LaunchPod - Michael Krafft === Michael: [00:00:00] We tested a new iteration of the page and yes, literally saw almost a 50% improvement in our conversion there. Which drove our CAC down substantially, which also helps us fuel growth because then we can reinvest that in the business and and keep kind of [00:00:15] fueling the growth of the company. Welcome to Launch Pod, the show from Log Rocket, where we sit down with top product and digital leaders. Today we're joined by Michael Kraft, CPO at Founders Card, a membership community for entrepreneurs and executives offering exclusive perks in VIP treatment. [00:00:30] Michael previously worked in investment banking, led product teams at American Express and played a key role in turning around a light. An HR tech firm ahead of its IPO. In this episode, we discuss how Michael uses AI to be a one person product team at a profitable startup. With over [00:00:45] 250,000 members, his AI driven discovery engine that automates interviews, segments, users, and turns feedback into roadmap, shaping insights. Why working as both product leader and IC lets him move from idea to prototype to product in record time with [00:01:00] tools like V zero and lovable, and the simple test that boosted paid conversions by 50%, and how he runs growth experiments far beyond his team size. So here's our episode with Michael Kraft. Jeff: Alright, Michael. Hey man, welcome to the show. Thank you for coming on. Michael: Thank you so much for [00:01:15] having me. Jeff: Yeah, I think this will be a fun one. You're an interesting kind of case study in polar opposites, right? You started out in iBanking and Finance, you went to business school. I think you're the first person I've met who, who went to business [00:01:30] school, ended up out of finance after starting there. At American Express, you moved into product. You've, you know, managed teams of 70 PMs, 35 PMs, all, you know, hundreds of people over your career. But I think the thing I really wanna dig into with you right now and start with [00:01:45] is you've kind of gone back to the roots a little bit and you're at a company called Founder's Card where you are, what? It's a 15 person company and you, you are the entire product team. Michael: That's right. I'm the, I'm a one person, chief Product officer, product team at Founder's Card.[00:02:00] Jeff: Yeah, I, I love this because you hear so many people talking about online. You know, this buzz of like, ha, have we seen already the inception of the first, you know, one person company that's going to be a billion dollar company, or all these wild claims about what AI is allowing [00:02:15] people to do? You guys aren't quite one person. I, I, I think that might be hyperbole, but like 15 people. I, you know, it sounds like you guys are doing a lot getting a lot of impact with, with a lean team. So give us like the TLDR, what does Founders Car do and what, what drew you in from this career? Kinda [00:02:30] large enterprises to come back and say, I'm gonna, I'm gonna do it myself, man. Michael: Sure. Founders' Card is a premium membership subscription where members get access to loyalty status. And really good preferred pricing, [00:02:45] really good rates on travel benefits, airline, hotels, car rentals as well as a suite of business benefits, lifestyle benefits. And so generally speaking, you know, we say entrepreneurs and those that are entrepreneurial minded are [00:03:00] attracted to our, to our product. This, this company, yes, it's a 15 person company very profitable and generally pretty. A pretty great growth company and as I was sort of looking at my options found, found this [00:03:15] company and I've, I've sort of been attracted to this type of company over the years where and I think I've been, been known for coming in and helping drive growth and also drive the modernization of a platform. And here was a company where I sort of deeply understood the business model. I had [00:03:30] spent 10 years at American Express, just sort of in an adjacent space. It spent time in, in FinTech and consumer tech and felt that even though the team was, was really lean, that I could come in and add a ton of value. And, you know, I didn't feel like I had to have [00:03:45] a team of 35 people or what, a 70 people in order to drive impact because I was confident in my AI skills, which I had sort of honed on on the side a little bit. And so, you know, that kind of was. Why I thought that this would [00:04:00] be a great opportunity for me. I joined about six months ago. Jeff: Yeah, I mean, I can definitely feel you on that. I've been at everything from, you know, one of the first 10 people. Early on, actually here at Log Rocket, up to, I've been at several thousand person companies running, you know, large [00:04:15] teams and in charge of huge initiatives. And what I liked about the small companies and, and that kind of, you know, right at the very beginning couple, you know, handful of people was you can really, you do something, you can see the impact of that in the overall results of the company. [00:04:30] But it was just a lot of grindy, kinda the mind numbing stuff that you had to do back then that you did because, you know, you had to, and you wanted to see the outcome, but. I think that's where, and I'm kind of jealous of where you're at right now is like with some of the gains in ai, it presents the opportunity [00:04:45] to, you can't offload all of that, but a lot of that can be sped up and you can do a lot more of the impact stuff. So maybe let's, let's just look at this a little bit, right? And like, it, it is a founder led company is scrappy startup. Chance to get [00:05:00] back to the basics a little bit here. , How did you come in