10 Ways AI Is Changing Sales. For Real: Live Workshop Wednesday

SaaSter Workshop Wednesday. Very excited to have everybody. Um, we are gonna have some fun talking about— I'm gonna give my learnings, my thoughts. Um, there's a lot of things like I'm 100% certain of, like what UiPath made in earnings last quarter, or the relative ratio of revenue for Sales and Marketing HubSpot. This is something that I think I know a lot about now. Um, where AI is going to change sales. Um, And I see a lot of stuff on X and LinkedIn and, and folks that are skeptics or don't believe in things, or maybe honestly the truth is maybe their views are a little dated because a lot is changing AI for sales. And I thought I would share what I've learned across 32,000 conversations with our AI, across investments I've made. Gorgeous has 20,000 customers using their AI agents in e-commerce, a bunch of others. I want to change how I think it's going to change sales. Sales. And, um, so I'll share 10 things that I think are going to happen and like 4 or 5 things that— well, 10 things I know are going to happen and 4 or 5 things that I think are going to happen to change in sales. I'm not sure, but I think so. 10 plus 4, and then we'll definitely take questions if there's time. But we're going to do a deep dive on this and so much more. I hope everybody can come to SaaStr Annual this year. It's May 13th to the 15th in the warm, sunny Bay Area on our 40-acre outdoor-indoor campus. It's going to be super fun. And a ton of this content is gonna be, about half of it will be about AI overall, and about half of it will be GTM and go-to-market and sales and where it's going in the future. And we'll have the nexus. We'll have folks like the CRO of Anthropic talking about both, the head of marketing from OpenAI talking about both, but we're all gonna talk about how the world is changing in B2B. So please come if you can. It, it's gonna be kind of, kind of awesome. Um, but with that, let's dig in and Sorry, Amelia, can you mute whoever that is?

Yeah, give me one second.

Yeah. Sorry if we can find it. Yep. Okay. So before we get there, just a couple of things I want to share. There's a few data sources that's going to come from these learnings. One is our own AI. And sorry, just going back. Sorry, a little out of order. We launched our AI about 4 or 5 weeks ago. I'll show you. It's already done over 30,000 chats and we've trained it on 17.785 5 million words of SaaS content. So I've learned a lot about— personally, I've learned a lot about training. And this has 12— not only it has 12 years of SaaS content, it has 1,000 podcasts and speaker sessions. So it knows what Eric Wan thought about competing with WebEx back in the day. It knows what Brian Halligan and Dharmesh from HubSpot thought at a million in revenue and $2 million and more. It's got all of them. It's got all of our learnings, all of our feeds. And, certainly one thing I have learned, if you're a skeptic of AI in general, if you're a skeptic of AI because of hallucinations in sales or marketing, it may be because you haven't seen an AI that's well trained. And it is work to train an AI. Again, at Gorgias that does, AI support for 20,000 SMBs, for the bigger merchants, and it's mostly SMB, for the bigger merchants, it can take a full month to fully get a merchant trained on their AI to know every single product, how to answer all the questions, all the issues. If you think an AI out of the box is going to do magic in 60 seconds, I think this is where a lot of folks are, are disillusioned or don't believe because they try AIs or they go into ChatGPT or Claude and it doesn't work or it hallucinates. I think it's because you haven't trained it. Ours is pretty good. And again, just go to saster.com and click on our little bubble at the bottom or do our AI and ask it any question on B2B. I think you're going to be pretty impressed. But there's a bunch of reasons. One of it is where technology really is today. It's changing so quickly, folks. And two, again, a lot of training. Okay. A lot of training is my life. It's not that it takes forever, but you really, you do have to invest. So again, this is our AI here on the left and we've done 33,000 chats already. So I think we've learned a lot from that and from other folks I have to do. And so let me tell you where I think it's going. And a lot of folks will think some of these things are nutty or aren't going to happen or sound crazy. But I'm going to tell you, these all are going to happen and they're going to happen this year. Or very soon. Okay. The first one, and I know some folks, a few folks that are deep on— there's a lot of like, uh, AI note-takers here. There's Otter here, which has a little bit of this capability already, a little bit. But what's gonna happen within 12 to 18 months, and this is probably the biggest change in sales, is your AI will join you on every sales call. Every sales call. I'm not talking about a note-taker. I'm not talking about some random little Slack thing. I'm talking about— look at this. This is me and Brian Halligan, chairman and co-founder of HubSpot. This is his digital AI and me together talking. And the product we use, Delphi, up here, if you look, you can see it already enables an integration where my AI can come to all of our Zoom meetings. Now, this product Delphi does a lot. I love it. This feature is a little— isn't quite ready for primetime yet. Okay. But here's the thing about AI. It's close. Look at the bottom. Brian Halligan and I can already have a deep conversation. And in a couple of months, this digital Brian or digital Jason will join our meetings. It will join our meetings. And it's more profound than that because digital Jason, again, is trained on 12 years of SaaS content. And like, take a sales use. We have sold $110 million of sponsorship to SaaS events. Our AI has ingested our prospectus word for word, every sales conversation, every meeting it has. And so I might not know, hey, you know, let's imagine we're doing a sales call with it with Google Cloud and they want to know, can I have churros at the SaaStr annual? I mean, I don't know if they can have— Amelia knows, but I don't know. I don't think our sales team knows. I'll just ask Digital Jason. Hey, Digital Jason, is it— can Google bring churros? And they will say, yes, you need this type of permit. It has to fit in the size. You have to have the approved vendors. It knows the product cold and it's almost there. It is almost there. I know some folks will say it's already here today. Mostly vendors, mostly vendors trying to sell you something will say it's already here today. And others will say this is impossible. What I'm going to tell you is just a couple of months out. It's just a couple of months and this will utterly change. I don't know about you guys, but I can't tell you how many sales calls I have been on in the last 24 months where the, where the, where the person knows nothing about the product. Literally nothing. Does this integration work? How could I do this? How much? I don't know. I'll get back to you. There will be no more I get back to yous. There will be no more stupid sales calls. Your AI will come and they will know. And again, this is gonna be a theme. It will know the product cold. And this will utterly transform the way we do sales because we will go from 10% of sales calls being great to 95% of them being great. At least the call, because the AI will know it. This is coming very, very fast. We will all have literally a digital person on it with us. And, and it will work in full duplex. And we'll just say, Amelia, does that work? And Amelia will tell you it won't be a fake. Amelia won't look— maybe it'll look like a robot sometimes, but I think it's going to look just like Brian Halligan here is. And you won't know or care if it's a real person because it will be great because it will know the product cold. Okay. Related point. And we just kind of talked about this, and some folks know this that have deployed AI across their They're like sales and websites. Other folks are skeptical. You shouldn't be skeptical. AI already can be the best BDR in the world. BDR, not SDR. SDR is a different topic. How to automate outbound is complicated. And I think many of us actually, um, are too jaded on AI and sales because we tried some AI SDR product 6 months ago that was just spam on steroids, right? We just— and it didn't work. It didn't work, right?

