CHURN FM

E263 | From Optimizely to Limitless: Dan Siroker on Building Products That Wow

Andrew Michael Episode 263

Today on the show we have Dan Siroker, CEO and Co-Founder of Limitless.

In this episode, Dan shares his journey from founding Optimizely, growing it to $120 million in ARR, to now leading Limitless, a company focused on building products that enhance memory.

We discussed the challenges of transitioning from B2B to B2C, including how AI is reshaping consumer retention and the importance of product focus and execution.

Finally, we wrapped up by diving into Dan’s philosophy on building products that truly "wow" and how startups can achieve that through ruthless prioritization and attention to detail.

Mentioned Resources


Churn FM is sponsored by Vitally, the all-in-one Customer Success Platform.

[00:00:00] Dan Siroker: And I actually think one of the mistakes I made as CEO was I wasn't decisive enough in terms of focus. Now, I'm a big believer in the saying, "The main thing is the main thing, should stay the main thing." Instead, back in Optimizely, ten years ago, I was making up for lack of focus by just sheer hours. I was just working more, and in hindsight, again, in hindsight, the vast majority of the things I spent my time on, our company spent our time on, didn't matter. It's only really a handful of things that we did exceptionally well that actually made the difference. And I think, had I been more ruthless in our focus, both my time and the company's focus, we could've done a few things much, much better.

[00:00:36] VO: How do you build a habit for your product? You need to invest in it. And we saw these different… You don't just gun for revenue in the door. 

[00:00:43] Andrew Michael: This is Churn.fm, the podcast for subscription economy pros. Each week, we hear how the world's fastest growing companies are tackling churn and using retention to fuel their growth.

[00:00:56] VO: How do you build a habit forming product? We crossed over that magic threshold to negative churn. You need to invest in customer success. It always comes down to retention and engagement. Completely boosts the strategy, profitable and growing. 

[00:01:10] Andrew Michael: Strategies, tactics, and ideas brought together to help your business thrive in the subscription economy. I'm your host, Andrew Michael, and here's today's episode. 

[00:01:19] Andrew Michael: Alright. Hey, Dan, welcome to the show. 

[00:01:24] Dan Siroker: Thanks for having me, Andrew.

[00:01:25] Andrew Michael: It's great to have you. For the listeners, Dan is the CEO and co-founder of Limitless, a personalized AI powered by what you've seen, said, or heard. Prior to Limitless, Dan was the CEO and co-founder of Optimizely, which he grew to 120 million in ARR and 450 employees. So my first question for you, Dan, is who were you meeting with three weeks ago on Monday at 10 AM, and what was the summary of the meeting?

[00:01:52] Dan Siroker: I know exactly that answer, and I have this great app that helps me remember. Three weeks ago was a long time. But, three weeks ago I was in a product review. That's a great question because it's one of those things, memory is one of those things, you don't know what you don't know, and you don't realize how much you forget. A lot of people don't, some people admit, okay, I have bad memory, but even people with good memory forget 90% of what happens after just one week. It follows what's called the forgetting curve. Not only that, but as we get older, our memory gets worse. Typically, your memory peaks when you're 20 years old, and then every year thereafter, your memory gets worse. So that was one of the key insights that inspired us to start the company. It's like, how do we build a tool that's the equivalent of maybe a hearing aid for hearing or glasses for vision, but for memory instead.

[00:02:36] Andrew Michael: Yeah, I love the premise as well, and as I mentioned [for the call act], just miss out investing in the last round just due to timing, mostly on my part, but, personally, I love the notion of it as well because my family suffers a bit from Alzheimer's and from dementia. So my grandmother had it, now my aunt has it. I think I have the marker as well of it. So yeah, I think it is what it is. But I think it's interesting to sort of see like now how we can start to leverage technology in different ways, and maybe this may not be it, but it's like a step in the direction of us exploring what's possible with technology today. So really excited about what you're building at Limitless.

[00:03:11] Andrew Michael: The thing we wanna chat about today, obviously, is like general attention and Limitless, in itself, bringing you unlimited memory and trying to be that backup for you. You have direct experience from a B2B and a B2C perspective, like coming from Optimizely, purely B2B. I have some stories, all that I'd love to touch on there and get your feedback from others, but what has been the biggest challenge for you coming from that B2B background into a B2C environment now when it comes to general attention?