and look at the company and say, here's where we need to apply, like this high leverage, but high skill level amount of kinda work as well. Yeah. Michael: Yeah, coming, coming into the [00:05:15] company. You know, I think as a, as a product leader, even though, you know, technically I'm an IC with no team as a product leader, you still have a responsibility to come in and help. You know, clarify and crystallize the vision and the strategy, the product strategy. So there [00:05:30] was a bunch of work that we needed to do there come in and also, you know, make sure that we were gathering as, as many customer insights as, as we could. And so, I think one of my initial priorities was around how do we turn on [00:05:45] the, turn on the, the engine or the flywheel for just. Kind of auto generating customer, customer insights and surveys. And then also just trying to clarify the roadmap. I think, I think the company had done a really great job of sort of smaller [00:06:00] optimizations around the, you know, acquisition funnel and on the product. But I think where they needed somebody to come in was to really help them figure out how do we drive stepwise change in their product experience. And their member experience. And I think [00:06:15] that's where, you know, in order to drive retention and, and, and acquisition and that's where coming in as a seasoned product leader, sort of, I was able to kind of know what the steps were to, to kind of drive that strategy and also some of the [00:06:30] execution. And as I know we're gonna talk about really use AI to kind of move very quickly on gathering a lot of that and putting together a lot of that. Those inputs myself. Jeff: That's the thread I think we don't wanna miss here is, [00:06:45] you know, it's when you come in as a senior leader early on like that, it's funny because , when you're one person team you have, you don't escape the responsibility of having to inform and help, you know, your [00:07:00] founder and, and the other leaders go forward and figure out the strategic ways that you are going to, you know, shift the company if they just wanted to ship a couple. Product things or they just want support in product, you know, that that founding team would hire the junior [00:07:15] PMs that we're all used to. But when you bring in a leader in, in any function, it really is, you know, you can't escape the strategic impact. You have to have the come in and the large, you know, macro step function impact you have to have. But then you also get to [00:07:30] kind of, you know, do the micro stuff and implement it and see the fruits of that labor. So let's start there. I guess you came in and like. The foundations of, you know, figuring out where you're going are just, how do you even listen? How do you do customer discovery? This guy, [00:07:45] aji Dewe, who's, you know, a great thought leader and product, he spoke at mind, the product recently, he's been on the podcast. He joins us around the country. In these networking dinners for product leaders we put on, has talked about. This kind of imbalance in speeds across teams. And [00:08:00] while engineering teams have greatly sped up from the help of ai, it seems like kind of discovery and customer listening and all that is still, we're still generating ways to speed that up a little bit more and kind of break this imbalance. So let's, let's start there. Like, how did you kind of come in and build this engine? Allowed you to [00:08:15] have meetings, meet customers, listen, synthesize, build, roadmap, all without that supporting team. Michael: Yeah, , it's a great question. I think you know, scarcity sort of drives the ingenuity because you know, I think [00:08:30] initially I spent a few weeks sort of doing a diagnostic of, you know, where, where we were and sort of where we wanted to be. Of course, , I have chat GPT as sort of my co-pilot going back and forth and helping to diagnose some of the issues. But. There was a huge opportunity at the [00:08:45] company to gather a lot more customer insights. And that was through interviews, NPS surveys. We, we didn't really have a lot of that. And so part of it was realizing that if I was going to do this customer [00:09:00] research in any sort of scalable way, I had to drive automation. It's sort of, you know, how do I use tools like. Calendly and HubSpot and, and AI to sort of craft surveys, to sort of auto set up interviews. So I, [00:09:15] I set up a ton of different customer interviews and I could test different levels of, do I pay them $50, 75, a hundred, you know, that kind of thing. And so I was able to sort of auto set up a lot of interviews with customers. I was writing surveys that, again, I was using a lot [00:09:30] of AI to generate those, those surveys and send those out. And. Ultimately, I, I ended up focusing on a group of users who were on our elite product, which is kind of the, the main product that we monetize [00:09:45] at, at Founder's Card. And the reason there was, I didn't wanna boil the ocean, but I knew that a major strategic priority for us was driving, you know, more acquisition and retention on our, on our main products. So I've focused on those users. We did not have at the time, a [00:10:00] great. You know, behavioral segmentation or customer personas. And so all the while that I was meeting with customers, I was making sure that I was capturing those transcripts, cranking those through chat GBTI was gathering survey, survey data. [00:10:15] I was looking at Trustpilot reviews, Google reviews, and using AI to sort of pull in all of these disparate data sources and, and the interviews to construct a behavioral segmentation of these users that we had. And then through, you know, going back and [00:10:30] forth using AI a lot, that behavioral segmentation, we knew kind of the needs and pain points of these different segments that we had and that could inform, you know, opportunities and roadmap. And again, it's sort of me going back and forth with, with the AI [00:10:45] tools to sort of figure out what is going to drive the most impact prioritization and sort of ultimately turn that into the, the roadmap and things that we're building, which now has turned into features that we, that we're building. Jeff: Right. So, am I right in kinda guessing [00:11:00] it started with some surveys that you sent out to target members as they responded and, and you answered those, they could book time for further feedback with Calendly, or was there more than that there? Michael: A hundred percent. And we, you know, I had an idea of [00:11:15] creating what we call the member advisory circle. So one of the last questions in every survey, every NPS, we set up a, you know, A-A-N-P-S survey that would go out to all members after 30 days. And I added a question of, would you like to join our member advisory circle? [00:11:30] And so a lot of people opted into that, which was great because then I had a group of people who didn't mind getting emailed about research and surveys because. Like many subscription businesses, founders card, you know, there's this delicate balance when it comes to emailing your [00:11:45] members. If you email them too much, they might unsubscribe. You know, they may, they may cancel their, their, their service. So identifying users who were keen on giving us feedback was, was really a, a great insight that we kind of had. And now I've [00:12:00] built a set of about 200 users and always, always growing. Jeff: That's a lot. Michael: Yeah, well, right, but it's not, we, our product is a really good product. I think it's a little bit of a some people have challenges with a little [00:12:15] bit of benefit discovery. Our strength is also our Achilles heel. We have so many benefits. That sometimes people maybe don't sort of see what all of those are. But we have people that are very passionate about our product, which is great. And something that when I think about joining a new company is, is [00:12:30] this a product that many people already love and that you have product market fit for? So identifying those users who can, you know, provide good feedback and don't mind getting emailed, you know, every couple weeks with a survey or an interview request is, is amazing. And the [00:12:45] fact that we have these NPS surveys that go out every 30 days. And ask this question, would you like to join our member Advisory circle is a way that I'm constantly adding to this list of people that are want to participate and give us feedback. Jeff: I mean, I think two things there. One is [00:13:00] that's always been a huge red flag for me is if I'm looking at a new company and I can't find evidence of users that, that really love the product, or, you know, they can't even, at least talk to me about and maybe help me, help me introduce me to a [00:13:15] couple. Users or customers who are really just infatuated with and, and love what they're doing, that's kinda red flag number one there. Unless I think I can come in and like really influence the product itself. That, that's, that's worrying to me. But what I think is really cool here is everyone I think looks at [00:13:30] these, you know, how do I start using AI to magnify my impact? And they think you have to have it through every step and like, oh, I gotta start with AI to, no, like you are, you know, NPS has been around. Forever. Like, or, you know, for [00:13:45] a long time HubSpot and, and you know, sending out those emails and surveys and allowing people to book time in Calendly while Cal, while I love Calendly it is not some new, you know, thing that is really disrupting our workflows right now. [00:14:00] What kind of the magic happens after that? Right? Like you can get, because you have these users who love the product, you can pull them in. But the problem becomes , if you talk to 200 people every year, you know, let's say you talk to each of them or you talk to half of 'em, even a hundred, [00:14:15] like, you're not gonna be able to synthesize and pull data from all of that and regularly go back and pull insights when you need it or aggregate it. Well, but the fuel, there's no AI magic and getting the fuel there. It's what you do with it once you have it. Michael: Yeah. And, and it's just, [00:14:30] it's I mean, there's AI and then there's just simple automation, right? And I think in a prior role, I would've said, okay, well I need to have my UX researcher and the UX researcher's gonna reach out to people. And it's like, no, you, you can do this all yourself. And I think you know, one of the things that I think about [00:14:45] is, you know, I've led some decent sized product and design teams. And you know, now technically yes, I am an an ic, but. Does that mean that I'm sort of functioning at like a junior PM level? Definitely not. I think the fact [00:15:00] that, you know, I do think quite broadly about, you know, how do I drive impact? I understand the drivers of the business. I understand what, what we need to do to drive features that our users will, will love. And kind of connecting all of those dots, it makes [00:15:15] me quite effective and efficient, like using these different tools. To, you know, help us, help us get where we want to go in a pretty efficient way. It's just, it's just not necessarily how I've operated over the last, you know, five or seven years I've been [00:15:30] a product VP at a bunch of different companies, like, like Amex and, and Alight. Jeff: Right. And like, you know, at