The Exchanges

Q So we get that analytics leader, we get that setup in place, What are we looking for then? What is good metrics, and what metrics should we be optimizing for? Is it acquisition? Is it retention? Is it spend?

A When you're first getting started, uh, the main thing that you optimize for is, like, it's the most important event to you as a business. So SaaS, it's, it's the subscription, or let's say the, the start trial, or in Whisperflow's example, it was, you know, the download, right? Super upper funnel event, um, because we had a desktop client and an iOS app made it super hard to track. But, you know, pick what that event is and essentially validate that, okay, we spend a hundred K. What is traffic cost? What is our acquisition cost per user, right? And, and just focus on acquisition cost because PLG and the product team are going to focus on retention as long as you bring in the quality user, right? And that's why over time you'll probably change the conversion action to be what the most important event is, right? So for example, at Victor, uh, we have, it's like a pretty hard funnel to go through. It could be an admin in Slack or let's say you're a smaller team and anyone can add apps or You need to be a tenant in Microsoft Teams. You go through onboarding, and then you add Victor into Slack or Microsoft Teams. That's our conversion action. Super down funnel, but we know that we've optimized for that, and like, because we're super high ARPU, we can have a couple thousand dollars in CAC, and the economics will work, uh, but you need to, like, take the time to find that out, right? …

AI assessment note: “just focus on acquisition cost because PLG and the product team are going to focus on retention”

Q How do you do 400 to 500 new creatives a month?

A You have a creator program where, like, at, you know, at Victor, we have, we work, we have our own creator program where we give a percentage of ad spend to people to essentially create ads for us. We obviously, you know, have some storytelling and They use the product, so it takes, you know, a little bit to acquire them, but once they do that, they create three to four videos a week, and we have a couple hundred of them. We work with five agencies. We also have our in-house creative team. So again, Ecom model. Ecom has been doing this for over a decade. That's, that is how the entirety of Ecom has functioned. UGC, creator programs, hundreds of thousands of variations of creative, and they've been doing that, and that's like in their DNA. SAS, Had to grow up to this. And like, because I come from that world, that was the, the instilling factor that I always brought in.

AI assessment note: “we have our own creator program where we give a percentage of ad spend”

Q We mentioned the three channels, Meta, Google, Lifecycle. What are your lessons on how to do YouTube well? We hear about YouTube being the acquisition engine and machine that it is. Any lessons that I should know as a young founder?

A So, I mean, it's a totally different type of creative, right? The, uh, so if you look at, uh, Meta and YouTube, they have, like, similar but slightly different, like, hook curves where, like, the, the hook happens Slightly later than like, because on Instagram and Facebook, you want that immediate validation while on YouTube, you're kind of there to spend the time, digest the video. So you have a little bit more of a longer attention span. And a lot of people actually are terrified of YouTube because they're, you know, the best performing YouTube ads are 16 by nine. So they're landscape. And most people are like, well, I have a UGC program. I just have story placements. What am I supposed to do? Um, so we did a really cool thing at Victor. Uh, we had Victor build an app for us that takes any story video and adds it into a static, uh, template that has like logos that use us, G two rating, and then a CTA, and the video is the replaceable asset. And we turned all of our story ads into YouTube ads. So every single one ended up on YouTube. Not all of them worked the same. We had to edit them a little bit to, to, to fit the format. Uh, but that's usually, people are, like, terrified to get started with YouTube because of that reason, but just, like, go into, go into Figma, make a template that's a static asset, throw your video in that and just export that. Welcome to YouTube, right…

AI assessment note: “it's a totally different type of creative, right? The, uh, so if you look”

Q One of the things that is supposed to be a growth engine as well is referrals, referrals programs. Any lessons from Whisper, from Superhuman that are really important for me and founders to know on how to build great referrals engines?