Wow.

Well, we sent 10 million emails. That were barely personalized and we didn't magically get $10 million AR. So AI doesn't work. It's the wrong way to think about it. First of all, I'm not talking about that. Secondly, everything's getting better. But third, I'm talking about a BDR. I'm talking about somebody to qualify a lead. And first of all, no one wants to be qualified. No one wastes their time. No one thinks it's fun to talk to sales. I mean, a few people do, like, and so you're routing leads to a BDR that takes between 10 minutes and 3 days to get back to a prospect and knows nothing about the product. Typically their only job is to set up a meeting with an AE And then I gotta wait for another meeting to be set up. It's a terrible experience. Our AI already answers, has already answered over 100 sponsor questions that, and over 20 sponsor inbound requests. And here's one. What are the specs on a Super Bowl booth? I mean, I don't know about a Super Bowl. I shouldn't have to know this stuff. This is a better answer for a prospect than a human can give. Here's the size of it. Here's how speaking works in the sponsor stage. Here's how activations work. Here's how a demo works. Here's how many passes you get for your team. Here's how lead scanning works. Again, it knows the product cold. So AI already can be better than all but the best BDRs. And the AI works 24 hours, and the AI is polite, and the AI is correct. The AI is correct. So I really don't— I see that the classic BDR to qualify, to score and screen a lead, I see this becoming borderline extinct over the rest of the year. You just don't need them. Are there exceptions in everything in the enterprise? Yes. The more enterprise we are, the more human and personal it's going to be. But we just don't need 20-year-old kids that don't know anything qualifying out prospects anymore. AI is so, so much better.

Okay.