[00:03:41] Dan Siroker: I would say, the bigger challenge is that it's harder. I think its, it feel like startup on hard mode versus startup on easy mode. Not that optimizing was easy, but at the end of the day in a B2B company, certainly B2B SaaS company, if worse comes to worst, you just pick up the phone and you do the sales resolve as the founder. When you're going to consumer, you can't do that one by one. You have to build a product that to some extent, sells itself. You have to build channels and a sort of growth engine that sustains itself beyond just hiring a human being to be the glue between the customer and your product.

[00:04:12] Dan Siroker: And that's probably the biggest gap, is the role of sales. I actually like the challenge cause it forces us. It's sort of more pure. It forces us to build a product that can get away with better marketing or positioning or sort of overpromising. It really has to deliver for people to actually continue to use it. And that's been, I think the nice thing is even though it's on hard mode, once you figure it out and start of building that well, you now have seven billion potential customers on this planet. As opposed to, you're not limited by, for example, how many sales rep you can hire and how much quota they can fill. You're more limited by your ability to go out and capture the market as a market being the whole world.

[00:04:49] Andrew Michael: Yeah. I saw somewhere recently as well, like somebody was discussing and debating the difference between SLG and PLG and what's better and what's worse. And I think ultimately, like any company at sort of any scale, they end up doing both. But the argument was that sales-led growth is easier to get started with than product-led growth. And I think it's, in B2B, it's the same thing as to the problems that you mentioned today. It's that you can get by with a good salesperson being able to sell things, but in product-led growth, you really need to have, like, poly engines running, optimizing activation, and all of that to efficiently be able to build that business and [scalets]. Do you see [crosstalk]?

[00:05:24] Dan Siroker: I have a bit of a hard take on this. And I do think, you know, I'm proud that the concept of the product-led growth was actually rooted in some of the early things we did at the Optimizely. I mean, even before the term existed, some of the early things we did at Optimizely were then described as the example of what product-led growth is. And as a B2B company what that sells you, I can tell you that  there really is no such thing that isn't product-led growth. There’s  just different degrees of successes product-led growth. You handle the big business initially, just selling it, but eventually though the product has to deliver, and the world were moving toward, you can't get away with just some SAP, sales crappy product that sort of, somebody signs a contract and they're stuck with for 20 years. 

[00:06:04] Dan Siroker: At the end of the day, if the product doesn't deliver, if it doesn't deliver magic throughout the customer journey, not just deliver ultimately, but throughout the journey. Meaning, you deliver magic that first time you see it. You deliver magic when they're right up for [inaudible]. You deliver magic ongoing overtime. They're just not just gonna stick around, they're gonna churn. And so I do think, the difference really, is how much and how well you get people to go on that journey. And I actually believe, if I'm really being cynical or honest, I think even the concept those sales that growth really is more coming from the insecurity salespeople have around their role in the world, not from actual reality, like at least my personal experience. 

[00:06:39] Dan Siroker: You know, at the end of the day, great salespeople love selling great products. They're not, they don't wanna sell a pile of crap. They don't wanna, they're not a glutton for punishment. As long as they hit their quota and they're making money, their job is easier when the product sells itself. So I, and I don't wanna be dismissive of the role of sales. It's certainly very critical in helping understand the value of the product and helping people understand the problems that it can help solve, but there's no really, I don't believe this idea that there's this antithesis between product like [inaudible], at the end of the day, the product that needs to deliver.

[00:07:07] Andre Michael: Yeah, I definitely see that tradition completely. I think obviously now with GenAI, and the ability to actually do so much more with so much less. And the real differentiates that comes down to the experience and it's gonna become like a necessity, not like a nice to have, where in the past, like you just, a nice looking CMS and you could overtake Salesforce because there's a few bells and whistles, but now you really like to stand out. I even question the notion of sort of the lean startup anymore and what the MVP should look like. Is it like a minimum viable product or not? Does it really need to be a minimum lovable product to actually give itself a chance in this environment? I think things have changed. 

[00:07:45] Dan Siroker: Yeah, I certainly have a lot of opinions on that topic. I do think the reason why that actually comes up over and over again is that there's just kind of two models of people, of builders out there. And I think they both have pros and cons. There's kind of like the craftsman or the crafts-focused builder, which is the idea that you want to put something out there that is impeccable quality throughout., and then I think there's the somebody who's more focused, I would say on, not to be cynical or [positive] or judgmental, but more focused on minimal scope. And so it's, you can have a very loveable product that does just a very few things well. That's the ideal for an MVP, and that's what it meant to be. It's meant to be viable. And if the definition of viable means something that's gonna help you succeed or learn, then quality and craftsmanship need to be part of it. 

