SaaStr 852: The Agents #003, Our Agent Now Runs Campaigns on Weekends, Plus Why We Pay More for Salesforce Than Ever Before

25:19Of course, we'd have to have a conversation, Amelia. But I think the meta thing for folks to take away is it's worth it to us to pay more, right? And that the Seat story, even at Salesforce, even at the classic, the folks who basically invented the classic Seat story is not that simple. Now that we have 20 agents running and really four or five core ones that run through Salesforce that are pushing data and pulling data out, mostly the sales folks one, but we may do more marketing. Net-net, we're paying 80 % more, and we're going to get better at it, so we will drive some costs down. Our agents will get better.

25:54We'll use less data. Maybe Salesforce will be, pricing will come down a bit. But in the wash, it's a good deal for both sides. Yeah. Right? People are missing the story here. And so as more folks go headless, this, I think this is, I'm, I'm for everything, for every core app. I think it's because we have some, it's such a positive to be headless. Right. And now I don't, maybe in six months, 10K, our VP marketing, maybe we'll have a different agent. That's like our support. But right now our VP marketing runs everything for us. Right. It is becoming our hub and it just makes Salesforce so much powerful to see Salesforce, not this isolated app with leads, contacts and opportunities, but it's integrated into every single metric, every single attendee, every single dollar.

26:36It's tied to how many newsletter subscribers we have and how many social media followers we have. It's all integrated inside of 10K. And it actually, it means we can do anything and run any analytics and any query we want against it. This headless thing. So anything else we, I think that the economics and how we use it or anything else that is interesting running headless that folks might not get? only because i think we're a little they just announced this right so like a week ago so i think this is from what i've heard from the folks at salesforce right this is still a newer concept for most of their customers community that they're you know i think people were still getting their heads around just asian force and now headless asian force they're like what does that even we were just getting on the asian force train and now you're here you're you're like here's the shiny bullet train to get on, which is headless.

27:29So I think that piece is interesting, but just because we've already been living it for six months, the headless piece, you know, to some degree, I don't, because the API is good, I think more apps and tools will become headless, right? Like as long as you have a central, I think the why headless for Salesforce in particular works is one, we already had a bunch of data in it. So to Jason's point, Like, did he need to log into Salesforce? No, he just needed a Vibet, a dashboard that hooked up into the API so he could see what all the rest of the team and the business is doing. Plus, he can talk to it.

28:05I think the biggest thing about headless with the API is you can talk to it in a way a CEO can't normally really talk to their data in Salesforce. You can talk to Slack bot if you have enough data, and it kind of replicates that vibe now, which I think is pretty cool. So I think they're catching up in that regard. But outside of that, I think being able to just talk to your apps with your API and say, tell me what's in the data. Give me ideas the way our AI VP of marketing has. Tell me where we're behind. Give me a list of folks that have churned that maybe I should reach out to that the sales team hasn't reached out to.

28:38You can just talk to it. And I feel like that is really the piece of it that is helpful, I think, for us. That sometimes Salesforce, I think they'll even admit, sometimes it's a little bit clunky to get things done. And it's hard sometimes to find things or know what the right filter is to get the right report or see the right dashboard. Sometimes that stuff is hard. And as a CEO, you may not have time or the want to just go through all the filters and figure out how to do it properly. But now if you just talk to your agent or talk to Slack bot, hey, run me a report of all these things and it does it for you, maybe it doesn't matter what's on the back end.

29:12As long as the API works and it can get you that outcome, I feel like more apps will start to become this way. Yeah, it was interesting. I'm on the board of a startup called Owner that you know. That's sort of a next generation agentic toast or so for restaurants. They're at about nine figures in revenue growing triple digits. And the CTO just made a loom and he shared it with me. You saw it too. They run Headless Salesforce and he built this before they called it Headless, but he did what we, he was inspired by us and he wrote his own on a hackathon. And as a CTO, what he did was a couple of things that were pretty amazing.

29:45I mean, they built their own AI VP of customer success that we did, but it does even more. It's pretty cool, right? They didn't really have this before so that one human can manage thousands and thousands of restaurant issues now. And it runs all underneath on Salesforce data. More interestingly, as a CTO, he took momentum, which we used to that Salesforce bought, and he used all the calls with all the prospects and one lost and one won. Okay. He looked at both. and used AI and Salesforce to analyze every single call to figure out what the top feature gaps are that they knew to win these deals and then the real reasons they chose owner over other solutions.

30:27And that was pretty cool. And at any point, what's great is like, I used to not enjoy working with Salesforce. Like it's not as a CTO, right? I mean, I love Dean, but at the end, I didn't used to enjoy working with Salesforce. Now we have this incredibly powerful data set that is figuring out why we win customers, why we lose them, and can in real time analyze every single one of our 10 ,000 existing restaurants and customers headless, that is so powerful, right? And I suspect they'll pay more for data, right? They probably have 200 seats, 200 salesforce. So this headless thing, it's a big deal.

31:00Okay, next topic I want to talk about is very micro. So that was very macro, headless salesforce. I want to talk about another thing you built this week. It's a very micro N equals one app. as I think I'm going to say that Darmesh Shah coined this term, but maybe he didn't. Maybe I copied him. Who knows? But an N equals one app, which is something that's super powerful, but only you need, right? So in a way, our AP and VP of marketing is that and that the custom version we worked used only this, but everyone needs an AI VP of marketing. But you built an AI parking pass app. I really think - Literally, no one needs this but us.