Amex, you're, you're gonna have a team, right? I, I don't think anyone's saying you're gonna have a company the size of Amex with a single VP of product, and that's it. But I, I, I do think it'll be interesting over the next, you [00:15:45] know, several years to see how far this model can be pushed and, and some of that stuff that you would've historically just gone, all right, well, you know, in my case I'm gonna have some, like, you know, entry-level marketer reach out or configure something to send. Like the light work is configuring [00:16:00] to send a, a regular cadence of surveys or the NPS is what you do with it. And, and when you have that interview, you can probably pull so much more on your, so like, okay, so we, we are at the point where we've pulled, you know, you've talked to these people, you've gotten recordings, I assume, you [00:16:15] know, where does it kinda go from there to turn, you know, that dozens of conversations into actionable outputs that can impact and, and drive founder's card forward. Michael: Yeah, and I think I, I kind of hinted a little bit at this before where you, you [00:16:30] are taking all of this disparate data, which includes all needs and pain points and. And, you know, as, as you can imagine, the AI is quite good, even if you're just dumping in Trustpilot reviews, sort of understanding who people are. And people do provide some [00:16:45] pretty lengthy comments, especially in our NPS surveys, when you give them kind of freeform text fields that you know, some people will write some nice things and some people write some, some, not some nice things. But I think the key to having the behavioral segmentation is that allows you to make sure that you're targeting.[00:17:00] Your features and your product strategy around the users that you care about most and who are gonna be the highest value for you as, as a company. So for us, those are generally people who are interested in, you know, premium [00:17:15] luxury travel. Generally, who are are small business owners themselves, but who can sort of take advantage of the whole, set of products that we offer. And then, you know, you do have a lot of people who kind of sign up and then they maybe a trite because they were [00:17:30] only there for one benefit and it wasn't as good as they thought. Or they say, oh, I had it on my, my credit card. So the point of bringing up all of this with the behavioral segmentation, make sure you're focusing on the right set of users and then. Understanding their needs and pain points and then solving their [00:17:45] problems. And so, you know, the AI is very good. Obviously you're gonna be thinking about how to solve their problems yourself, but the AI is also very good at prompting you for, you know, lists of ideas of things that you can do. Especially if you are identifying those [00:18:00] needs and pain points, not just listening to what they're saying, but really kind of solving their problem maybe with something that they didn't think about. And I'll give you one example. So, at a company like Founder's Card. Benefit Discovery is one of the biggest challenges that people [00:18:15] have. So they come onto our site. We have, you know, over 500 benefits, how it's not reasonable to think that somebody's gonna sit there and just sit here with all the filters and understand all the benefits. So one of the things that we are building, which is our first real AI feature. Is an [00:18:30] AI concierge, and this AI concierge is a conversational ai. So imagine having chat GPT sitting at the top of your, you know, at your page when, once you've logged in and it prompts you for things that you can do. For example, do you wanna enroll in your [00:18:45] elite tier benefits? Do you wanna, you know, get a present for your wife for Mother's Day? Do you wanna travel to Sydney, Australia? And so by providing that sort of conversational AI interface, we believe that users will be able to much [00:19:00] more easily sort of digest and understand and access the, the benefits that we offer. And again, nobody asked for that specifically, but that came from the needs and pain points analysis of sort of, you guys have so much stuff, I don't know how to deal with it, and sort of. I think the AI can sort of help you in [00:19:15] some of your, your brainstorming there. So that is sort of the, the main feature that I'm working on launching in the next sort of four to six weeks. Jeff: And that's where I think, right, like the, the main goal of products, and I've said this over and over again, is not [00:19:30] creating software. It's what are the user's problems and how are you solving them? And in this case, you know, you, you guys have found cards, have a lot of different. Ways to solve problems because it's a lot of functionality. So how do you kinda help people discover that and you know, [00:19:45] looking back to, you have the surveys, you have feedback, you have feedback interviews you did. One thing I've found kind of, you know, chat GT is, is my choice of tool, but you can use any of the LLMs is really good at, is kinda you can feed in 20, 30, 50 of those interviews. And [00:20:00] ask it to output, you know, aggregate the similar problems or like kinda group them into general, general things and you start to see, you know, these groupings. You go, oh, maybe like this person and this person actually talked about different things. Let's break those out. Or, Hey, these kind of problems are [00:20:15] actually really similar. Can we aggregate them? And just so much faster you get to these kind of macro problems and you go, all right, well, how do we solve? Right, people figuring out how they want to use all the stuff we offer them in this case, or, you know, on our end we were going [00:20:30] through and, and launching our Galileo AI platform back in, back in spring. And as we were trying to figure out what areas do we focus on as a value, how do we phrase it? Like, what's the problem? That. How do people talk about the problem that we, that we're solving? Like what's their [00:20:45] phraseology for it? We were able to kinda go, oh, it's not, you know, it's not this other way of looking at it. It's that they see that they have a ton of tools that are really hard to go across and understand kind of how all of that impacts the digital [00:21:00] experience they're delivering. 