[00:08:30] Dan Siroker: So I do think sometimes people say, MVPs are too brittle when their orientation is around jankiness and the UI experience, and those things are really important, but it also, the other mistake I think more commonly happens is the builder decides to overscope, overbuild, and they make perfect the end of the good. They spend three years building something nobody wants. And so the whole idea [but the leans are], but is still very true, which is you got to get yourself out there. You gotta learn and iterate. You can't be 100% proud of what you build. And ideally, what you build and you launch, you're 100% proud of the one or two things you do well. You don't wanna be 80% proud of 20 things you do poorly. Like that's, I think, where maybe a lot of this sort of minimum lovable product concept comes in. But I do think the lesson that most people still don't internalize is that you don't need to do everything, you just need to do a few things exceptionally well.

[00:09:17] Andrew Michael: Yeah, I love the framing of that as well and sort of, it still goes to being a minimal viable product, but viable actually means it needs to wow at to some degree as well on a certain aspect for people to take advantage. 

[00:09:29] Dan Siroker: There's also really, the important part there is like, wow for who? The Who behind the product you're building in the early on is really important. Ideally, if you are building a product that is creating value, there is somebody out there who is screaming from the rooftops that this is the problem they have. And in fact, there's a whole lean startup movement defining this person as the earlyvangelist. They're somebody who is making, they're trying to make this product out of parts. Maybe they're using Airtable or something like they're gluing together things they can find to try to solve their problems. They care so much about the problem. That's the person you wanna start with.

[00:09:59] Dan Siroker: And for that person, minimum viable product, it just needs to deliver a little bit of the value that they're trying to capture. It doesn't need to do, you know, SOC 2 compliance and all the long laundry list of the future factory that you can build later. It just needs to solve their problem. And I think that's important to remember as well is that for a specific Who, you can make 100 people really love your product if you nail the problem that you're trying to solve and you understand the Who.

[00:10:20] Andrew Michael: Yeah, exactly. [Can you try to send me] who's the Who for Limitless? Then like, what did that look like, the discovery process for you?

[00:10:28] Dan Siroker: The Who for Limitless? 

[00:10:29] Andrew Michael: Yes. 

[00:10:30] Dan Siroker: So we define them broadly as, I'm pulling up the dark here, how we sort of [codify] the Who. It's people are busy. They're often back-to-back meetings. They're context switchers. They move from one task to the next. They wanna be better. That's a big part. For example, they like to use [slipper finish trackers]. They're more [pirate the Navy]. So they're willing to seek forgiveness, a permission when it comes to new tools and technology if the tools aren't useful enough. They're nonconformists. They identify themselves as early adopters of new technology. They're likely or to using AI in some parts of their life and they wanna use it more. And they value products that are more utilitarian than more stylish. So they care about outcomes not perception. So that's generally how we define our super specific Who.

[00:11:07] Dan Siroker: What you'll notice there, it's not a job title. It's not a brawl in a company cause to some extent, we're not actually selling to that. We are selling to people who classify themselves in the ways I just described. Those people can be lots of different roles and lots of different titles. And the fact that they're more [pirates the Navy], means that they're individually deciding to buy this product. They're not getting IT pushing it down to sort of corporatewide mandate.

[00:11:28] Andrew Michael: Yep. All of the specificity of the resolve because it also gives you clues and indicators and tools, sort of a channels that you can use them to acquire this, uses the messaging that you can, and hit to reach them. So what did that look like then in contrast with Optimizely? Like, if you can recall, what did the Who look like for Optimizely initially, and how did that evolve of a time?

[00:11:46] Dan Siroker: Yeah, initially we built the product, I wish I had the Obama campaign in 2008. In 2008, I was the director of analytics. I showed up as a volunteer, tried to figure out how to help. And we started doing A/B testing, and the value, that was tremendous. We helped the campaign tremendously through A/B testing experimentation. I was a first campaign to do that, but the pain of actually setting up these tests was really cumbersome. you need an engineer to be every part of the process. You are limited by your ability to sort of deploy these things. I was writing Java script in the front end to sort of tweak the Obama website. And so the initial Who was me. It was like, how do I build a product that I could've used to do 10x the number of experiments we wanted to do? Cause we were always bottled at on our peak capacity to do those experiments.