31:35Well, maybe others like us, maybe Coachella and WebSeph need it. But why did we build an AI VP of Parking Pass? And for us, it's actually disruptive, right? Coachella could use it. Coachella. For the shuttles, hit us up. All right. So just for some context and background, I've talked about it, I think, on episode one. But because this relates to everything of like costs and efficiencies and people. So just a quick backstory. We typically use an agency or two this time of year to help backstop, backfill some of our humans. Even now, still with AI, we still use a couple of agencies to help backfill our humans.

32:15Slash, you know, we have some folks that are really good at things that the agents can't do yet. So we always backfill every year. So I said a couple of weeks ago how much our bill has gone down with both of these agencies because we have such great efficiencies with our AI. And actually, they like it because then they don't have to do these tedious tasks that they used to have to do. They can focus on the fun stuff. So we have more fun stuff at Saster. And they can help me more with like the human in the loop support and like speaker support, again, that they previously didn't have time to because all their time was spent on like mundane, you know, step and repeat tasks.

32:47So that's a little bit of context. So every year, typically we'll see our bill go up the month before annual, which is now, because we have to distribute parking passes manually. Super niche problem, right? 4 ,000 parking passes manually. They're all individual PDFs. They're not just a code. They're 4 ,000 individual PDFs in a massive file, right? In one massive file. So the first thing I did is I gave the folder of 4 ,000 PDFs to Claude Cowork. I said, separate the PDFs. Because last year we had a person who did this, and it took them about a week to go into Acrobat. So tedious. And literally separate the patient one by one and then save them as their own file.

33:32Oh, my God. Not only, I know. That's 4 ,000. Cut in Acrobats to individual pages. Save it as a file. Get the file name right. Not blow your brains out because you did the same one three times. Yeah, that's a full-time week. It's a full-time week. Super tedious. So not only did Claude rip through it in, I want to say, two minutes. He said, here you go. Here are all your files individually. I put it in a folder for you on your Mac. I said, what the? i said actually if you're done in two minutes these parking passes we get from the venue don't have the address on them it'll say like you're assigned to the east lot or the west lot i'm like most people don't know what that means they're not gonna go look that up i said can you append the actual address the street address where they need to go so when they're driving to the venue they know exactly where to go he said sure thing maybe five minutes later i got all 4 000 back in a folder, not duplicates.

34:28He just updated the existing files, which I thought was pretty cool. He updated the existing files with the address in a nice banner at the bottom. So he highlighted it and stylized it already and said, this is the actual parking lot address you need. And I said, fantastic. So then I took the folder into our replet that powers that's radio annual 26 that you see here. Building a form super easy. I was like, hey, here's all the info I need to assign these. Okay, great. Then I said, okay, here's the kicker because this has to work now. Can you take this folder of 4 ,000 parking passes and wire up logic so that if I say I'm an attendee and I need all three days that you will give me the right parking pass?

35:14And it said, yes, here's what I need you to do. And it said, put the parking passes in a folder by day and by type, which Claude just rearranged the folders in five minutes. So he did that. And then give me the logic of who gets what lot. Like, okay, attendees go into like the east lot, you know, speakers go into the south lot, sponsors go into the west. So he was like, give me all the scenarios. If I'm filling this out, what do I need? And then I gave him all that. I gave him the folders again. And he said, okay, but can you put that into Google Drive so that I can actually access all the PDFs?

35:49So I did that. then he said okay now give me the logic he did some test runs right okay give me the logic if i say i'm a sponsor and instead of picking all three days i say i need you know just tuesday and wednesday i think he said i think my theory is i think i should give them a parking pass for tuesday and a parking pass for wednesday instead of blowing one of your all three days parking passes, which is good for all three days. I think I should give them individual passes for Tuesday and Wednesday. Is that correct? I said, yes, that logic is correct. Typically when we've done this manually, if somebody doesn't request all three days, we give them the individual days that they need.

36:31He said, okay, great. He said, let me rattle this up for you. I don't know. He's probably working for like 15 or so minutes in the background. And then he said, okay, it's ready. do a test and I said okay let's let's freaking try this so I filled out my email I'll fill it out right now and I'll show you guys what it gives me so I'll do my email plus like 12 or something because I tested this so many times

37:02alright so I'm going to put in my name I'm going to put in my company does it really need to ask for this but it's it's what's what it's doing is it's checking our api integration just lightly to see if this person seems like they have a ticket to saster versus giving it to a spam bot so i was like i don't know if i need all these fields but it felt better about giving it to actual people and not you know getting hit on the form constantly by putting that in so i'll submit that request okay check your inbox so your parking pass and lot assignment have been emailed to you okay so now let's pull up my Gmail here.

37:43Okay, so here it is. It's the first email in my inbox. It is my two parking passes because I picked two single days. And magically, here are my two individual PDFs. So here's the address that, you know, Claude appended. That's cool. So I picked Tuesday. There's Tuesday. I picked Wednesday. There's Wednesday. And then the last thing I did with Replit before we launched this is I said, make sure you burn these. So like, whenever you use a parking pass, they are - Don't send it to someone else. Yeah. Don't send it. I said, whatever you do, don't send it to someone else. I was like, hard code that in your memory.

38:23Don't send it to anybody else. This is one and done. Never repeat the same PDF or parking pass twice. And it said, okay, you got it. And then it said, But also, just as a backup, he said, give me a Zapier webhook and I'll put it into a Google sheet of which parking passes I assign to which person. I'll store it in the database. So if you're ever worried that I assign the same pass to people, you can spot check it and make sure I didn't. So super niche use case. But this, not only did the setup of it used to take us a week, literally distributing parking passes was a full time thing for all of us.