'cause it just takes hours to aggregate all this data across these things. That's how we point this as, or that's how we kind of position it as it's not. Some other, it does the same thing, but it also gets into how you talk about your benefits, right? Or in your case, how do you communicate them [00:21:15] to your members? It's what do you wanna do? And then showing them how they can best Michael: A hundred percent. In our case there is a, there is a question of, of understanding from users. What do you do when you come to our, our site? And I [00:21:30] think having that good understanding has helped sort of us ideate again, understanding their needs. Delivering against those, rather than just doing kind of what they say, Hey. 'cause people will tell you you guys need to add more benefits or, you know, [00:21:45] something like that. Why, you know, why'd you lose that benefit? And really I think you can solve the problem that they're asking for, or you sort of understand what they need and sort of solve it in, in, in a different way. Jeff: Yeah, and kinda the neat thing here, [00:22:00] Ray, is, and I think we alluded this briefly, but like it kinda takes a full cycle, is the fact that you were the one kind of hands-on throughout all of this. You know, like you said, it doesn't erase. Your VP skills and suddenly you're just doing like [00:22:15] menial, you know, the entry level work. It rather you're able to apply that level of craft to all this and probably get to a better synthesis of what are the things we need to do a lot faster than having to kind of have a 10 person team where you're all kind of going through and, and you have some reporting up to [00:22:30] you the results from interviews and you kind of, you know, help them synthesize it or, or figure out what you're gonna do from all this aggregated data. You're in the weeds and up high and able to just do that faster and so. Presents a a probably better, faster outcome here where you can, where you [00:22:45] can really move fast and this is why you tend to see, I feel like these kinda small up share companies moving so incredibly quickly and really just nailing what is gonna move the needle with the customers. 'cause like you are the one in there and you have just so much more context now. Yeah. Michael: I [00:23:00] a hundred, a hundred percent could not have said it better myself. The fact is because I'm the one aggregating all of this customer info, I'm also the one, you know, conducting a lot of the interviews. I feel closer to the customer than I've ever been, or at least that I can think of in a [00:23:15] long time. And then using AI to sort of create vision documents for a feature that I want using AI prototyping tools to actually come up with what this would look like. So I'll use V zero. Or lovable. I'll upload, you [00:23:30] know, pictures of our, of our website, screenshots of our website. Say, okay, I wanna add this, this feature at the top in the AI concierge, and then I can put that in my vision document , and that turns into a PRD and all of this [00:23:45] stuff that would've taken me so long or that I would've been relying on other people. We would've had a zillion meetings and, oh, let's review it. Let's change it. We can move really fast. And you know, it doesn't mean that, it doesn't mean that we get it right. And I, I don't think that I'm the single source of truth for [00:24:00] like, what we should do. So maybe you do lose a little bit in sort of the, the lack of sort of diverse opinions because you sort of are trusting one person to sort of drive [00:24:15] the, the insights and the solutioning, you know? But I do feel like. Because I have, for example, chat, GPT and an AI prototyping tool, like as my thought partner. And I spend a lot of time and I actually sit and we'll talk to chat GPTI turn on [00:24:30] like the dictation and I spend a lot of time just talking to what people will kind of laugh at me and say like, who are, who are you talking to? But now in my office, they know that I'm talking to chat GPT and go back and forth. So I don't it, it's not perfect, but. [00:24:45] I can move really fast and not spend tons of times in ti time in alignment meetings because I can move very quickly from customer insight to an actual description of the solution. Jeff: And one place I've found that kinda replaces or, [00:25:00] or helps augment that, the risk around, you know, maybe one train of thought is 'cause I, I've made a big push as well to have a lot more customer conversations and. And talk a lot more in the market with people. And there I've been able to kinda bring up all, you know, here's kinda how we're looking at [00:25:15] doing it. What do you think about this? Is this gonna work? Is this interesting? But very quickly, people, you know, people are very happy to engage with you on like, oh, it should actually be like this or that. And you kinda push back on them and, and you know, a lot of those use cases I've been able to, you know, maybe [00:25:30] replace the robust internal discussion. What is probably just as good is, is talking to the actual market