A I think it's one of the, like, core things, like we talked about MetaGoogle lifecycle. When you think about PLG, like, have that referral program set up, and it be very front and center and easy to discover in the product to make it effortless for someone to share the product and get something, you know, for doing that, right? And, uh, at Superhuman, it was, you know, give one month, get one month. And, you know, we had, uh, we had people that had hundreds of referrals, Hundreds of months. They were never going to pay for the rest of their lives. Um, and because it was, it was delightful and it was easy to discover, but it was, it was very tactile where like I'm getting something right for, for doing this. Uh, so it was a similar thing with, uh, with whisper where, uh, when you hit your, uh, let's say 2000, uh, word limit. So now that, you know, we're, we're changing a little bit of, of how that trial works, but, uh, essentially finding the right moment to show the referral program And make it tangible. So like, you're right about, you're like, you're right about to hit your, your word limit. Hey, did you know that if you share this with your friend, you can get the next month free? Okay, I'm going to take five seconds and go and do that. And I'll tell one friend, hey, sign up for this. And if they sign up, I get that one month. And now I have another person that can spread the…

AI assessment note: “finding the right moment to show the referral program And make it tangible.”

Q have this hundred K, we've done three months, I, I, I'm really enjoying this, because as I said, for me, it's nice to follow the chronology. You have a hundred K, it's going well, I think. How do I know whether to pour fuel on the fire, keep the budget as it is, and just keep testing, but in a more measured fashion, or uh-uh, this is not fucking working?

A You want to chart your total spend to your acquisition and how elastic it is, right? There's always going to be a lag where, like, I spend today, and then I convert to paid in 14 days, right? Let's say, right? And that's like, we're looking at two main things, like revenue is the most important thing from an ARR standpoint, because we're in SaaS. As I spend up, does that ARR also spike, right? And how elastic that relationship is will immediately tell you, oil on the fire, or let's pump the brakes, because as we spend up the, you know, this is a very common thing that you'll hear, how incremental is your spend? It's a, it's a, it's a question you should ask relatively early, because it's, uh, once things are working, it's really easy just to, like, dial it up to 13 on a scale of 10. Sure. And what ends up happening, if you're not very diligent, all that additive spend could have not been there, and you would've gotten the same results. And then you're just giving money to Meta and Google, right? And there's that fine line. And that's why I'd say, like, true people that understand this in growth, Are on the rare side because like there's usually like I am a expert in meta or I'm an expert in some other channel, but being able to sit above the whole thing and say, okay, here's how the interrelationship of this much spend here, this many sponsorships, this many, you know, podcast …

AI assessment note: “You want to chart your total spend to your acquisition and how elastic it is”

Q Dude, I have said, I totally agree with you. You could take, like, landing experience. Give me three examples of the best landing experience in PLG in the last year. Uh, I, I, I completely agree with you. Um, If we go to the team itself, the requirements for talent changes. How has what you look for in talent changed in the last year or two?

A Yeah, I mean, I think it's, it's a complete one 80, right? Like, uh, I'd say, uh, now a year ago, I'd be looking for someone that has 10 years of experience in, let's say, meta ads. They've scaled to hundreds of millions of dollars, and they're, let's say, the best in the world as a Meta ads media buyer. And that would be the main thing I'd be looking for. Let's say that's the person I'm hiring. Now I'm looking for that, but they don't necessarily have to be the best in the world historically. If they're AI native or a systems thinker and can deconstruct what makes their job hum, that person plus experience will outcompete someone that just has experience and is not an AI native. And that's like, I'm seeing that happen where like, I'm going to have a hot take, Here, I think, um, a lot of teams are trying to make their teams AI native, and they're failing.

AI assessment note: “If they're AI native or a systems thinker and can deconstruct what makes their job hum”

Q Dude, I'm gonna love this show because most people say the same thing. And what everyone tells me is that, no, no, you should wait to do paid. You should wait to be paid. It's a dangerous drug that you can get hooked on. When is the right time, do you think?

A Right away. Right? So like at the end of the day, uh, paid is the easiest way to validate that you have PLG, that you have a product that can scale in any way, shape or form, right? Like there, there's a ton of narrative that like focus on brand, go, uh, organic content, build all that base. That just takes too much time, and yes, do that motion while also doing all of paid, because you can refine messaging, do creative testing, test your funnels, test your positioning, all within a week on paid. How long is that going to take you? If you do some content writing, you talk about it on a podcast, you see if it sticks, you're going to do some user interviews. Great. Do all of that, but then you have a much faster engine of validation on the other side on the paid amplification end.