38:55Like we would all be on the inbox and we'd be like, someone else is requesting a parking pass. We would have a Google Drive. Like we'd have to make sure between the humans, we didn't use the same PDFs for a person. A lot of human error in that process with multiple people who didn't want to do, who did not want to do this job for weeks. I want to send people parking. Yeah, you have more important things to do. But it took so much time that it was one of those things I had to help the team with, right? I was like, it's just one of those things that's such a big time. It's like every year I have to give people parking passes.

39:24So I said to myself this year, instead of that, I'm going to take this maybe took me 90 minutes total with testing and hooking up all the PDFs and getting it all together. Like I'm going to take 90 minutes, one shot on Sunday and get this all done and routed versus hundreds of hours leading up to SASTR, which is a huge disruption. OK, this is a good transition to our next topic of where you find your Amelia. So you were able to have the experience and the foresight to identify, OK, what issues are coming up for a large event? And everyone's got their own issues, whether it's a release, an event.

40:03What are the big blockers that we could create an app, vibe an app that could solve massive blockers? Right. And so you quickly identified one that takes hundreds of hours of people's time is incredibly tedious, prone to error, prone to friction and creates hundreds or maybe even a thousand or more tickets and questions. Right. Because everyone needs one and there's issues. Thousands of it's you'd have to have a massive team to support this. Or frankly, what most folks do is they don't even try. They don't even try it. Like you just show up to a large event and it's terrible. Like we showed up to Coachella this year on our team trip with our agents and our three people.

40:37And the three-hour wait to the bus at our hotel, right? They just decide not. These problems are so complicated. Without an agent, you can't solve them with humans, right? So you solved it. You had the foresight to see it, to care, to scope out which top issues that were worth solving, right? Because we can't solve everything agentically. And then you executed it in about 90 minutes with a lot of experience. So this isn't now this isn't the first app you've built, right, or applet that you've built. So I want to give you a quote and then ask you a question. Where do we find it, Amelia? And Aaron Levy was on 20 Stebbin, Harry Stebbin's 20 VC this week and said there's going to be a whole new category of 500K to a million dollar New Year jobs for operators over the coming years.

41:23They're somewhat technical like you. You're not an engineer. You can't really code. But at this point, you're getting somewhat technical. They're deep in AI, MCPs and CLIs and skills. We need these folks. And I know we would both agree, absolutely everyone needs one of these folks, right? You don't even need a hundred, but where the hell do you find you? Everyone needs, like what I just described, every organization needs this, right? You can imagine a hundred blockers you have in your company, right? Someone could even just vibe code the fixes to Marketo if Marketo had a couple of folks like this, right?

41:57So any thoughts on this? Because I think Aaron's right, but I think he's understating how hard it is to find these folks and what it takes to build them internally. Yeah, I almost feel finding these folks internally may in some ways be easier than going out to the world to find these folks because how will they have the – it's chicken and egg, right? Like even we've tried to loosely look for somebody who can help with our agents. And like, there's just no. Yeah, we'd hire another Amelia if we could find her and pay her well, right? Yes. So if you're listening, hit us up. But we can't find another Amelia.

42:37We literally can't. I've had people message me on LinkedIn and say, OK, you know, I've done this XYZ with agents. I know you guys are hiring. I'm like, that's not good enough. Like you just employing this little agent that doesn't do a heck of like, I'm glad you deployed an agent. but it doesn't do enough. Like I need you to understand all of our agents 24-7, but more than anything, I need you to understand the product. And I feel like that is the harder part. Maybe it's a little bit easier for folks to find somebody internally. So I'll give you guys a good example from Salesforce, right? So Lindsay, who's our FDE, it was not an FDE previous to probably November of 2025.

43:21she had a different role at sales port but at sales force she's pretty you know she's more technical than me less technical than you so she's pretty technical but not a full-blown you know full stack engineer um the way some of their fdes are she was more on the se side right because she sat somewhere in the middle but not only does she know the product she's so good she was so good at helping people like us get deployed that they have now promoted her to like lead some of the FDE team because for them to find this person internally they were like okay who's hungry who's already kind of spending some of her free time on agent force just out in the world who was spending her time you know in slack bot testing all the features trying to break things who was intellectually curious about getting their customers deployed who had the most deployments who had like the best like ratings and she it's because she knows the she knows all the here's the thing she knows all the products salesforce has so many freaking products she knows enough about all of them she'd be like oh i know you need to use this xyz thing that we have at salesforce or this connector to do that and you can use slackbot to do this and you can use agent force to do that and now you can use qualified to do this and momentum to do that like she knows enough about all the products that she is i consider her one of their amelias right because she can absolutely spot problems before they happen because she's helping with deployments.

44:47She knows the products. She's helping with deployments every day. She just told me she's got like four customer deployments today she's doing. Plus a demo with me at the tower for their summit. So like that's the folks you want. You want to find - That's a good insight. If I had to hunt, if I wasn't going to homegrown my, you know, VP of agent operations, right? If I wasn't going to homegrown that person or head of, Maybe head up. We don't want someone that won't do the work. Today, as we do this, maybe recruiting someone that's a strong FD anywhere for this job. That just doesn't want to be like they don't want to do the 28th Replit deployment.

