don't bring those deals home, I'm going to look like a shithead, oh, and then, oh, by the way, I got to make sure I don't enter the next quarter, next quarter bone dry. So I better make sure I provide same opportunity of working early a funnel to make sure that when I get to next quarter I actually have the mid stage funnel to go close right? That job is possible. There are plenty of AI platforms out there. You tell it goals and anomalies, it will tell you, hey, across your teams, across these goals, here's where you struggle. It will actually give it to your rep. Hey, here's where you struggle. Oh, when we meet on a one on one, let's talk about this. Oh, and if you struggle here, here's a link to some sales enablement or content or. Training, right? And then we should rethink through sales enablement. Sales Enablement, by default, is sort of one size fits all. We have a small team of sales enablement. They aren't content experts. We help product marketing right stuff, and we do a lot of it for boot camp, but nobody has started working, and there's no context, and we don't know where this human actually is going to struggle. But like, what should sales enablement be? If I'm a Sales name, a person I'd want to know like, hey in our org, what's our biggest struggle? Is it creating pipeline? Is it moving from top to middle? We're really bad at doing discovery and creating some level of economic urgency that somebody wants to fund a project. Is the issue that we're really bad at beating competition and bottom of funnel? And is that true across our org? And do some org struggle worse is somebody terrible in enterprise deals like not doing multi threading, like, what would a best rep look like this? And as sales enablement, how are you getting that data? These tools exist. You can get the data, Sean Lane 45:52 yeah. And how are you getting it fast and knowing it's right every single time, right? Because otherwise, like people like me are your bottleneck. Yeah, Brett Queener 45:59 Pete and atrium and team. Can go do that for the managers. I could do that tomorrow, right tomorrow, but you have to rethink through how you should do your job, right? You can't be like, what tools you should be like, if I was really doing sales enablement perfectly, what would that be? What you're trying to say is, I have a perfect rep. Okay, what is the perfect rep do in each of these situations? Okay, then how do I understand where those biggest deviations are in aggregate? For one size fits all, one to many content, and then where is it most common? Such that I could do individual trainings and I could point individual to get better. So then, God forbid, as a manager, I know somebody struggling here, I could point out to them, you need to get better. Here. I can point them to content get better. And if they don't, then maybe I can do a pip. But today, somebody's struggling you. I gotta put you on performance. You're not good at this. You can't diagnose why, if you could diagnose why, you can't expect a manager to be the expert on each of those. And you don't actually make the rep better. And the reality is, a pip is not a one time thing. A PIP is as long as a rep works there, you're always improving performance, but like, so you have to kind of rethink this. But the cool thing about this is we can expect a lot more from reps. Like, What excuses do reps have? The other key thing to think through in these roles, as you think about these tools, is, well, what am I hiring? Sean Lane 47:15 Yeah, because we're going to challenge all the ratios. You're going to have unlimited leverage right, to put this in front of people. And I think the best byproduct of everything you just said is it keeps you really, really close to all those end users. You have no choice, right, because the thing that they need and run into and are tripped up by is the next thing to train the agent to do. Right? Brett Queener 47:37 Why do you struggle? Why do you struggle? Why do you struggle? And so you have to rethink. So what I tell companies, it's interesting. We used to be like, employee 15 or 20. I'd be like, hire a rev ops. Person. Like, Nah, that's just because you're a rev ops. You're biased. Like, all right, good luck. And then I'm now telling founders, like, by like, employee 10, I'm like, go hire somebody I don't know. 2520, I don't want them having worked in CRM for too long, because they'll be like, I do it the chili Piper way, or I'm six cents sense a, okay, fuck off and have them think from a first principles perspective, like, Well, how should we do this? And if they can't write code, does some of the kids write some Python because they did R in school or business school? Have them go look at these declarative tools, and they spend all their time looking at it, and are constantly thinking through how to automate that. And I almost call it the Chief of Staff. I'm calling it the founder associate. It's a really smart kid works for the founder, because the founders like, yeah. And when it happens, right? I talk to founders that are like, Oh