AI assessment note: “Right away. Right? So like at the end of the day, uh, paid is”

Q We're going to go into the kind of team composition later. I want to stay on channels. So we now have A working program. How do we determine how much fire to fuel to pour on the fire?

A Yeah, I mean, I, I think it's, it's all dependent on, right, that elasticity that we defined. So like, we at Whisper did, what I think everyone should do, is go as hard as you absolutely can, and then see it blow up, and then understand where and how it blew up, pull back, and use that as information of now we know what is incremental, we know where our ceilings are, We know that we hit audience saturation or we need more creative. So we did that last year at whisper, right? Like we, we like five X the budget from one month to another. And where did the pieces fall through? Right. But the thing is, the reason why we did that instead of trying to figure it out over time is like, you don't have six months to say, okay, we need to be growing 30%, 40% month over month over month. And the only way that we can do that is, okay, let's see where our ceiling is. Cause maybe we would have been surprised that we would have been scaling infinitely. Which didn't happen, right? But we then knew, okay, meta needs to be second to Google. Here's why. Here's how all these things worked together. These newsletters were shit. This, you know, podcast doesn't work. How, how, how did the pieces fit together? Um, so, I mean, in my opinion, go hard to understand what doesn't work and where your mix needs to be adjusted, then bring things down and then scale back up with those learnings, and then you ca…

AI assessment note: “go as hard as you absolutely can, and then see it blow up”

Q How do you think about it? What does a systems thinker mean?

A So, to me, that's being able to step back from, let's say, A task that is a part of my job. Being able to take one degree of separation from the task itself and say, in my role, what are all of the moving pieces that essentially I need to do on a day to day basis? What are the interrelationships? Where's the boring admin? Where's the reporting? What's all of that? And being able to map that out and understand how your job as a system actually works, because then you can actually say, okay, I can apply an agent or AI to this part, and now I can focus on this higher leverage part of my job. Right? And a lot of people go and go to ChatGPT and ask it a question, and then they go on still doing their job the manual way. Right? And if you take a step back and say, what would it take for me to fully automate, let's say, all of my job? Of course, it's not going to work. Things are going to break apart. But run yourself through that thought experiment. And the thing is, is if you ask a lot of people that, it turns out they don't really understand their job. And because they're not systems thinkers, they don't actually know What are all the inputs and the outputs? Uh, so that, that's the line for me. So like when I, when I interview people to fill in the team, uh, it's, it's like partially like an engineering interview where like systems thinking is the bar.

AI assessment note: “being able to map that out and understand how your job as a system actually works”

Q When you look back at, say, superhuman, if we go back in time chronologically, It grew very, kind of, craftsman-like with founder referrals. What's your reflections or lessons from that experience on Superhuman's growth that you've taken with you?

A The, the taste making of what it means to create a strong product, uh, like, that for me was born while I was at, at Superhuman, uh, because at the end of the day, no matter whether, you know, if you've used Superhuman, you've likely stayed using Superhuman just because every little element of that product Is, is hyper refined to make you want to continue using it. Um, but I think the, the difference of what was the early scale of superhuman, where it was super founder led, word of mouth referral, uh, there was essentially a, a asymptote that, you know, the, the product hit where they had it, they had to follow a different playbook. And, uh, there was a, like, this was, you know, three years ago, roughly. Uh, and, uh, back then, you know, what was SAS now we call AI SAS. Uh, everyone was running more of the, Uh, you know, just PLG motion, referral, word of mouth, without doing paid, and they were laughing, like, you know, Ecom's been doing, you know, the, the UGC programs, hundreds of thousands of ads, like paid ad spend is the way to drive performance, and the resistance was definitely there while I was at Superhuman, uh, but, you know, seeing what, you know, I've been able to do alongside a great team at Whisper, and now Victor, uh, my philosophy is that the Ecom playbook Is the right playbook for SAS? Because if you look at e-comm, you have every single cent needs to equal a…

AI assessment note: “The, the taste making of what it means to create a strong product”

Q When you hit, when you hit audience fatigue, do you keep spending waiting for a next unlock? Would you just go, I've tapped out this channel?