45:22They want to do a little bit more. Maybe you can steal them from Replit or Lovable or Rusell or wherever and pay them well because they're not cheap and say, hey, you get to deploy 20 agents for us over the next year from diverse vendors. Most people will just stay put at a hot startup. But in theory, you get that person. That's a good persona, right? One of the top FDs because they've seen it all. That's who you want. I think there'll be a lot of folks going to the comment you saw, the folks that reached out to you linked to, there'll be a lot of folks with very limited agentic experience, which is still better than nothing, right?

45:56It's still better than being terrified, like so many people are terrified of rolling out an agent. But their ability to have translated one chatbot they released over six months to running 20 agents for you, it's too big a leap. Yeah. And people are going to make this bet, and they're going to make the fancy LinkedIn bet. And I think these hires, these air-leaving hires, are going to churn at a higher rate than CRO because they'll talk the talk, but they won't be able to deploy the agents. Yep. But the other thing I've seen commonly, I was just on a panel recording and the CEO, one of the CEOs that was on the recording for this Inc thing that's coming out, I think in a week or so.

46:34He was talking about how he solved this problem. He's got like, I think, 120 humans or something. And he has a lot of agents. He's like a Salesforce partner. And the way he solved this, which I've heard from other folks, too, is he hired an ex-CEO. Like he hired an ex-technical founder who was burnt out on his startup, didn't want to do this for himself anymore, but was technical to be, I forget what they call them, like their head of agentic operations. And now he's the one. That's always one persona that works. Yeah. That's the persona that works. And he said he's had this person now for over a year.

47:06Like he hasn't gone anywhere. So retention has been good there. But I know a few other companies have done that too, where they've hired ex-CEOs that were technical founders to come be their head of the agents. So that's another one that works. But obviously pros and cons to that one, right? Like me and Lindsay are like a little bit younger. If you're hiring somebody who's maybe a former CEO and a technical founder, they may be a little bit more burned, right? So I think you just got to get different things. That's the biggest race. Sometimes they're burnt. They're too burnt. Yeah. But I will say, you know, almost everybody, you know, folks have been using this playbook.

47:37Parker Conrad quietly kind of, I think there's hundreds of ex-founders that work at Rippling now, hundreds. Yeah, there are. And everyone's done this. Adam from Owner has quietly done it. everyone should have a couple ex-founders, if you can. Listen, they may not go if you're slow growth and more traditional, but if you have a reasonably hot startup, hire a few founders, if nothing else, to kickstart new things. It could be a new BU, a new geography, but an agendic deployment is a fun one because a smart generalist can replicate what you did. You're a smart generalist, right? And so today, that's a good persona, like the super smart generalist that loves product and knows what an API is, that's probably enough, but they really have to meet those criteria for real.

48:19You can't cut corners here. And I feel like founder mode works in that scenario within your company because founders will figure out what works. And I feel like part of this issue that Eric kind of raises is if you go out to deploy AI agents for anything, but you don't actually know internally what success looks like with an agent or don't take the time to figure that out in the deployment, you're just gonna you're gonna end up with slop right so i feel like that's why founder mode works a little bit better maybe if you can find this person could work you gotta manage the burnout factor but i think in some ways it can work maybe better because these are folks that will you know take the time to figure out okay what do i need to do in sales to get the agent to work what do i need to do in cs and marketing they'll they know every function right they they so across the board for functionality, they get points, right, for setting up your agents and deploying them because they'll figure out, okay, that sequence didn't work or that context didn't work for my sales agent.

49:16Let me try something else versus just letting it run. Yeah, I think one thing, and watching what you've done in particular, but also a lot of the FDs we've worked with at vendors, I'm actually, it's not something we would do, but I'm actually more optimistic that as the year goes on, consulting firms and agencies that are good at this are the key. Right now, if you hire a mediocre agency that's been doing this for six years with the same people, you're going to see a lot of recycled folks that aren't agentic enough do it. Just like so many folks are trying to recycle their customer success folks in the FDA is it's a failure.

49:48But I think as the space matures, if you've done six agent force deployments or you've done 20 artisan deployments or you've done 10 Monaco deployments, there'll be agencies that specialize in all of these things. And I do think they'll get pretty darn good. It won't be cheap, but they'll get pretty darn good at doing the 20th or 30th deployment. And I think maybe they'll add even more value, these top consulting agentic firms, than they did when it was just setting up a Shopify store for you. That seemed hard enough back in the day. But actually, if you can set up an agentic store, it may be more valuable than just uploading assets and connecting you to a few APIs.

50:26Yep. Yep. It's funny. We used I'm still a little anti-consultancy firms at Saster, and I'm still a little anti-fractional folks. But listen, if you can do it for real, if you can actually get these agents into production, I think those folks are like the top one or maybe 0.5 % of agencies. There are some out there already that can do this really well. There are some. There are some that can do this. I think those folks are going to have such a boon. and the agencies that, you know, waste all your time picking tables, you're not going to need those people. And I don't think you need them now, but you're definitely not going to need them in three months because you're going to say, you know, this other agency can actually deploy my agents.

51:10Maybe that's what I care about more. Yeah, I think it's going to solve itself as they bifurcate these consultant agencies. The one Captain Obvious point I'll say, and then we'll go on to the next one, because we do a lot of it, do reference calls. The reference call died sometime in 2020 and 2021. I mean, literally it died during hiring. It died everywhere. But we get a lot of them. Even I get them. You get more. Hey, how is your implementation with Qualified or Artisan or Monaco or Agent Force? Like we had a lot for Agent Force. Do this is do it. Like if you're going to and do it. But I mean, don't just do it yourself.