This one is going to be controversial to some. And this is something that many VCs think actually has already happened and it hasn't. Like there's the, there's the AI, there's the AI super fans that don't work in the real world. And then there's the Luddites. But I'm pretty darn convinced based on using our AI and watching what's happening in support that within 12 to 18 months, AI will handle 80% of one-call closes. One-call closes. I'm not talking about putting on a tie and a suit and going to meet a 7-figure customer in their office. I'm not saying AI is going to replace that at all. But if you've worked in high-volume SMB environments, if you've worked in $2K, $3K, $4K deals, even smaller, you will know that honestly, even, even pre-AI sales and support kind of merge a little bit. A lot of times folks will try a product on their own. They'll, they'll, they'll use a free edition and they'll have 1 or 2 questions and then they're ready to buy. Okay. And we call that sales and it is sales and you need sales process and you need techniques and you need objection handling. But AI is gonna do better than this. And it look, here's our AI. Now our AI is like 1 step, probably 1 release for our product away. Here's, here's me. Can I buy a sponsorship directly? No, you can't buy one. But let me know if you want me to connect you with someone specific. I think in 3 months this will just send them the contract. Now, if you need to negotiate it, if it's a 6-figure deal, you need a human. Don't get me wrong, our, our, and our little tiny sales team of 3.5 at Saastr is great. They will help you for a big deal. But to just buy a gold or silver booth, you're not going to need a human. We're almost there in our AI already. It can answer all your questions. It knows the answer. It will— it won't oversell you. It will tailor it to the size of your company. And instead of connecting with someone, it's just going to close the deal when it's $2K, $3K, $5K, $10K. You're not going to need humans. And again, I don't know how many times I can say this. This is the most profound thing that I've learned from our AI, that Amelia, that our team has. It knows the product cold. If you train a great AI for sales or support, no human being, even Amelia, who's the best on our team, no human being will ever know the product as well as the AI. They'll know every integration, every spec, every objection. They will know every similar, you know, hi, I'm Drada. Is there anyone like me coming to master annual. Yes. You know, you can say Vanta has been there for 5 years. The rep might not know that. The rep will know every similar customer, every competitor, everything that knows it cold, and it doesn't judge you. It doesn't judge you. Um, okay, sorry, I guess I thought— I guess I skipped 4. 4 wasn't that important. Okay, 5. 5. Um, this is one that a lot of folks challenge a lot. I see this all the time on LinkedIn. AI is not going to work because people want to talk to people. AI is not going to really work in sales because people are going to want to talk to people. Look, let's be clear. If you have the world's best sales rep available 24 hours a day that is there for you, I would rather talk to the single best sales rep at any company than an AI, the single best one. If for no other reason, I can get a sense of the product. They'll probably tell me a few things they shouldn't. I'll get some war stories. Of course, I want to talk to a human. But I have to tell you, across 32,000 chats, people love to chat to a great AI more than a mid-pack or mediocre human. They love it. Here's a few. You're awesome. Thanks for doing what you do to our AI. Again, thank you. Great help. Thank you. People love it. They love to talk to a great AI. And if you're skeptical that people want to talk to people, honestly, here's what I would say. Go, go find 5 SaaS products today and go inbound. And see how terrible the experience is today. See how many days it takes to set up a meeting and then ask them a, a, a slightly complicated question that you could figure out from their website. Okay. How does your Zoom integration work? They're gonna have no idea, but it's right on the website. You know who will know? The AI. The AI will know. And people already across our 32,000 conversations, they prefer to talk to the AI. So if you're skeptical, if you think people wanna talk to people, I'm not saying you're wrong, but I am saying it may only end up being true like 10% of the time. And what you see in the, in the support space, Gorgias, again, 20,000 coming up on $100 million in revenue, 20,000 e-commerce customers with AI. What they found rolling out all these AIs is the AI is never the number one person in support, right? These, these are, these are, these are merchants with typically 5 to 500 folks doing human support. And they found that it's never number one. It's never the number one agent that the AI is, but it's always in the top 10%. It's always in the top 10% of CSAT. Think about that. The AI isn't perfect. It does escalate things to humans when it needs to. AI is already really good at escalation, but it's always in the top 10% because it knows the product cold. This is the thing. If you don't get AI for sales, this is the part you really need to know, that it will know the product cold better than 95% of humans. And so where do I think this will go for routine sales? For routine sales. And again, using Gorgeous as an example, this is a fun slide. Because support is just earlier to this than sales for a variety of reasons. And here's Gorgeous on the left, and here's a bunch of other competitors or adjacent folks blacked out. They did a survey of hundreds of folks. There's all these other folks in support. I'm not going to tell you which of these bars they are. Some are Zendesk, some are others. It doesn't really matter. The point is the, the, the bar on the left, I think it's pink or something like that. I don't know, salmon. I'm colorblind. They all claim that they resolve— AI can resolve 60% of customer issues. And the other chart is how much it is in practice, between 25% and 50%. I think this is where it's going to be in sales. We're all— a lot of products and folks are going to say we can close 60%, 80% of deals with AI, and it'll range from 25% to 50%. But if your AI is great and people love to talk to it and it's a routine deal, um, people are going to prefer to talk to the AI rather than the kid that doesn't know the product and takes 3 days to get back to you. So we're going to see this play out. I think I wasn't sure even a couple months ago that this would happen because support is more routine, support is more issue-driven. Sales is— everyone would say sales is different than support, but now I'm not seeing it. Now that we've run 32,000 chats through AI, I'm not seeing it. I think we're going to see this exact same rough— we're going to see 25 to 45% of sales be resolved overall by AI for routine sales, for routine sales. Um, Okay. This one is super interesting if you guys haven't seen it. If you haven't seen it yet. And I think a lot of folks that aren't, that are in other parts of B2B haven't seen it. But this is one of the most profound things happening in sales and support. And it's already happening in e-commerce, which is where I see it earlier in, in the, in the gorgeous Klaviyo Shopify ecosystem. But it's coming every, which is that a great AI, a great one. And again, you gotta understand great to understand this whole presentation. It's going to start to merge sales, support, and success. It's going to start to merge it. So here's a, here's just a little silly example I did with our AI. But what ticket should I buy for Saastr? This is a support question. Okay. What size should I buy? A team pack? VIP? Big co, right? Great support question. Answers it really, really well. At the end, it makes a recommendation. You're probably a startup guy. And then I wrote at the bottom, okay, I'm ready to buy. Okay. Right now our, our, our, our AI will direct you to the ticket page, which is doing it. But my point is, what's the difference here between support and sales? There's nothing. And what our AI can already do is then remarket to you on its own in a couple days, in a week. You can set up triggers and workflows. So you ask what tickets you buy, it shows you, it shows you to buy. If you're not ready to buy, it will do the marketing for you and the remarketing and the drip marketing all in one AI. And typically, if you look at most products today, most software products, you've actually got 3 to 4 different routes. You've got You've got customer success. You're trying to triage support between large and small customers so that your support team doesn't get burned out. And then you've got contact sales, my least favorite button, but I know we have to have it. Okay. Why do you need all of these? Why do you need big support, small support, contact sales, marketing? It should all just be one. AI is going to know it all. It's going to know it all and handle your issue. And so what we're already seeing this in e-commerce where because e-commerce is so transactional, you'll go to a lot of like I went to buy a Secret Chair. Gaming chair a couple months ago, it had 4 links on it. Click here for support, click here for returns, click here for ChatGPT, click here for sales. Like 4 links. Like they weren't even in like the same font or colors. It was like a mess on the website, right? That will— should be one link. AI will make it one link and it will make it one for you. And the line between sales, support, success, and marketing is gonna be very, very fluid because of AI. Because AI, you— it's not 4 people arguing at your startup who should do it and not 4 people who don't wanna pick up the phone because the deal size isn't large enough and not 4 people pointing at each other whose job it is. Is it, is this customer success or is it support or is it sales or is it post-sales or is it the solution architect's problem? The AI will handle all of it. It will handle all of it and it's magical. This is magical and it's already happening. So exactly where this will happen, we will see. But bear in mind, the walls between these functions is already starting to break down and it's going to be completely changed the way prospects interface with you in the future. The fact that these are not 4 silos— sales, support, success, marketing— huge change coming already sort of here in e-commerce. It's early, but it's gonna come everywhere over the next 12 to 24 months. Hey, point 8, and, and, and some people again get slightly triggered when I say this, okay? AIs over the next 12 years will just