to go, oh, you don't like that? Why not? But I, you know, but I think other people will like it because of this. And oftentimes people will go, ah, yeah, you know, this part of the market's gonna like it. This part's not. And you [00:25:45] can get there pretty quickly that way as well, Michael: you don't have to talk to a million people. You come up with a solution and put in front of five people and you'll probably get three different responses and that'll probably be good enough to give you sort of a go no go or help you if you need to sort of pivot your design a little bit. , and [00:26:00] we've done some of that and we've done some, you know, I, I built a quant survey in SurveyMonkey that, took me probably longer than, than I needed to. But you know, getting that, that feedback is, is really important before you go and, and build something. Jeff: Yeah. It's interesting. We had a woman named Aloe Ucky. [00:26:15] On who's she's SVP of Product over at US Bank, and she has this theory around what she calls the rule of five, which basically it is a step function increase, you know, kind of practice of, of talking to customers. But her first step is always kind of talk to five [00:26:30] people. And if you see kind of across five people, a really heavy preference for all the same thing or like. A strong opinion about one thing that kind of spans with people. You can feel pretty confident that you have high signal that that's a really [00:26:45] good way. And now you kind of narrow your focus on the next group of people you talk to. If you talk to five people and they all say different things, you, you gotta go back to the drawing board, man. Like you're, you're still off base here. So, it's interesting to kind of see all this practice from. Really smart people who I've talked to [00:27:00] kind of coalesce around some of these things to, to how do you do this? Makes a lot of sense. So I do wanna kind of get a little bit more tactical here. As much as we talked about, you have really, you know, big impact. And, and I, I do love the AI kind of chat discovery. 'cause I think that's a big [00:27:15] large piece about how do you kinda move the, the breadth of the company forward. But you and I got to sync up a little bit before we actually jumped on. And there was this test you were able to run. It lifted like 50% conversion rate on a key revenue page. And I think that's another thing is [00:27:30] the, everyone's still always a sucker for a good growth story. 50 percent's a lot, like 5%. Usually people are happy. About 20% people are stoked. About 50 percent's ridiculous. So let, let's do it. What's this one? Michael: This, this one, it still, still makes me laugh. I, I feel like there was a giant [00:27:45] element of luck in this, but, Jeff: it usually is for 50%, Michael: Right. And, and I'll tell you, this page was our core monetization page where when somebody signs up for a trial, you know, we show them a really strong offer. But it's kind of [00:28:00] where somebody can make a decision whether they want to go straight to the paid product or if they wanna just continue on the free trial. The paid product gives them, you know, the, the value for what we charge is outstanding. And so there's sort of this decision point [00:28:15] for. For our users. And this page had been tested 15 different ways with sort of modern designs, but the page had sort of been kind of stagnant for about seven or eight years before. And I think, you know, we, we had an objective of, you know, driving [00:28:30] more conversion on the page and also doing more of like a UI uplift. Because the, the page design was a, was a little, was a little dated. I spent a good amount of time in V zero just sort of testing out different, different ways , of running this page. [00:28:45] And also was using chat GBT to sort of. Think about what are the elements that might be most important in a user's decision at this critical point where they're deciding like, do I wanna pay money for more or just keep going? And I think some of the insights that, that we generated were you know, [00:29:00] people care a lot about what other people think. So testimonials, which, you know, I always had like, love hate relationship with testimonials because you sort of are like, eh, these are very like, curated. But for a product like ours where the value may not be. Completely clear [00:29:15] to somebody upfront because it's a lot more about like, Hey, sign up and then you're gonna be able to save money here, here and here. We actually pushed some testimonials further up the page that really spoke to the value of the product and how people could save, you know, hundreds or thousands of [00:29:30] dollars. We did do sort of a UI refresh and then. Through, through that process, we tested a new iteration of the page and yes, literally saw almost a 50% improvement in our in our conversion there, which drove [00:29:45] our CAC down substantially. Which also helps us fuel growth because then we can reinvest that in the business and, and keep kind of fueling the growth of the company. I mean really, really one that I'm still excited about. We've, we've had a lot of testing success. Using AI [00:30:00] and using these prototyping tools to help kind of generate ideas and test things out. So that was, that was a fun one. Jeff: We had years ago, a large blog for front end engineers and we were testing a bunch of kind of conversion [00:30:15] optimization stuff and we kinda lucked into a similar thing where we were, we were trying some different creatives and finally our growth leader said, you know what, we should just for. For the sake of completeness in doing a real test here, we should have a null version where we remove this kind