51:43If you're going to hire an agency or a contractor or consultant or a fractional firm, great. Maybe only 1 % of them can really do what we do. Maybe it's 5%, right? But ask for three references and just email them. And we will maybe Amelia is more patient than me. I won't get on the phone with most of these folks. I don't have time, but I will do the reference. If you LinkedIn me or email me about one of the earners, I will give you that reference. I'll give you the five or six bullet points. It's super valuable. And if we tried something and there's not much, but if we tried something that's failed, right, because mostly we've invested in things that have had success.

52:14I'm not going to be a jerk about it, but I'm going to tell you, I don't think it's a good fit. So this is the, as you're trying, I know this is kept an obvious, but as the reference check has died, please bring it back for third parties to help you on agents and deployment, whether it's hiring or consulting or agencies. And I say this because I've, you've gotten more than me. I've gotten more reference requests on folks to help and vendors in the last three months than I've gotten in my whole career. More than my whole career. And it's heartening me. Again, I'm not going to get on the Zoom, but I will respond.

52:46Even I will respond if you say it helpfully. Hey, I need 20 minutes or can I buy you coffee to talk about your agents? Quick question, right? How is working with Monaco been? I get like three or four of those a week. We get a ton of them for Artisan, right? We get a lot of them for Agent Force. Make it a low lift request on folks that you trust like us and do three. Like just do that and then go for it. So, okay, related. Two last points and then we'll wrap up. really simple, a low-level one relates to this on support. And then I want to talk about who CIOs are getting rid of and how that compares to our experience.

53:19So the low-level one, on the one hand, this is really, really, really, really minor compared to a lot of stuff we have. On the other hand, this has been emblematic of our agentic journey. Okay. So a tool we don't talk about much is one called GetRecall. Only I really use it. Okay. It is an early agentic tool. It's great. I think it's been around since the beginning. It might've been the second tool thing I used after Delphi for my digital mind. And it's another version of a digital mind. It will take every YouTube you've done, every blog post, everything you've ever read, everything you've ever seen and create essentially a digital repository where you can talk to all of this with AI instead of just talking to whatever chats you have in Claude.

53:56And so what I've done over our history is put every podcast we've ever done, a lot of our blog posts, but especially all of our podcasts and YouTubes for history. And at first I used it because to get this to work, They needed better transcription service than I can find otherwise, because if the transcription doesn't work on the YouTube, you can't really talk to the YouTube, right? And you can get a transcription out of the API and there's other things. But it just, you know, I asked ChatGBT or Claude, what should I use? And a year ago, it said, use this one called GetRecall to get the best, easiest transcripts.

54:26And so it's been great. I've used it every day. And they did a big release this last week, which I don't care about because I only use a subset of the functionality. and for about an hour it was broken. Okay, for an hour. Now, they fixed it, no criticism. And like a lot of times, I might've been even the only one that noticed because how many folks are uploading their YouTubes and really using it for their better transcription? Probably only a small subset, but I noticed it was broken. It didn't bother me that much, but I tried it like eight or 10 times like most folks would. And then I saw that the error message wasn't the something a product person would write, but it was a classic engineering error message.

55:06error number 72 or API unavailable, right? Just it's clear there's a bug when you see one of those messages, right? That looks very, that looks, that was not written by a human or, yeah, you know, there's a problem. I've been around software. So at an hour or so later, it did resolve itself, but it's a long story. Here's my point. For the first time ever, I tried to reach out to support for GetRecall. Doesn't exist as near as I can tell. There's no chat window. There's no email. There's nothing I could find. I had an almost worse problem trying to get my own clerk working, which we talked about before, right?

55:39With almost non-existent support. So my rambly point is, and maybe, Amelia, you could add to it. I just want to challenge everyone in the agentic space. They seem to have skipped support. Listen, I love Replit. We have the best relationship with their FD and the engineering team. They're the best. Like, I love this team. But when I started on the Replit journey 12 months ago, they had no support. They've added it. They've upscaled. They brought in a whole team. It's great. And I get it. A lot of these are developer first products. I don't know that Claude has great support. It probably doesn't, right?

56:12We've never used it. So if you're an engineer developing, if you're a developer developing something that's developer first, maybe figuring out how to do support is not high on your priority list. But I got to tell you, like I just have these with GetRecall for the first time with Clark for weeks. I just have endless, I try to be self-sufficient and then I try to ask Claude to help me and then I play around and then I come back in an hour. But if I can't get past a problem in an hour, I need help, man. I mean, I got stuff to do, Amelia. Yeah. So anyhow, you could add to it. But I just find that so many of these products that we talk about and love every day, unless we have an FDE, unless there's something, we get no support.

56:50So do better, guys, because it remains the best investment you can have in references, referrals, NPS, and push it down. Don't just do FDEs for your Fortune 500 companies. Try to do them for everyone if you can. It's okay if you break even. Like having Epic experience requires not just Epic onboarding, NFT requires Epic support. So that's my rant. I just, and I don't want to, they fixed it in an hour. I just, maybe they have support. I couldn't find it with my, you know, limited patience. And I can tell you more stories, but I won't. So anything you want to add to that, Amelia, to my push or cautionary tale to please, please invest in support, hot AI company.