simply take over the QBR and the check-in, okay? I know when I say this, a lot of folks in customer success get upset with me. But I've written this up. Literally every single QBR that Amelia and I have had over the last 24 months at Saastr has just been an upsell. We come to the QBR. What problems are you having with this product? Well, we're having these problems, 1, 2, and 3. Oh, someone will get back to you. It's like a waste of time, right? It's just a disguised upsell, usually with someone who doesn't know how to spell Saastr and knows nothing about us. It's a total waste of time. There is no point in having a mediocre QBR or check-in call when an AI can do a great one. Right? Because again, it's— I know some of you will challenge this, but spend time with RAI or GreatAI. You will find you will trust it more than 80% of CS people who are just there to sell you a module you don't even want to buy. Okay? And look at our little AI right, right, right down here. It already has weekly progress check-in. We're going to enable this over the summer and it will just check in with you automatically. And it's not a dumb cadence or a dumb random email. This This AI, it has memory. Like, memory is a trendy thing, but memory is not new. Our AI remembers every conversation it's had with you. If you go into our user AI and then come back a couple days later, it remembers all the conversations. So it will know how to follow up with you on whatever the progress is. If— forget about sales. If you used us, like, to, to decide if you should hire an AE, a lot of folks are uploading resumes, uploading VC decks, uploading board decks. It'll follow up with you in a week. Hey, I know you, I know you interviewed Amelia. You said you were worried it was a tough interview. How did it go? It will know, it will remember. Or if you bought the product, everyone knows that, um, like you wanna do the 30-day check-in, the 60-day, the 90-day check-in, but humans don't want to do it. They don't want to do it, guys. It's the truth. AI don't mind. AI is happy to do a 30-day check-in, the 60. And if you complain and you say, you know what, I'm disappointed that integration with Squarespace didn't really work that, that I was promised, the AI will say, I'm sorry, you're right, it doesn't work. Here's what we can do to solve it. It will be much better. It will— 80% of QBRs and check-ins will be done by AI in the next 12 months. There's no— and it's already here in our AI. It's already here. It doesn't— it's not quite all worked up on our backend connected, but man, it will be soon. It will be soon, like a lot of these points. Okay, point 9, which, um, I think— and, and then this goes into things that I think are going to come and happen. Bear in mind, um, we've been able to do a cheat a lot in SaaS, a cheat. B2B, which is that, um, look, we'd all love to have apps that are fully self-serve and all this, but if your app needs some onboarding, if it needs some hookups, if it needs some integrations, we, we plug the gap with people. We plug the gap with people. So many— almost anyone that's— that, that sells to the enterprise plugs a lot of gap with people. But our AI is going to push us here to do more self-serve because the AI is going to want to get you going on its own. The AI, what the AI can't do on its own is get 4 consultants to deploy your app as humans. AI, it's gonna, I mean, I, it's gonna be a while before AI can, can replace, uh, um, forward deployed engineers and, and, and, and, uh, and humans configuring your software. And so what, when the AI can close deals on its own, when it can do the follow-up on its own, when it can do sales on its own, you're gonna want a true PLG self-deploying app. You're not gonna want an app that requires humans. You're not going to want— you're going to want an app that goes live that second because the support's going to happen, the questions are going to happen, the buying is going to happen, and then the AI is going to want to deploy your customer that then and there, not this month, not when the team, the onboarding team has time. So it's going to push us to build better onboarding, more freemium, more self-serve products, even higher up in the stack, even for higher deal sizes, because this is the way AI is going to work. The AI is going to want to get that customer deployed instantly. so there'll just be more of it. And my last point here, and then I'll come up with some predictions. All this stuff outside of number 4, which I forgot. Okay, so I guess it's really 9. All these, I'm 100% sure are gonna happen over the next 6 to 18 months because it's already happening in our AI and it's already happening in e-commerce and post-support. If you're skeptical of some of these things, and I know many of you will be, you'll be like, there's no way this, this, this like Jason M. Lemkin AI is gonna join the sales call. It's not ready. Well, It's almost there, guys, but don't judge everything that it is today. AI is constantly getting better. For example, RA has only been live 5 weeks. 3 weeks ago, it couldn't review a VC deck. It couldn't take a PDF and fully review it. Now it can. And that was just automatically— the rate of change is so fast in AI that don't judge it where it is today. Judge it where you can see it going. Now, we've already had hundreds of founders reviewed their VC decks and board decks. If you're not sure if you could get funded, go to our AI, upload your VC deck, and ask if it's good enough. It can ingest that deck right now. And here's an example on the right. And I said, we're almost ready for me and Brian Halligan, our AIs, to talk with each other. We try this now at the office. We bring up both our AIs and we have them talk to each other, but they can't quite hear each other right. And me and Brian's digital guys, we talk over a little. Like, it's close. It's close. We try to have these fun conversations where I ask digital Brian and digital Jason what they think of topics. Okay. But And they're kind of like two old codders out on the stoop. They can't quite hear each other. They need a little bit of help. What'd you say, sweetheart? What'd you say? They can sort of hear each other. But then here's the founder of the AI we use, Delphi, intelligence interruptions coming soon. They'll be able to hear each other and interrupt each other. Coming soon. It's all coming soon. So anything that you see out there that sort of, it's kind of working or it's close, instead of being a naysayer and saying it's not gonna work, Invest, dig in. The rate of change is insane. In 5 weeks, our AI does so much more than it did 5 weeks ago. Okay, so just don't judge completely as it is today. Squint or you're going to be left behind. Okay, now a couple of quick thoughts and then we can open up to questions. Things that I think are going to happen in sales. I'm not sure the other ones. I'm sure I know you're not sure. You're skeptics, but I'll bet you dollars to donuts all these, these last 9, 10 points happen. These are things I think are going to happen, but, but, but I'm not sure. But but, but people may be missing this. Okay. The first one is I'm pretty sure when the AI closes deals on its own, you have to have more pricing transparency. And if you look at so many of the folks that have broken out early and, and we can learn a lot from DevTools and support because those are the areas that are the biggest, the biggest beneficiaries of B2B. I mean, Windsurf, Cursor, my God, you know, Lovable Bolt. These are all businesses that have rocketed to hundreds of millions of dollars with DevTools in a couple of months. But it has also utterly changed Zendesk and Gorgeous and Intercom and Talkdesk and Genesis and all those folks too, really early. One thing we can learn is that transparency is going to really help. If the AI is going to close a deal on its own, there's only so many pricing games it can play. It can play a few. It can tell you what list price is. But for example, on our AI right now, like if you want to come to SaaS Annual, I would say 50% of the folks that I've seen do that ask about SaaS, they ask for a discount. Okay. And the AI just gives it to you. The best discount we have today is Jason20. If you want to use it and you're coming to SaaStrato, go to our, go to buy tickets and type in code Jason20. You get 20% off. Okay. And then they come on. Then they ask the AI, come on, there's gotta be a better one out there. And the AI is honest. You know, there were, but we're less than 30 days out. There isn't, there used, you know, 6 months ago there were better codes and they're like, come on, you've gotta be hiding a code. And the AI is like honest, like I'm not hiding a code. It's Jason20. Now we could probably train the AI to hide that JSON 20 code, like make you ask a couple of times before you get it. But ultimately you're gonna have to dial in a certain amount of transparency if you want the AI to close on its own, right? I'm sure there could be games you could try to get 18% off or 16% off, but, but the, the endless games that sales teams like to play on hide the ball, it's just not gonna work as well. You have to give the AI, just like a tight CPQ system has to really tightly lock down discounting. When you get there in bigger and bigger deals, when you have contracting processes for, for configure, price, quote, you're gonna have to do the same with your AI. You're gonna have to tightly lock down discounting rules and just live with them. And I think it's gonna be great. I think it actually— the, the more friction you take out of a sales process, the more that closes. But I think it will accelerate some level of pricing transparency. AI, it's, it's got to, so the AI can do its job. Um, related to this, if the AI is gonna close deals on its own, like you got to be, you got to have customer-centric rollouts. Like if you want the, if you want someone to talk, discuss, sign, buy, and deploy on the run without a human, maybe a 2-year deal is not the right choice. Maybe it's, you're asking too much trust for a 2-year deal. We may need more pilots, we may need more opt-outs, we may need more monthly deals. Like we may see things tilt more toward monthly deals with clear opt-outs rather than screw you and auto-renews, right? I, you know, I ask my AI, hey, I'm going to ask every AI going forward, hey, is there, is there an auto-renew in this contract? Yes, there is. Can you take it out? No, I can't. Like, I might slow it down, right? Maybe the AI adds, adds the clause for you. But I think we're going to have to have more customer-centric terms so that the AI can close the deal on its own. Right. Just especially.