A Well, you want to unlock the next audience, right? So it's, uh, who is, so like, you know, for example, uh, when, when we're looking at, at Victor, uh, we, uh, when we, when we launched, we, we purposely had the, you know, the AI employee for everyone. And we did this to see who came in the door and we had, you know, agencies, econ brands, and SMBs that first showed up. Uh, and essentially what we did is we went through in order, right? We went, okay, we have evergreen stuff. We'll continue that running, but now let's go agencies. Can we acquire them? What does the funnel look like? What does the creative look like? And then once we've let that, like once we unlocked one audience and it seems like the economics work, we can then, now there's a repeatable process to refine it. Then we can go on to the next audience. So you, you essentially want to, you'll hit audience fatigue in what you're currently targeting. You, like, marketing now is very ICP specific. Like, you, we, we have all the latest and greatest tools that can tell you, simulate as a pro, like, in a prompt, whether using Victor or whatever tool, simulate this person, like, what do they read? What do they think about decision making? Run that prompt, and then you have, like, the richest ICP research you possibly can have. Create, create it for that person, create an experience from a landing, onboarding, and product e…

AI assessment note: “Well, you want to unlock the next audience, right?”

Q We mentioned another one that was very interesting. You said AEO, which is obviously kind of, um, kind of answer engine optimization. We invested in peak, which essentially optimizes this. Um, how do you think about that as a new channel? How do we embrace it? What should we know?

A AEO in general followed the, like the core tenants of SEO in general, where, uh, what like used to be old school SEO, like Pumping out thousands of pages came back again, because like Google, because they own the, the SEO algorithm, essentially penalized people that were just pumping out pages for SEO. So like over the last couple of years, people stopped doing that. But the sheer amount of pages that you have in AEO, that is also changing a little bit, but like, you know, if you look at, um, you can, you can ask any, uh, agent, uh, to analyze the site map and essentially do like cron. So we do this with Victor. We look at every competitor and we look at their site map and we look at, uh, what, and like, we see how many pages they're making. All these companies are doing like a hundred to 200 pages a week. It's like a lot of it is just like generated AI slop. Um, so that's also what you don't want to do at scale. You actually, like, want valuable content that says valuable things because the, the AI crawler essentially ingests that, and then what you're saying there is what gets written about you when, like, what is the best tool for this, or like, I'm trying to do this, and it's, it's doing these citations. So the things that I'll say that are the most important to AAO now, it's like, yes, have an engine that you send content through your website on specific things, but the mo…

AI assessment note: “the most important thing is YouTube, Reddit, and like the, the social narrative.”

Q Can you even build a team on that?

A So, but that's the thing, right? Like, um, I think the team composition of, of today and tomorrow Is way leaner than it was before. So, like, we likely will, like, for, you know, Grammarly, you know, let's say, you know, bigger startup in SaaS. Startup is air quotes there. Big corporate SaaS. They have a influencer marketing team, right? They have, like, three, four, five people. Today and tomorrow, that is one really good person with an amazing suite of agents that will only ever touch, like, imagine you scale to A 100,000 influencers that are posting on you in a month. If you have an agentic system and you know how to weed through the data, understand where you need to be plugged in, you could do that. And that's why, even though it's one percent of candidates, It's fine. If I find that one person that is gonna be able to fully own that, build the systems, and, like, right now, uh, and this is, like, my bet for the next, you know, three, five years, you know, 80% of our work is still manual, 20% is agents. Hence why, like, my bet into Victor. They did it to, like, essentially solving all of knowledge work. I think in three years, companies will essentially be, like, board of directors, where 20% of the work is the strategy and the thinking, And then agents just do 80% of the execution, right? Like right now it's still like, we're still, you know, editing is manual, sending an…

AI assessment note: “the team composition of, of today and tomorrow Is way leaner than it was before.”

Q I'm not going to name companies so they don't want to shit on people, but like, if you look at some of the big providers, Are you seriously saying in your Anna plans and your Coopers and your, that anyone has a fucking clue what you're talking about?