my god, third time founder. Like, you can't imagine what I could just do and Claude myself, not even a tool, just him building some little projects in Claude, where, in the past, he would have hired 10 people across these functions in sales, and he still wouldn't have known why things weren't working. And he's like, You know what? I can do this, and I could get these answers, and I can now, which I didn't do in first two companies. When I hire somebody, I kind of know what I'm hiring for, and I have the beginning of a recipe, and he says it's life changing, and the capital that I've raised will last me longer. So I'm still writing on that for me to be the expert on go do this, and it's these tools. I'm not going to be Hawking tools, but I think it's just an opportunity for us all to go, like, I'll give you an example. I go to a foundry. He's like, Hey, look at my agentic assistant in my product that helps the user. I'm like, All right, what if something's not working on the product and they want to, like, log a bug? Oh, they go over to this tab over here, and there's an intercom agent. What are you talking about? You have an agent who's answering their question. Why? It's just a different question? Why wouldn't it be the same agent? They're like, Oh. And the reaction is like, Oh, my God, I'm a dumbass. No, you're not, because you were doing it the way we used to think about it. But like, why would we do that? You can do that. That's not hard. And Sean Lane 49:52 I think founder associates, first 10 employees you would have normally had. Anyways, those really smart ops people, like, all those people, are going to. To do now is just inject their creativity and how to solve these problems. Brett Queener 50:03 Yeah, like when I was rev ops at Salesforce, the first rev op, I had to create the category to like, justify why I was getting sued by Tom Siebel, right? It was just two short, angry people. But anyways, what did I know about rev ops? I ran a factory before I was kind of smart, that complicated. We're trying to, like, talk to people, get them interested about us, move them down the funnel. Seems like a process. Okay? What would we measure Okay, in the process? What we define in terms of what the person would need to confirm to move to this process, and then what would we try to do to move them to that? And today, every time you look at a customer and they have this pipeline and the rest of it, I'm like, Hey, what's your sales process? Like, what do you when something is to stage four? What does the buyer have had to confirm in your stage it says giving proposal like you're doing this. What did they have to confirm before it moved there and they look at you like you're an alien? I was like, yeah. Like, it's not really a selling process. There's a buying process that you need to sell effectively through. And they go, oh, and so, like, is a reason why Blackstone used to only hire stem kids, but did you see they're all hiring liberal arts kids? I think the big thing about this is it can seem kind of scary, and it is a little bit, but it's really, actually exciting, because as a founder or running a company, you can actually build a product that an end user is, like, hot diggity dog. Can I slap your hand if you give me that mascot, I'm not giving it to my dog. I'll turn to the mascot sassy, and I'll give them a high five, right? And then how you have to run your company? There's one line I had which is like, what is customer onboarding? If your product works in English the way somebody would want it to, what is onboarding like? What is that? Think about that like? How much better is that? So there's a lot of things that are painful that you have to read these blogs and listen to these podcasts for the OGS that have done this, because we've been through that pain. You're like, oh, that's going to be painful. Oh, that's going to be painful, right? Oh, you're selling to a large org, and you haven't set up the right metadata in terms of org structures and sub departments and the rest of it. Oh, you don't have your admin screen exposed for customizable you know, like all of this stuff goes away. It's kind of freeing. You music. Sean Lane 52:24 Thank you so much to Brett for joining us on this week's episode of operations. There's just so much to unpack there. If you couldn't tell we didn't do the lightning round this week, because we packed as much as we could into the time that we had with Brett. And if you want to continue to learn more about these topics, in particular, what Brett has written on these I would really encourage you to check out his substack. You can find [email protected] if you learned something today, make sure you're subscribed to our show so you get a new episode in your feed every other Friday. Also, if you learned something from Brett or from any of our guests, please leave us a review on Apple podcast or wherever you get your podcasts, six star reviews only. All right, that's going to do it for me. Thanks so much for listening. We'll see you next time Unknown Speaker 53:09 you.