Right.

Take friction out of it. And I think if nothing else, we may just, we may see an acceleration of monthly deals and acceleration of easy opt-outs. I know, I know sales folks hate opt-outs and they hate pilots, but put your customer hat on. Like, if you're buying from some vendor you just discovered on, on, on a ChatGPT search an hour ago, you really want to sign a 3-year deal with no opt-out? Of course you don't. Of course you don't. So I, I, I, I think that to, to, to, to accelerate the AI closing deals on its own, we're going to want somewhat more transparent pricing, and we're going to want more customer-centric terms. And I think it's fine. I mean, Honestly, if you, if you, if you, if you, if you force, if you make it really, really hard for a customer to leave, but they still leave, you know what, maybe it makes it a little bit easier to hit the plan for the month, but it doesn't matter because when they leave, they leave. All that matters in B2B is your terminal state revenue, your ARR in 2 years, in 3 years, in 4 years. It doesn't really matter what your ARR is today. So masking churn doesn't really help. And I think AI will accentuate the fact that masking churn doesn't help. Okay. The third one, I have less ideas here. This is what I'm curious. Um, but someone wrote something about our AI the other day on Twitter, on X, and it, and it was, it's, it was obvious, but it was profound to me. They're like, a great AI fills gaps. A great AI does things you didn't know it could do. For example, um, our AI alone, like we have thousands of speaker submissions to SaaS Annual, to our events, to our, for our podcast. We, we used to have a team of humans process them and they couldn't do it. Our AI does it now. Our AI process reviews their, their presentation, reviews their ideas, reviews the speaker submissions. I had no idea our AI could do this. I mean, I get it now. We didn't even think about it, but Amelia threw in one for now. Everything goes through our AI. Okay. That's an example of a gap. I didn't know our AI could review VC decks really well. Trust me, put your VC deck in. It's honest in a way humans aren't. Um, what one thing I think all AI should do, ours isn't tailored to do it, but yours should be. Is it should automatically follow up with a new case study. Let's say you just closed OpenAI as a customer. Your AI should automatically push that case study to everyone in, in, in AI in the space. It should put— if you, if, if anyone that's in the funnel that would be interested in OpenAI as, as a case study, it should push to them. So what I can tell you is our AI does 20 times more things than I thought it would do when it launched 5 weeks ago. I thought it would just answer questions about SaaS or content. I didn't know it could do ticket sales. I didn't know it could, answer sponsor questions. I didn't know it could qualify sponsors. I didn't know it could review sponsors. I, I didn't know if it could review speaker content. I didn't know it could review submissions. I didn't know it could review VC decks. So I don't know all the things it can do in sales and go-to-market, but if you hook it up properly, it's gonna fill gaps. Not just the gaps when your sales team is sleeping or doesn't wanna work or is out going for a run or is working their side hustle, but it's gonna fill all sorts of gaps for your prospects and customers that you didn't even know you needed filled. It's magical. It's magical for us already. Um, okay. And I think the last point— hold on, I might have 2 more points. Yeah, the last 2 points, then we can take questions remaining time. Here's what I think is going to happen, and I, it's possible I have this backwards, but I already see this, and maybe this is the most existential change today I will have to sales. I think we're no longer going to tolerate the slow and vendor-centric sales processes that we've had since the dawn of SaaS. Okay. Right now I go to a vendor and I click on a link and between 5 minutes and 3 days later, an AE sends me their Calendly. Okay. And listen, I love Calendly. Tope's gonna be there. He's gonna come to Saastr and he's gonna answer all your questions about Calendly in the Day of AI. But this, this is not a customer-centric workflow to take days to get a Calendly link from an AE. Our AI never sleeps and is always kind, is always ready to take the meeting. And I think you're not gonna be able to sell in 2026 the slightly laconic and on— I hate to say this 'cause I'm gonna make people mad— the somewhat lazy way we've sold the last 20 years, 10 years since I've been in SaaS. Once you have a great AI, this is the thing you have to— going back, so many people thank our AI, they love our AI. Once you've had that, and so many people go back 5, 10, 20 times already in the first 5 weeks. Once you've had that wonderful experience, you're not going to want to tolerate a mediocre AE, a mediocre SDR that doesn't know the product. You're not going to tolerate it. And you're going to expect, you're going to expect the human experience and the overall sales experience to be as good as that AI. You're going to expect it to be as good. And so whether that means we're going to have to step up our game, step down our game or whatever, we're not going to tolerate the way that selling has been the last decades in SaaS. I know some folks are going to challenge me on this, but pretty sure once you've had that great hamburger, once you've had that great taco, once you've had the best steak ever, it's really hard to go back to eat at Denny's. It's just really hard. And a lot of sales is like Denny's today. It's a terrible, it's truly a terrible experience. And I'm just going to not tolerate it anymore. We tolerated it in 2021. We bought everything we needed because we were locked down. I don't think we're going to tolerate this once everyone has a great AI. We're not going to tolerate slow, plodding, somewhat ignorant sales process. It's just not going to be tolerated. Okay. And here's my last interesting question. I don't know the answer to this one. Then we will open up to questions. Will human AEs and SDRs get better or worse in the age of AI? Okay. I wrote a post a couple of months ago talking about when AI was just starting, but before we ran our own AI, talking about mech AEs, mech AEs that I thought what AI would really do is make every AE great. And I still believe this. Going to our first point, I believe very, very soon every AE will come to every Zoom with their, with their great AI, with the digital version of themselves. And that will make them only better because they're mech. They will know everything about the product. They will help them. They won't judge them. They will know all the integrations, all the corner cases. They will know if you, if you, if you barely know your prospect, your AI will know it cold. And if you're actually not sure what your prospect does, your AI will know it's that two of its competitors already use your product. It will be magical to have this. And so the MechAE and the MechAE will automate follow-up and the MechAE will automate everything. And so I think it will make them great, but I'm also worried it may make them terrible because if I just wrote this article on SaaStr, this experience we had where I, where I'm seeing is a lot of AI SDR outreach is actually really, really good now. It's really, really good. And the follow-up is terrible. I gave an example here of You know, we, we, we have a podcast, but we really need it to be edited great. It needs to be pushed out. It's like a lot of work to produce our podcast. So I got this outbound from this, uh, I guess this SDR, they wrote me the best email in the world. We can fully automate your, your, your, uh, your podcast. We know all about the SaaS podcast. We know what you need. We know how to improve it. We'll take it all over for you. We'll do all the clips. We'll do the social promotion. We'll upload it to iTunes. We'll do the YouTube. We'll do everything for you. And he is like, here's our other, clients where we do the exact same thing. I'm like, this is great. This is what we need. I'm in. I didn't even ask about price. I didn't care. I just said, I'm in if you could do this today. And then it turned out it was an AI that wrote it. The, the, the sales guy got back to me, he's like, um, what's your podcast called and how many do you do? What's it called and how many do you do? If you guys know anything about podcasts, and I'm sure almost everybody does, just go to iTunes or Spotify or YouTube and just see how many we do. You literally, We've done 799 podcasts. We're coming— 800 will be with Marc Benioff. It's gonna be great. It's gonna be all about AI and the sales, the world's best outreach. And then the AE asked me how many we do. It's an insult, guys. It's an insult. So I made this little image up of this guy below. Sorry, talk to my AI on the Zoom. This might be where it goes. We might get better AEs, and I think we will for sure. I think the best AEs— we've got a, a great new folk that joined our tiny sales team at Saastr. He's already using our, our AI to answer all these questions. And he's already much smarter about SaaS than anyone that ever started before because of AI. He's being a bit of a mech AE. But I got to be honest, I know a lot of folks are going to be like, this guy at the bottom, just talk to— I'm not just like this, this, this podcast outreach. I'm not going to do the work at all. Talk to my AI. I'm just here to collect my commission. I think it's going to cut both ways. I think it's going to make some AEs much better. And I think I'm concerned it's going to make most of them worse. And it may be in the contact center where We look at them and like in, in support, a lot of folks are already laying off 30 to 40% of their support team that isn't better than the AI. I think you may be at risk as an AE if you're not a MEC AE, you may get laid off in the next 24 months. They, especially for transactional stuff, the AI may, it, it may, you may rely on the AI to do your job, but then we may not need you. I don't know. I think it's gonna cut both ways. I'm not sure. I'm waiting to see. This one is an interesting one to see whether it makes salespeople better or worse. And for sure the answer is both, but Okay, those are my thoughts. And I think, Amelia, if we've got any questions, I'll try to stop the share and we can just chat about it to extent folks have questions.