57:32It's funny. I feel like so many people have said in their data, and I think it's true on Replit and Lovable and certainly Agent Force, that support is the number one use case. There's the irony, isn't it? That's the irony. I'm like, but a chatbot does not equal true support. And I feel like that's the difference. Like, even what we've done with our AI, VB of Customer Success QB, we added a chatbot to him. But I've said now for three episodes in a row, he's not a replacement for what we do. Like, there is a chatbot. There's a million AI on Sassery Annual, which sometimes people chat there as our customers.

58:13They still email us, either me, David, anybody on our production team, they still email us. We still respond, you know, as quickly as we can. It's not a replacement. Just having a chatbot on your website is not a replacement for actually fixing the problem. I think it's worse if there's no chatbot because then you just have nothing to your point or you get a contact us form like a Marketo. But having that does not, that should not exclude you from actually backfilling support, from finding FDs, from finding folks that know the product, from finding people that can actually solve your customers' problems.

58:48I feel like the irony is that too many folks have said, oh, I've got a chat bot, so people will just figure it out themselves. I'm like, no, especially for agents. This is why I'm not a fan of self-serve with agents either. You need support to get these up and running. As good as Jason and I are, we need FDs at all these top companies to help us get them and maintained, right? We've needed some help with maintenance from our FDs. So I feel like, yeah, that's really the true irony of April of 2020. It's an irony. I will say, and you're right, I will say even before that, just at a deflection level, I am shocked how many products just don't have a well-trained support agent.

59:27It's so easy. Go buy one. Or we just built, like Amelia asked me the other day if we could do a good AI chatbot on Qubia, right? The customer says, I said, I can build for that for you in five minutes. So we did it in five minutes. And then we told it to train on the data it had. And then we had to keep training it with top questions and then build a document and retrain it. So it takes a while. but there is no excuse to not have a support agent that is trained on all of your website documentation that ideally is connected to some sort of backend to get some resolution for customers. Most folks pre-AI have this.

1:00:01Okay. Everyone in e-commerce has this, but I'm shocked how many AI folks, yes, you need a human, you need escalation, you need all this, but I'm, but they either have nothing or it's untrained. It's like literally they buy a tool because they don't care and they stick it stick the javascript on the website and they ignore it just like we talk about these are ai leaders so just buy any support tool on planet earth and then have someone's job to train it and then to update it every week like just at least do that right at least do that um especially when you get that right it will route the cases to the right humans right like that's where the magic are qualified and our qb right now is it's not one it's our replacement but two since it's like it will give people answers but if it gets stuck or if it thinks this person still needs help it will route it to the right person on our team not just a generic oh we're you know we're going to route you just to all of customer success and we'll get back to you or all of customer support and we'll get back to you it's yeah okay i've tried to solve this person's issue i've given them some live help versus just sending them resource links like i've tried to triage them by you know chatting with them they still need some help here's their specific problem, here's the right person that it needs to be escalated to.

1:01:11Yeah. Okay, the last one, and maybe we've hit some of it, and then we can wrap. But just because this was super popular on Saster, I wrote up a Redpoint report, Redpoint Ventures, top VC fund, investor and owner that we talked about, and many other leaders, Stripe and others. They did, they just surveyed 141 CEOs the other day, and 54 % said they're consolidating vendors. They want to shrink the number of B2B vendors they have. and that 45%, 45 % of all this AI budget, which is leading to explosion, explosion in clod, explosion in hot legal startups, explosion in everything, 45 % of that is coming, it's displacement.

1:01:50In other words, it's being stolen from old vendors that they're either canceling or cutting back the number of seats or cutting back. That's massive. 54 % are consolidating vendors and 45 % are grabbing budget for all of their spend. It's not from net new budget. 45 % is not net new budget. In other words, 45 % of AI spend is that CIOs are trying to grab from existing vendors. And they listed, and I just want to get Amelia's quick thoughts we can talk about, and then we can break, because we hit some of them. And what are they consolidating the most? Customer service, we talked about. Finance operations, we can do a quick update on.

1:02:23Project management, we can talk about. And sales automation, which we've done a lot of. So customer service, I mean, we have spun up pretty good agents on our AI VP of marketing, our AI customer success, digital Jason and digital Amelia, and we haven't used any legacy vendors. Did you ever think to use Zendesk? Did you ever think to set a time with the Zendesk reps calendly for next month? No, never. It's just, if I can vibe code for you a very strong AI powered customer success agent for QB, it's pretty disruptive. it's disruptive he also with any third-party tool it's going to be out of the box right everything that you do that's custom to you you have to kind of force through or find a way to do and the magic of our ai vp marketing and our ai vp of customer success is it's built fully custom from the ground up for us for sastra just for us for our customers the way that we have unique customers who are sponsors and media sponsors, like in our, you know, segment vertical, like everything is hyper personalized, both disaster plus our customers, plus, you know, the attendees who are customers of us as well.

1:03:43So everything is hyper personalized to them. Now, we've always said to like, we're still in the 90%, you know, buy versus build mode. And that is still true, right but i feel like when it comes to cutting things or or reallocation of budget i've we're reallocating budget right now like we were just talking about how we've we've reallocated some now this is a slightly different example but we've reallocated some budget from other tools to increase our salesforce um tooling but i took from other tools we were using that we don't use anymore put it into salesforce there are other tools we don't use anymore that i reallocated into things like ours in Monaco.