A No, no, no. I mean, yeah, I mean, that's, that's also why I think the companies that do will quickly gain market share against those that don't, right? And that's why I think we live in very interesting times in terms of competitive environment where the incumbents that don't adopt this at the senior level Will, essentially, their growth rates will, will stand still while the hyper growers that are doing exactly this framing will essentially gain that market share and outcompete them, right? So I think you're right. Like, I think leadership needs to change. And like I said, I think a lot of teams, and my hot take was probably fire most of your marketing team that is not a systems thinker. Like, stop brute forcing people in. Hire the right people into the roles because the role has changed. The JD is no longer the same what it was a year ago, right? And, and, and people are still brute forcing people into these new JDs. I think there needs to be a bit of a shake up.

AI assessment note: “No, no, no. I mean, yeah, I mean, that's, that's also why”

Q When we actually break that down, If I'm an early stage founder, what should I actually do? How much money should I take to spend on page? Should I do it on one channel versus 10 channels? If I'm an early stage company, I don't have mega budget. What do I actually do?

A Yeah. So, I mean, I would say it's, it's definitely start with what I say is the, the core two. Uh, so like, uh, I say there's the core three of any acquisition engine. You have met a Google and life cycle, and obviously like your site and all the rest is, is in the middle, uh, because you have your, The strong video intent platforms, both in meta and like YouTube ads. You have the search intent of like people not knowing your product exists, but they're searching for something, right? Via the keywords. And then I add in life cycle as well, because if you don't have a net to like nudge people and show up in the right place, email, SMS, whatever it is, push if you have an app. Um, you want those three things spun up because you can scale to your first 1,000,010 million ARR, Just off of those three things. Um, and then obviously, you know, you're, you're investing into the site, you're investing into founder-led content. There's like the other things in there, but in paid, just focus on Meta and Google and you'll totally be fine. Everyone's like, oh, we have to do TikTok and Reddit. Like there comes a time and place to add in a million things because in the same way that like, uh, a lot of people get lost with agents is they like start doing a million things with agents and they do them all poorly. Same thing here, right? If you do a million channels and you do them all poorly, i…

AI assessment note: “in paid, just focus on Meta and Google and you'll totally be fine.”

Q How do I set that up? If we don't have it in SaaS, how do I set it up as a founder?

A This is what I essentially did both at Superhuman and Whisperflow, now at Victor, Uh, is essentially coming in and defining what that looks like, right? And, and it comes, uh, to have a strong analytics leader, uh, as well as like a strong, uh, analytics dev that can essentially set up all of that on your website, set up all of that inside of your product to make sure that you are, because most people stop at the, we have analytics to track our product. Are all the marketing platforms getting the same signals? There's like, there's a thing called maturate on like meta, and then there's like a, Enrichment score on, on Google. If you're not maxing that out, you're, you're like missing 50%. Like you're, it's just ghost people. Meta doesn't see them. It's not attributing conversions. And the way that the platforms work in today's environment is each platform's algorithm does what you tell it. So let's say we, we have poor conversion tracking, uh, that doesn't track when someone subscribes to our product. And it's like a 50 50 match, right? And if I optimize for that subscription, Uh, the higher volume amount of events I have, the more data it has of what a subscriber looks like, and when I optimize for that, Meta will give me more of those people, but if you have poor conversion tracking, Meta doesn't know who those people are, and it just randomly targets people, and you, you'll h…

AI assessment note: “have a strong analytics leader, uh, as well as like a strong, uh, analytics dev”

Q How does that change, then, with a whisper and a victor, where you do have a token cost being much more significant than you had in a SaaS world where there were no token costs?

A The biggest differences are first in that, like, Meta and Google are not optimized for, like, B to B SaaS usage-based products. Like, they go absolutely nuts with a, so you're telling me there's a predictive revenue of a one person could be a 50,000, you know, annual contract, while another person will be Right. There's just like huge variance depending on single person can be a high user and like the, the algorithm has no idea what to do with that. Um, but, but the other side of it is, um, we obviously have way higher average revenue per user, but, you know, average cost per user is also higher. Right. So you essentially want to make sure that you actually have a good grapple on what cost is. Uh, and a lot of people were like, oh, well, that's my anthropic bill, but what about modal? What about inference costs? What about all the other things? That is allowing you to service your products and, like, have a good finance leader that's gonna be able to essentially gut check, yeah, you've been missing these cost items, this is what you should be optimizing for so we don't burn money. And that's a lot, I think, a mistake a lot of people make is not doing that analysis at first.