of [00:30:30] additional CTA that has just always been on the site. And oddly enough, the cleaner one without that CTA on it did the best. We still had another one kinda at the top and one at the bottom, but that mid one, it turns out, was actually really negatively impacting somehow conversion. We, we re, we [00:30:45] reran it like three times to make sure in every single time. It was the same in kind of where we finally landed was maybe a little bit similar to the, you know, thinking hypothesis of, of why it might work here. Is doing that made it feel more authentic when [00:31:00] people actually hit, you know, where the CTA should be. And developers can be a little touchy about that. Maybe it seemed a little bit more natural or whatever it was, but it's always funny to find those little locked in huge gains. Michael: So one of the ways that I think I've been scalable in, in our [00:31:15] testing program, because I think anybody that's done a ton of testing knows that that can be very time consuming. You're sort of generating hypotheses. You're coming up with designs, and some of them are gonna be, you know, CTA tests, hero banner tests, you know, reorganizing the order of the page. One of the decisions that [00:31:30] I made after I'd been there for about a month is to hire a like a specialized conversion optimization firm. So we hired one of those that, you know, they charge you sort of a monthly fee but they give you a little bit of like arms and legs to sort of help you on [00:31:45] ideation. That was something that I, I hadn't done that type of thing before, but has I found, been very powerful sort of helping me. From a scalability perspective, drive, you know, more impact and increase our test velocity. You know, we're launching a [00:32:00] couple tests a week, which, you know, may not sound like a lot to certain people, but when you're going from, you know, one a month to, you know, a couple a week, like you, you do need , a little extra leverage. So that has been a powerful a ad as well. Jeff: I mean, if you can find the right [00:32:15] agency or the right kind of partnership there, that can be really high leverage. If you're not, you know, hiring on dedicated people, especially if you don't, maybe don't have a full person to work. Worth of, of stuff there. It's a good way to kind of like get a lot of that done. But , to the point of you can be in the weeds, but maybe there's some things you don't need to be [00:32:30] that in the weeds on like actually building the, the test iterations or something like that. Okay. I do, you know, we've talked about this a lot, but I, I think it is important to recognize like, this is not, you know, some background of yours where you have, have been at these tiny little companies [00:32:45] over and over and over again and you just have, you know, this huge background in. Being the, the guy or being the person who kind of comes in and, you know, leads vision as well as getting it done, which I think some people kinda look at and go, well, that's, that's the model I like. And people often [00:33:00] feel like I do big companies or I do small companies. But you, I mean, you were at some of the most, you know, Amex, which is one of the most massive companies you can be at. You were at a company be called Alight. And this is actually interesting because, you know, it's kinda the contrast to small companies is you were [00:33:15] brought in at a light. To kinda help them through the IPO process, which is the antithesis of a 14, 15 person company and kinda help them get, get on track and, and really get ready for a successful run here. So let's, let's just run through the quick version of that story real Michael: Yeah. So, yeah, [00:33:30] I'm not all a hands-on tactical product leader. I, I do, I do have Jeff: You can do pure vision sometimes. Michael: And, , I had been at Amex for 10 years. Made to vp, but I think I was looking for a new challenge. And Alight, which is in the [00:33:45] HR tech space was a really interesting challenge. They, they had tried to go public maybe six months before they'd filed their S one, maybe six months before I got there. They did not get the valuation that they were looking for. They were sort of being viewed as a business [00:34:00] process outsourcing company instead of like a SaaS tech company. And so they were very keen to invest in their, in their technology. And so they hired me to run their digital product team. And, and so I was there to kind of help them build out their digital product strategy. [00:34:15] Didn't really have a super clear vision at the time. They were having some a RR compression from their, their different clients. And so my mission was sort of develop that product strategy, the roadmap. And try and, you know, get them to this, this IPO. [00:34:30] And so essentially what I did was I led sort of cross-functional work stream with a lot of the senior leadership at the company. It was also during COVID. So we were sort of sitting in rooms very far apart from each other and trying to align on a, on a strategy. And, you know, I think. [00:34:45] Part of it was identifying what wasn't working. So the mobile app was sort of an underutilized resource, so we really focused there. We also focused on, on AI and how do we drive more personalization to end users. And, you know, through a lot of, you know, regular product [00:35:00] processes around discovery and design and testing. We actually built a kind of a new face of their digital platform. That was called the a Alight Worklife platform. And not only that, I was heavily [00:35:15] involved in pivoting their go-to market strategy, which had been aligned around their business units health, wealth and payroll, and turned