1:04:21Like we're doing a bit of both, right? Like we are doubling down our spend for one of our legacy tools in Salesforce. And we're also reallocating budget. Right now that used to be in salaries and people into these AI tools. And I feel like folks will start to see, you've already seen it in the data, but I think we'll start to see more of that. But I, you know, to our point though on customer service, I feel like that's where maybe people take it a little bit too far of, They're like, okay, I've got an AI chatbot that fixes all the problems. Maybe for e-commerce. You still need humans, though, especially if you're an agentic product.

1:04:57You still need a lot of humans and a lot of FDEs, however you classify them, to get your customers up and running. All right. The second one that CIOs are consolidating is finance operations. Right now, we're Brex, Bill, and I guess some QuickBooks is sort of our slightly embarrassing stack for finance. Am I missing any finance apps that you can think of? Are those at risk? Do you think Brex and Bill and QuickBooks have a gentic risk at the little Team Sasser? Yeah, and I'll tell you why. I think QuickBooks is, for me as a non-finance person, listen, I've always had a login for the last few years.

1:05:40It's like the Salesforce example. I haven't tried to hook up their API yet, but if it sucks, I'm getting off it. Like, I need some sort of agentic API that I can use with my agency. Oh, my God, I need that, too. When I ask that, like, some of the external finance folks, it could be a month or two until I get an answer. Not only that, as we've said, they're so bad at collections. Yeah, collections is the number one issue that we have, right? Collections is the number one issue. So I haven't even tried to hook up QuickBooks. I'll see if I have time to before our next recording of this. But I'm going to find if that doesn't work, which I suspect it's going to be dated because even the platform itself is pretty dated.

1:06:21Like I suspect, much like Marketo, if I can hook up the API, it's going to be fairly limited. So I'm probably going to move off of it. But I'm going to find something that is at least the API can talk to our agents. But also the thing about finance is it has to be secure, dude. Like I can't have you leaking billing information and banking information. like that's hypersensitive, right? So that's a double-edged sword on the finance piece, which is, I think that piece is why it's so surprising that the finance thing is so high because you're like, that's maybe the most sensitive data all of us have is our finance data.

1:06:59Yeah, I don't know if folks are necessarily thinking of replacing their ERPs or QuickBooks. I think they're, maybe they are. I think it's the rest of the stack maybe that's at risk, right? For sure. when we build an agentic accounts receivable agent what will that displace will it displace quickbooks will it displace brex brex bill and quickbooks have a lot of data in them so they have modes and they have a double mode bill and quickbooks have a double mode in that our third our contractors and our agencies know them well like accountants and stuff and we No, Bill and QuickBooks like the back of their hand.

1:07:40So moving off to that, to Bob and to some hot company from my Combinator instead of QuickBooks and Bill, is it like we're creating massive friction with our contractors and agencies, right? For sure. Yeah, for sure. Might be hard. All right. Project management is a third category. I mean, we're not going to ever buy a project management. I mean, we do use Notion, right? But Notion, even for us, is at the edge of stealth churn, isn't it? I mean, we barely use Notion anymore, do we? Do we even need Notion? I don't use it anymore. We've definitely self-turned off. Well, it's funny. Oh, yeah, we don't even use Notion for our stand-up because Notion used to be our stand-up, right?

1:08:16Now, I didn't realize our AI, VK Marketing, has replaced Notion as the head of our stand-up. You're right. I don't think I've used Notion in months. Yeah, there was a few weeks of overlap when I first built our 10K VK Marketing, because, again, I built it as a dashboard store. I built it as I couldn't get, and this is just a failing. I don't think this is a failing on Ocean, more of a failing on me. I couldn't for the life of me for one more week, log into something and manually update numbers. I was like, you know what? If I can vibe code, this is where 10K started. If I can vibe code an AI VP of marketing just for dashboards that has all of my ticket data, all my revenue data from Salesforce, all of my everything of like where, you know.

1:09:01And if Dave and our sales team doesn't have to update it because Salesforce just pulls his pipeline automatically, wouldn't that be awesome to save me an hour every Sunday so I don't have to do it on Sunday? And so it was for selfish reasons because I was tired of literally copy pasting into Notion that I originally built our AI VP of marketing. And so for a few weeks we had overlap because as we've shared, our Salesforce integration kept disconnecting. It's now been fixed. But since it's been fixed for the last couple of months, I haven't logged into Notion. You just realized you haven't either.

1:09:34I mean, we probably pay$100 for Notion, right, or$200 a month, right? But yeah, we stealth churned. We didn't – months, we stealth churned when we built our – now, it's interesting. Notion does have a lot of agents, and that is they're accelerating. I don't know what they do. It's possible if we hadn't started on our journey earlier and were good at vibe coding our own agents, Maybe we would have futzed with Notion's agents. And maybe if we have time, we will. Look, Notion remains a great pre-AI product. So if you can get the agents to do it, like if our Notion could have connected to our Salesforce and built dashboards that random people couldn't delete and all the other issues we have with Notion, like if we could have built it in Notion before we built it, we might have.

1:10:16But now we're gone. Like we're not going back to Notion. There's no chance we're going back to Notion. Like we're gone for, they don't know it in the bill. But, you know, I made this point on 20VC to everybody. Dudes, you got to look at the mouths and dows and wows. You got to in the stealth churn area, because if you're feeling all confident because your product's cheap, so people don't. So there is a laziness to churn off credit card payments, right, for a lot of reasons or whose credit card is it on. There's the prisoner dilemma with like a bill or a QuickBooks for us. But look at those dows and wows, because our dows and wows and mouths are zero for Notion.