AI assessment note: “we obviously have way higher average revenue per user, but, you know, average cost per user is also higher”

Q Um, okay, so we've got this hundred K, and we've got this UGC network. It's not really working. What does not really working look like? It's not doing the one-to-one. How long should I give it? Should we do this for three months straight? Should we do it for six months?

A Yeah. Yeah. I mean, you, you'll usually be able to say that like, if you just get started, the campaigns have to warm up. There are like certain things you have to have already in place. So like we said, you have your analytics, you have your conversion tracking, let it just run. And of course you, you already have to have customers coming in the door. And ideally that's why like, PLG and product excellence is still the beginning step, right? Like have a product that people genuinely love and they want to tell their friends about. If that's already happening, then your conversion tracking is working on your website or your app or whatever, and you get to 50 conversions, right? Then you can start paid because that's like the minimum line of whatever your conversion action is that the algorithm is going to say, okay, I have enough data to understand who your target You know, best customer is. Uh, and then I'd say within two to three weeks, if you launch a campaign and go through a couple, like you, you'll do two or three iteration cycles of we launched some creative, this worked, this didn't. Why? Ok, let's throw some diverse stuff, let's do some variations. You probably want to give it a full three months, but you'll, you'll be able to say, yes, I can start printing money today in like two to three weeks, or I have to go to the drawing board on all of these things, because there…

AI assessment note: “You probably want to give it a full three months”

Q I would fund that company today with millions of dollars. I'm being serious. Well, if you're JP Morgan or Goldman Sachs, and you put that at the bottom of Canary Wharf, would you not send every single employee there?

A Cause then you, you built the hyperscalers of tomorrow. I mean, what's, I mean, super interesting. Uh, so like I, I went to NYU, uh, for, for college. Uh, what's interesting. I did not take a single, I mean, zero marketing classes, and I learned everything via doing. And, uh, I think what's interesting is like the, the education around marketing, I think is also like still, and I'm like, again, I haven't been into a college today. I mean, maybe that has changed, but like, are they teaching you how to install a Facebook pixel? And no, right? Like the, The, the application, the, the, like, the, the framing of how it works comes after you start doing the execution, right? Like, you can do and then understand, verse, understand and sit in a room and learn, and then go and try to do the thing. You're gonna, you're gonna learn so much faster just by doing, so I, I agree with you. I, I've yet to see, like, true A to Z boot camp of what it means to be a growth leader, and I mean, like, Reforge and Brian Balfour, they, they've tried to do this, but, like, It's still-

AI assessment note: “so I, I agree with you. I, I've yet to see, like, true A to Z boot camp”

Q But my point is, are you able to do horizontal taglines, and then verticalize, and that's the right approach?

A I mean, I think because the category is so wide, right, where it's like, uh, and I wouldn't say we're the first, but we're the first to, like, say AI employee and actually, like, deliver on that promise, where, like, work is being given, and it's at, like, the usage is actually ROI, like, positive, actually valuable. And we started with that because it's, it's, it's again, how do you, like, how else are you going to frame it as being different than Claude or ChatGPT and these other players where, like, you have to, like, for us, it's the positioning around knowledge work. Uh, so it, like, there was nothing other than AI coworker and AI employee, like, there's no other positioning that, like, really makes sense. And because it hasn't been used a lot, I think it's confusing. But now that there are more use cases and stories to back it up, Then we can shift the framing to be a, I hired Victor and here's what he did. And then it's a, oh, it's a, you know, a guy's name. It's an employee. Oh, it's an AI tool, right? There's, you can invert the process now, but we had to start with something to essentially build those use cases.

AI assessment note: “we had to start with something to essentially build those use cases.”