Yeah, you do have some. There's a lot of just discussion also just happening. So let me go back because there was a lot of note-taker notification.

A lot of note-takers.

A lot of note-takers. A few people at the beginning asked and I answered, but I don't know if you want to share on this on like how we built our AI. I know we did a whole separate workshop Wednesday about this. So I pointed people to Delphi.

Yeah.

I don't know if you just want to give your quick thoughts on like, you know, we did train a lot of content.

I'll tell you. So you and I showed that screenshot. This presentation is published and we did one before on it. It's on our YouTube slash Saastr. It's a good question and I learned a lot from it. So it— listen, I love this app. You can try it. A lot of folks have. You do. You do have to ingest it. And what it ingested for us is it does a combination of things. One, it has scraping. So it scraped every single 10,000 Saastr posts off saastra.com. So we put in the domain saastra.com, we put in saastraannual.com. Now it didn't pick necessarily— the scraping doesn't pick up documents. So for example, our first sponsors, for our sponsors, we have a prospectus. At first it couldn't answer good sales questions like it didn't answer them right, but then we just uploaded the PDF of the prospectus and it instantly got like 98% of it right. Okay, so we had to upload some PDFs. So if you have sales collateral, you should upload it. Sometimes it got confused with old versions of the prospectus and new ones, right? For example, it thought SaaS Annual was in September this year in the beginning. It moved to May, but then, then there's just training. So I just wrote in SaaS Annual is May this year. Now it doesn't get it wrong. Okay. It's similarly, it made up like SaaS Europe we've done for years. It's typically been in June. It can't be June this year because SaaS Annual is in May. So it started to make up that it was in June. Because there was nothing. Then I told it it's going to be in December and now it's perfect. So we had to, we had to do a few of those things. So we had to upload some documents, some things where it's confusing because it wasn't on the web. We added, we added this. So probably we add 1 or 2 things a day to the database and then it actually connects with our YouTube automatically. So every single video, we have thousands of YouTubes from everyone, Eric Kwan, Stuart Butterfield, everybody, and we'll have 200 sessions from SAS Annual this year go up. It ingests them all, all the YouTubes, transcribes them and ingests them all. You actually— it's not as hard as it sounds. You can actually just get transcripts out of the YouTube API. It turns out it's pretty simple, but every single thing that's been said at every Saastr Angle since 2015 and every digital event we've done, CEOs of Notion and Webflow, it's all there. And all you have to do is just— we just had to hook it up to youtube.com/saastr, okay? And my YouTube. And then it actually is connected to my Twitter and LinkedIn. Um, and so every single thing I've ever done on LinkedIn, and I have thousands and thousands of LinkedIn posts and Twitter posts, it's all adjusted. So all our social media, our YouTube, uh, our, our blog, all our content, um, and, um, some PDFs like that are like basically sales collateral we put in. And we watched it, and then you, you kind of— and then so that's pretty good. And then you see where the, where the corner cases are, and you upload more material. But it's a couple of hours of work to do that. Um, but if you want to be great, you, you kind of need an AI ops person. Like Emily and I each do a little bit of it. So each day we spend 10 or 15 minutes upgrading it, fixing little things that like the hallucinations are rare. The only, you know, the hallucinations are like we didn't have SaaStr Europa 2025, so it made it up. That was a hallucination. Mostly it's just corner things that gets wrong and we just correct it. And but that's how it works. And it's, it's, and, and it is, you know, what I learned from Gorgeous in the support space is when they first rolled out their AI, they thought it could all be automated, right? That, but, but what they learned is small businesses don't know how to do this. So actually it is work. It takes them a week or two in many cases to get their cust— they have to train them. Okay, where's your pricing list? Where— because Gorgeous, they integrate with Shopify. So anything that's Shopify, they can ingest into their AI. But there's often other sources of data and other things. So they do have to kind of teach them a little bit how to train it. Um, but it's easier than you'd think if you do it. Just, just gotta do it. But, and it's, and the other thing I learned, if you asked, um, you can try Brian Halligan's too. It's on the Delphi website. His is fun. He's the chairman of HubSpot. Brian, I think Brian's a better CEO than I ever was, right? A better executive. I think our AI is better. He thinks his is very good and his is very good, but I test it a lot. I talk to Brian's AI a lot. It's pretty good. But we have 12,000 pieces of content, 17 million words. There's just, I've, I've written more And we have more executives that have come to SaaStr than he's written and trained his AI. So my AI is better. There's definitely a point of diminishing returns is what I've learned, right? But mine is better. So the learning is the more you can train up to a point, the better, right? So that's one reason it's very good. So it's just something to think about. I think this training can be— ours is muchly automated, but you're always going to want to add a human on top of it to, to get it to that nth degree. So that's my rambly answer.

That's a good answer. There's a, there was, I'm going to combine two questions into one because you kind of already answered about the hallucinations. There was a similar question on if we added any guardrails so that it doesn't like mispromise any pricing.

It's a good question. We haven't. I do think that generally guardrails are super important in AIs. Look, maybe I don't actually know that. Listen, what's really interesting is our tool is not designed for sales. That's why it's so interesting. That's why it's— guys, that's why this is so cool. When you talk about new use cases, this Delphi tool was originally designed to be a clone, a cool tool to clone. I mean, clone Brian Halligan, clone Lenny, clone me. And they built all this cool functionality to be a clone. And, and to some extent it's been architected as a support tool now. I don't know, the guys are coming that run Delphi, they're coming to Saastrano to explain how it all works under the hood. So if you really want to know, come to Saastrano because the guys that built RAA are coming. They're going to teach you how mine, Lenny's, and Brian Halligan's all work for real. You can ask all your questions. It's gonna be super cool. But what's interesting is it was not designed for sales, and yet it works pretty good for sales. So there may be guardrails in the tool. It's got a lot of stuff, but I don't— I think a sales tool would absolutely have guardrails, just like a CPU system. Um, what ours doesn't, but it does have— you can absolutely add, uh, limitations. I forget where it is. We could show a demo. It has a version of guardrails where you can say, talk about certain things, don't say certain things, here are limits on certain things. And I think if we trained it and had a more complex pricing, it'd be pretty easy to add. But we haven't added guardrails. And frankly, I haven't seen it make it— it made a couple goofy pricing mistakes that we fixed just by adding it, but I haven't seen it make a lot, which is probably pretty interesting as it is.

I'm just going to keep asking the questions and some of them are repeats. So there was a question from Alfonso. If you expect real customers are not going to be fast enough to follow the tech adoption cycle in AI?