1:10:50And I used to think these were like B2C consumer metrics. Now I think they're the number one most important in B2B. Yep. Because you're under threat everywhere from stealth churn, from agents doing part of your workforce. Our dows and wows and mouths, if you include API calls for Salesforce, are like 100 times a year ago, right? That's as green as you get on the health meter, right? They went from red to green. But our notion once, like I know no one's going to care to follow up with us for 100 bucks a month, but we've been red light for 90 days and no one's even emailed us. That's an F, isn't it?

1:11:19that's a wild I haven't even logged into Canva I've logged into Canva twice I think in the last 60 days I haven't gotten any email or anything from any agent these are you gotta track the thousand miles and miles these are stealth churn everywhere I didn't even know we'd stealth churned off notion but now that I've been like that oh cancel I don't care do a quick export but I haven't even looked at the dashboard in a couple months I don't care it's not as good as 10k I don't care I don't know my log stealth turn man that's brutal okay project management we already changed our whole stack i didn't even realize it goodbye love you notion but goodbye i don't need you anymore right you were great when you were great but we don't okay last one that folks are consolidating or turning is sales automation look we've talked about headless sales source on this we've talked a million times about our aisdr stack including today just and you and i haven't discussed it but is there anything else we're thinking of adding beyond this to our sales automation stack that would even more lead to change or consolidation?

1:12:19What are we thinking in sales automation as our final point? Yeah, I think where we're pushing this all now is how much can we get it to self-sell? And so right now, with all of our different AI SDRs that we have, maybe you guys have noticed, we don't have an AI VP of sales. Like we do not have a VP of sales by definition. That is, that's one kind of, we have momentum, which is joining all the calls, which is great because it'll send me, you know, summaries. It'll give us accident items. It'll move stages for deals for us in Salesforce. But it's not at that point of a true human VP of sales where it can join the call, say, hey, I also joined this other call.

1:13:06And this is how you should do the follow-up to this person. Like we get that very lightly right now from our current staff. very light, right? But right now, it's still a human taking the call. Our AI agents do. So sorry, there's a lot going on there. You want to figure out how we can close without a human, close a 50K or 100K deal without a human? Or what are you saying by that? What's your goal? Yeah, I would maybe not for 50 or 100K, but I think for something like our... Sometimes we do these add-ons for Saster, something like an add-on, something like a media, a podcast. Yeah, sponsor our newsletter, right?

1:13:42Sponsor our newsletter, right? Fixed pricing, fixed work to do. Fixed pricing, fixed availability based on like our current calendar, right? Yeah, that should all be automated in a way, right? Yeah, this month for April, we're really full and May and like Jude's getting really like, just take all that data. And instead of me or David having to get on with these deals, can't the agent get them either fully to close or maybe to contract where we just send the contract and say thank you to the agent for the commissions. That's where I want to see how much we can push the limits there, because I feel like our tools are really good and they are helping us close a lot of deals.

1:14:22But at the end of the day, we've still, you know, a human has closed the deal and sent to the contract. So I want to just push the limits there and see how much. So one is, can we get our agents to close more like 45K and below routine deals? Higher ACV. Yeah, that's the question number one. Like right now it's closing, our agents have closed a lot of ticket revenue, right? Like hundreds of thousands now of ticket revenue. Maybe, probably seven figures. Probably seven figures now, right? Yeah, because that's what I looked at a couple of months ago. But the ACV on a ticket previously was sub 1 ,000.

1:14:54Now it's 1 ,000, right? So low ACV, it can close on its own. It can send people to code. It can follow up with them. It's already doing a very light ACV version of this. So the challenge to myself now is, can I get that to work for a higher, for real ACV? That's like more like 25 to 50K. Can we get that to work? And so I think that's enough for today. I guess, you know, the last one, and you didn't make your list, so it's telling. All these terms, orchestration and this, they all drive me nuts. or even training, but the idea of having an agentic fabric, at least in sales, that can manage all of our GTM and sales agents, that would be nice, right?

1:15:33Is that on the... Do we think we could do that this year or do we think having a true agentic fabric that can manage all of our sales and GTM agents, do we think that that's just not a solvable problem we're solving this calendar year? I would like to. I think I call it more of like a genetic sales aura, if you will, of like something that's just always there, right? Like could I hook up our, let's say, five core different sales and now CS agent because it's part of sales. Could I hook them all together enough, either using Salesforce or something we custom build on top of our Salesforce and our other agents, to be always there, like an aura, like in the background?

1:16:12An aura. But it would have to also manage all the other agents, too, for it to fully achieve its goals, right? I like the aura idea too. Don't get me wrong. It's always there, right? That's my hope is by, you know, fall maybe of 2026, we can get to a sales aura that's always present, always on, can sell, can escalate to human. In the way we've done with some of our other, you know, lower ACV or marketing things that we've done. And I hope we can get there. And I hope it's in a way that can truly manage all the agents together. Because right now, that's our biggest roadblock, right? It takes you and I a lot of human hours to manually manage all of our agents.

1:16:55All right, Selzer, I love it. Okay, we'll wrap on that. If you've made it this far to the end, thank you. And if you are on YouTube or go to YouTube, literally put in a comment your question for me and Amelia. I can't promise you Amelia is busy. But if you look at what I do with Harry and Rory on 20VC, I literally respond to every comment there. it's a very efficient way like put a comment in the agents on YouTube questions you want answered on your agents and as long as you don't mind them answered someone in public at least I commit to doing my best so thanks everybody we'll see you next week Amelia you're the best and we'll keep learning and try and make this better thanks everybody