their kind of go to market positioning much more about delivering employee engagement and value through the tech platform. And [00:35:30] so, you know, we did a lot of testing with clients around what this value proposition would look like. And ultimately through a lot of the efforts of our team, we, we were able to, kind of stem, this a RR loss that we were having. We won a few, you know, [00:35:45] giant new clients, which hadn't been the case so much prior. And that a light work life platform that I was kind of the, the architect for, ended up being sort of the centerpiece of the IPO. And so that kind of became a lot of our messaging through the IPO. [00:36:00] And then we, we did go public back in, in 2021, so that was a very. Proud moment for me. Jeff: it's interesting 'cause a lot of companies will often. See themselves as software companies. And in reality, you know, if you are, you can be software assisted services, you can, you know, but [00:36:15] that, that's the risk here, right? Is, is software companies because of, you know, the ability to have pretty high margin, you know, almost zero opportunity cost for adding new licenses aside from just whatever the compute cost of. If your customer's using it it's very different than any kinda services [00:36:30] org or outsourcing where you were kind of bound by human capacity to do work while you, while you offer revenue driving subscriptions. And so, you know, here that ended up kind of biting them in the butt a little bit where the multiplier [00:36:45] was just not there. And what I think is cool here is you guys were doing, you know. Doing AI before AI was cool. A lot of this was around how do you answer questions automatically? How do you automate, how do you create better experiences for employees by [00:37:00] having a place where they can get these answers and without, you know, having to have humans who are, who are scale bound to do that. Michael: Yeah, I, I will say and shout out to Ibrahim, Kuri from my team over there. We had, we had a phenomenal chat bot. But even that [00:37:15] chat bot was very underutilized, but that was sort of the old school model of ai, which, you know, you had to like script all of your answers and there was a ton of manual testing and looking at what you know. But we had, we had a very good success rate on, on the sort of correctness of the [00:37:30] answers. It was over 90%, but yes, we. That was part of it. There was also just personalizing communications to end users. There was a big aspect of using AI to make recommendations to employers on [00:37:45] changes they might wanna make in terms of how they manage their company benefits. Because we had this giant data set, A light was, and I, I presume still is, serves about. Two thirds of the Fortune 100 companies. So they have this very broad [00:38:00] data set. They have 30 million end users on the platform in terms of how other companies are offering benefits and patterns on how much leave people take and salaries and all this kind of stuff that we were able to sort of take some of that AI and power that, in terms of. [00:38:15] Recommendations to employers. That made our offering a lot more high value than kind of, yes, we will just be a connector for all of your kind of company hr portals, which is sort of a little bit kind of, I think, how we were [00:38:30] perceived before. Jeff: right. And, and you know, 2021 went public via spac in 22. Revenue grew pretty heavily. You know, you were at you know, billions of dollars already and grew almost 10% to, to over three. Then just like 2023 grew again. So like [00:38:45] a good series. It's not like it was an IPO that suddenly just like, you know, didn't work out. There was great growth afterwards and the, you know, stock price grew. There was increasing revenue. The segment was a good one. So it seems like it did really well. All right. I, you know, as much as I love to talk about [00:39:00] ai, as much as I love to talk about how people have developed these cool companies and really, you know, done interesting things, I do have to give you part of your day back here. Michael, it was nice of you to join. Michael: Really love the conversation. Jeff: I, I think, you know, typically what I'll say is you know, if [00:39:15] people have questions that people think they can either be helpful to you, especially over, you know, over at Founder's Card or if they have questions for you. Is LinkedIn the best place to reach out? Is there another Michael: Yeah, absolutely. LinkedIn. LinkedIn is great. So yeah, please reach out. We are, we are hiring, we've got a a pm [00:39:30] role and a couple design roles at the moment. So, and I think that trend will continue. So yeah, shoot me a note on LinkedIn. Jeff: Nice. And, and so you are hiring. So PMs check it out. High impact work, but you know, sorry to say, Michael, it seems like. You guys are [00:39:45] not going to be the first single person billion dollar company. You pass that up now, but Michael: Yeah. You know, I, I was, I was Jeff: I dunno if I'd wanna work at a bi single person, billion dollar company though. Like that seems lonely. Michael: Well, you know, to me, I, I care a lot about company culture [00:40:00] and comradery and sort of building things together and pressure testing my ideas and debating what we should do. And I feel like that will be hard at a, you know, I, I, yeah, at a, at a one person billion dollar company you know, kudos to anybody that can do that. [00:40:15] But that may not be, that may not be for Jeff: Yeah, I like lean, but that seems a little too lean, so. All right, man. Well, it's good to have you on the show. Thanks for joining us, Michael, and keep in touch. Awesome. Talk soon. Michael: Okay, bye.