Well, it's a great question. I'll give two different answers. We did, we literally did a deep dive with Marc Benioff, founder and CEO of Salesforce. It's our next podcast. It's number 800. It's going to be great. And we talked a lot about this and what he, and he explained it was very bimodal at Salesforce, especially with Agentforce. He said they have customers like Lennar Homes, and Singapore Post. I think he was just in Singapore and like, they're all in anything they can do on AI to automate Leonard Holmes and this, they're all in. He's like, but man, let me tell you, the enterprise is slow. He's like, overall, it's gonna take a couple of years for enterprises to get there. So one, both can be true, right? It can be segments. But the other thing is like, there's different types of AI. I mean, if you look at how we use it at Saastr, it just, you know, we, we used to have Intercom years ago. It just replaced Intercom. So we always had chat, right? But our AI is 10— for us, it's 10,000 trillion billion. You know, we would have intercom and, and people would create tickets and someone would check them once a month. That's as bad as it gets. Now our AI answers thousands of them a month. It's like a billion trillion better times better than someone never answering that. So, so literally the fact that our, you know, chat bubble used to be intercom and is now our AI, I don't think that's a lot of friction in the enterprise, right? So we may see a lot of enterprise uses. In areas where there's low friction and areas where it's high friction going to marketplace, it may take years. It may take years in areas that are high friction. Certainly, folks are of, in a lot of regulated industries, folks are of two minds on AI. There's a lot of spaces where it's slow, but it doesn't mean the areas that are easy to do aren't gonna happen very fast in the enterprise. And that was my point as well.

Yeah, I'm trying to, I'm gonna try and hit a few sales questions before we run out of time. So one from Shirash is how does the role of sales enablement change in the era of AI?

Man, it's— I think get ahead of it. I mean, in some ways, half of this presentation was sales enablement, wasn't it really? Right. I just think I would— I don't know, but I would be all over this idea, especially for in-person selling. We're still going to learn. I do think AI is going to make in-person selling better, right? Because we're— but it's murkier to me exactly how the tools play out than when we're selling over Zoom. Um, but be ahead of it. Like, be ahead of it because this, this Zoom AI is gonna do 99% of what, what we, what we work really hard in sales enablement to get our team scaled up for. The way we train salespeople is gonna be radically different. Our expectations are radically different. And I think maybe at, at a meta level, what I've also learned in the sports space, the way we're gonna do QA, quality assurance, seeing how good reps are has already changed. I'm an investor in a company called Maestro QA, which is doing tens of millions for enterprise QA for contact center. And this space is nothing like it was 18 months ago because 18, 24 months ago, I mean, they have big customers, you know, huge e-commerce customers. And 24 months ago, these customers would have thousands of human reps and all the QA was basically manual right now for their biggest customers. 30 to 40% of the reps are AIs, but the AIs make mistakes. We talked about guardrails, we talked about issues. So the whole way we do QA has radically changed. The same thing's going to happen in sales enablement. We're going to have to create, not only is AI going to make our sales reps better at many things, but we're going to have to check it very differently than just randomly listening to gongs, right? Which is the classic way, right? In testing, like we're going to need to screen every single thing through AI and what that means and how we're going to QM and how we're going to judge them is going to, it's going to change. I don't have all the answers, but you want to be at the, you want to be at the leading edge here because the, just like QA is for AI, I don't think the space is going to be the same in 24 months. It's going to be radically different.

Okay, next question on sales is from Luis. I know you and Sam talked about this a couple weeks ago, but when— and you talked about this earlier, so for folks who joined late, um, but when do you see AI actually replacing the SDR function?

Well, I purposely— that, that's one I didn't write about because I don't know. I don't know. I, um, you know, we did a survey, I don't know, 6 or 8 months ago on SaaS. We got thousands of responses. Have you ever had— made an— has an AI SDR ever worked for you? And almost everyone said no, right? And even the folks that that said, yes, it worked. They're like, well, it generated pipeline, but no deals closed. Okay. But that was a while ago, folks. That was 6 months. I mean, 6 months ago, we didn't have an AI. 6 months ago, this full duplex thing of digital Brian and Jason talking wasn't possible. The world's changed so much. So I'm actually a super fan of the AI SDR now. I'm super excited. It's just, I don't think it's gonna be the way we thought about it in 2024. Okay. In 2024, there, everyone was like, okay, let's automate sales loft and outreach, right? That's what we're gonna do. We're just gonna take this well-known paradigm. And people building Cadence and let's have an AI build the Cadence. And I still think there's a role for that. And those tools are great because if nothing else, those tools make the emails more professional. There's no question that when you run, uh, SDR content through AI, yeah, can you smell it and see it when you see it? But sure, but that's sure better than terrible copy. Um, but the idea that an AI SDR on its own can close business was, you know, a bridge too far, right? So I don't know. But what does make sense to me, and I, so I don't know, and that's why I didn't touch into this. And we're gonna have a bunch of the vendors at SaaStr Annual, and let's ask all of them where it's going, because I am bullish on this space. Because if you step back for a minute, I think we expected the wrong things and expected too much from AI SDRs. Let's step back for a minute. If you take an AI that's already just as good as what SaaStr's own AI has, and you thoughtfully develop your target accounts, and you thoughtfully make sure the communications are tailored to them. And the AI then makes sure they're all followed up properly. The AI makes sure they all get the right case studies. The AI makes sure they're all invited to your weekly webinar. The AI makes sure they all come to the right customer dinner in the right city because the AEs don't. And they make sure that when you're doing the dinner in Austin, everyone in Austin is invited. And they make sure that they're all invited because you're a platinum sponsor to SaaStr Annual. You can invite 50 customers, that they're all in your VIP room at SaaStr Annual. Why can't the AI make that, that, that high quality outbound better? You know, this at the edge of ABM, it's gotta make it much better. It's gotta make it. ABM has never delivered what we've hoped it has been, but it's always worked. So I don't know if, if where AI sales outreach goes, I don't know. But, um, but making all of, making the human 50 times better, I think is achievable today. So I'm excited to see where it goes. I just don't know that, that, that expecting an AI to bring you $10 million of revenue because you pay them $199 a month is realistic today. These other things are. So, but I am, but I was pessimistic. I was skeptical like many of you were 6 months ago. I am not. I'm all in on the AI star today. I just don't think it's gonna be the 2022, it's gonna look, I don't think it's gonna be anything like the 2024 version.

Yep. Um, really 2 related questions. Um, What do you suggest using AI for to complement specifically founder-led sales?

Well, listen, the, the, what we will see as tools evolve, I think the reality is there's, let's simplify sales at the early stage. There's outbound, there's get into deals, there's being middlers. There's closing and there's follow-up. Okay? Founders almost universally are best at stage 2. They're great middlers. They're great middlers because they know the product cold, they know the why, they can commit to the future. Like if there's a product gap, the founder can commit to solving it in a way nobody else can. And customers love to talk to the CEO. They love to talk to the founder. So, they're great middlers. There's like, everyone gets energized talking to Marc Benioff. At that scale, anyone, and don't be intimidated because you have 6 people in your company. The customers will love to talk to you because you care so much about their problems and you care about them. Typically, founders deeply care about their customers in a way that, you know, Zoom doesn't care about me. I mean, nothing personal. Eric Yuan's one of the greatest founders of our generation, but the Zoomer on, they don't care about me. But, but you care about your customers. So we're good at the beginning, the middle. We vary on our other stage. And most of us as founders, we're just dead awful at follow-up. Follow-up requires a lot of process, a lot of task management, Um, and so I think AI is gonna automate all the follow-up in the founder-led sales. I think it will help us with the opener if we use these next-generation tools properly and we don't expect to do everything. It's gonna help us with the opener. It's not gonna help too much with the middler, but our AI will join us. It will, I don't know if it'll help with the close, but it, it may, it may for routine deals. You know what we'll do? Like, let's say your founder-led sales working on a $2K or $3K deal, it may just follow up with you and close the deal on its own. And it'll absolutely make your follow-up. Either in the pipe or after the deal closed, like 100 times better because all founders are terrible at follow-up. They're all terrible at it.

Yep. I am guilty of that too.

Well, you're too busy. The founder— listen, I know it's not that founders can't do sales or business development. The problem is they don't have 50 hours a week to just do sales or business development or marketing, right? That's why they're really good at like the relationship building, but all the follow-up, sales, marketing, and business development, they're terrible. That's why founders You know, beyond this, business development partnerships, founders screw it up the most because they build a great relationship with Marc Benioff or Tobi at Shopify, but then there's 100 people at Shopify you have to work with on the deal and they fail, right? So AI is gonna help with all of this.

Cool. I'll try and squeeze in one, maybe two last questions.

Okay.

How do you deal with, I, I, I think this The question is, how do you deal with people disagreeing with your AI?

I think AI is great for that. We have seen it in our 30, that we do see people argue.

Yeah, we do.

And I think it's great because I will tell you my learning. If you'd asked me this question 45 days ago, I couldn't have answered it. Our AI is so patient. It's so patient and it will hear your complaint and it will, and it knows that you're being aggro. It knows it and it deflects it and it answers it and it'll answer it simply. And then it'll say, you're right, but, and it'll say, I hear you. And then if you want, it will ask, can I go into deeper? Can I explain? It'll try to explain things in more detail and other ways to you until it kind of wears you down with its honesty of your answer. And then people just end the chat. But that is better than 99.9% of humans can handle. So I think it's wonderful. I wouldn't have known 45 days. I think it's wonderful.

Awesome.

And they do.

Okay.

They do.

Yep. The last ones are really just, it's funny, the AI note takers are giving us the key takeaways from the meeting, which one of them was to attend faster. And the other one were to see how AI can help improve your sales enablement and quality assurance. Which is a little bit funny. The maybe last question here. Hold on. Let me just Okay. How— it's a good— I think it's a good last question. How often do you see the AI getting either hung up or kind of losing a customer if it can't answer?

Honestly, never.

I know.

Never. Here's the thing, guys. This is why I made this presentation. Never. Never. I've never seen— listen, we're just— we didn't mean to be a sales tool, right? Listen, do— Amelia has been at SaaS for 7 years. All of our— we have hundreds of sponsors. They all know her. They all trust her. They'd all rather talk to Amelia than the AI 40% of the time. Not actually, not most of the time, because they just want a question answered. They'd actually rather talk to the AI, but, but they— but fortunately they'd rather talk to her. But, but no, sorry, what was the question again? I got, I got a little bit distracted. Yeah.

How often do we lose deals or customers in the AI?

Never. Never. And now listen, Now, okay, I'm being a little facetious. It handles everything well. Now, where I guess in theory, if you had a very— if you were selling a fungible commodity and aggressive where our AI is not like we're going to work on this after SaaS, we have a little bit of time. Okay. When it literally does a handoff to sales, it is not a sales tool. So it is not— what it doesn't do is immediately reach out to Mona and get her on the phone that second when you need to go to a human. Okay. That is what tools like Gorgeous do. Like they will immediately ask if they can't solve the problem in support, it will immediately find a human. So that is a gap. And while I can tell you that it does a great job answering all the questions, maybe we lost a deal because it sent it to, to email and we didn't write the connection. Absolutely could have happened. But I haven't seen a single bad experience. Not one for the prospect. Not one. And I can tell you, man, in the early— over the— we have a really good sales team today, but over the years we've had some really I mean, literally terrible experiences with sales, right? I mean, we almost lost one of our— a partner of us that has spent over $2 million with SaaStr over 10 years, and they said they would never work with someone that used to be on our sales team ever again. It was such a horrible experience. Would never work with them. And they churned. We had to get them back. And you're just never gonna have that with AI. You don't wanna think that's true, but I gotta be honest with you, when sales folks leave, when they get angry, when they struggle, when, when they have a dumb boss, it's sad what people say and do. And they sometimes they don't even realize it. Right. And the AI never does that. The AI never threatens you. It never says if you don't sign today, you can't come to SaaS or annual. It never says we're going to— we're going to— we're only going to allow your competitor to come and not you if you don't sign this minute to get their commission. It don't do those things. It don't do that. Maybe it should. Maybe that works here and there, right, for an individual. Rep, like being super aggro can work to get their commission. I don't think it's good for the brand of the company though, is it? When that happens, when that happens. And, you know, maybe the meta question I'll leave you, I put this, I did this poll on Saastr. Let me see if I can pull it up and then we could break. Let's see. And maybe this is, this is the last meta point. I did the survey a while ago on Saastr. I'll try to find the link and then we can add how many, how many folks, how many sales folks said their competitors lie to win deals. 93%. 93%. This may be a profound change from AI. We can train the AI to lie a little bit, but if you think of all this bad behavior that happens in sales, right? I mean, deal rippling. Come on, the AI is not going to do that shit. Okay, terrible. And a lot of folks think this deal rippling thing is crazy. I don't think it's crazy. I think it's just like a little bit more extreme than what happens all the time. I think it's just a little more extreme. And AI never does this. AI always treats the customer with respect. And I think if you're building a brand, if you're building something generational, you want to treat your customers— we, none of us treat customers with enough respect. I love sales. SaaS is all about sales, but for sales, it's about the deal. It's about getting the deal signed and getting their commission. For the customer, this is the one thing we forget. It is— this is business process change. This is changing how I run my business. This is not just the $29 or $29,000. The money does not really matter to the customer at the end of the day, as long as it's fair. It's so much more work to deploy a new tool or a new business process. So I just want to be treated well, and AI is going to mean we're going to treat our customers better. So thanks for the time, everybody. We'll keep doing this and we'll do a bunch of deep dives like this in just a couple of weeks at SaaStr Annual.

Awesome. See everybody IRL there.