[00:12:28] Dan Siroker: And then we sort of, and almost luckily, more than anything, we actually realized, wow there's a lot of people like me out there than just me. We found this class of users we described as me or marketing mortals, which was described being a group of people for whom they wanted to be more analytical, more data-driven. I wanted to prove the value of marketing. You know, oftentimes, CEOs always asking, what were the impact? What's marketing done for us lately? And so those people for whom they wanted to use data  to prove that they're valuable, they didn't have engineering resources cause the companies didn't prioritize it. We were kind of in their exact sweet spot. We were perfect solution to them. We were an oasis in the desert for the problems that they're having.

[00:13:08] Andrew Michael: Nice. Very, very interesting in the Obama campaign. I think one of the things I've heard, chatting from others at Optimizely, is that at some point the Who, I think you realized that the market wasn't as big as you may be potentially thought it was. And the reason was that there's not just not that many teams doing experimentation at scale and that level of sophistication. Like, is this something that changed in your understanding of the Who over time? And, how did that process go internally and evolve?

[00:13:33] Dan Siroker: You know, with hindsight, it's so easy to look back and see what we could have done differently, and I wish there was like a hindsight machine where you could just like, in this moment, like right now, I really hope, what is the thing in this moment right now that 10 years from now I wish I had known or done? In hindsight, I actually think quite the opposite of what you said. I actually think the market for A/B testing is massive. And we barely scratched the surface cause we were too literal and maybe too narrow in how we thought about it. Cause at the most basic level, what A/B testing is, is science. It's the experiment. It's experimental method. It's a scientific method. It's the idea of putting a hypothesis out there, and measuring it. 

[00:14:07] Dan Siroker: And I think, we actually made a mistake of falling out of love with that problem, and expanding too early to adjacent markets. You know, things like personalization, things like feature flag development, you know, the equivalent for engineers. Had we really instilled, by the way, optimization is so round, it's a $400 million business right now. It proves there's still a market there. I just think, you know, is when you're closed to, you kind of, you don't realize how big and impactful your product or idea is until it's far too late. So I actually think had we gone deep in A/B testing for ads, A/B testing for creative, AB testing for all the different things that we didn't actually deliver on. That could have been a huge market in and of itself. 

[00:14:44] Dan Siroker: So we actually made a mistake. I think of too willing to sort of move on, and had we just obsessed over mere marketing morals. What are all of the things that's gonna make a marketer look amazing in front of their boss, get them promoted, have them impact to their company. I think we could've built an even bigger company than we did.

[00:15:01] Andrew Michael: I see that completely. I think the notion of not that many levels of state of experimentation, it comes to like, when you think about A/B testing on the web pages, and I think for that to get to that [said] you need to certain volume and you need to be a certain size and then it narrows the market down. But the other use cases that you mentioned now, like I think with ads and things like you get to that [SIG] with this all sample sizes and audiences, and there's a lot of different use cases in interesting ways. I can definitely see how that played out.

[00:15:26] Andrew Michael: One of the quotes I loved as well was like we, at one point at Hot Show, we were working with an exec and I, forgive me because I can't remember his name. They came from Optimizely and we were going through an exercise for our GTM and really trying to understand where the market is going. And one of the quotes that he stood out, he said to us is that, the customer base that you have today is a direct result of the marketing you've done to date and the product you've built to date, but it doesn't necessarily reflect the best opportunity for the business long-term and where it's going. And I think there's sort of a little bit of lines with that discussion we've had now is like, was this sort of some internal debate that you had as well that Optimizely. Like, going through what were the paths and which was the best to take? 

[00:16:09] Dan Siroker: Yeah, we certainly did. And I actually think one of the mistakes I made as CEO was I wasn't decisive enough in terms of focus. Now, I'm a big believer in the saying, "The main thing is the main thing, should stay the main thing." Instead, back in Optimizely, ten years ago, I was making up for lack of focus by just sheer hours. I was just working more, and in hindsight, again, in hindsight, the vast majority of the things I spent my time on, our company spent our time on, didn't matter. It's only really a handful of things that we did exceptionally well that actually made the difference. And I think, had I been more ruthless in our focus, both my time and the company's focus, we could've done a few things much, much better.

[00:16:45] Dan Siroker: Obviously, with hindsight, you know what those things are. But in the moment you don't, so you have to kind of predict. Like, what are the things that are really gonna matter? But, I never had that mindset. In fact, I actually think having kids really helped me hone this skill because I just have a hard stop. I have dinner with my kids at five, put them down, I get back to work, but that period of time really forces you. I'm really productive between 4:45 and 5 p.m. every day cause I know that time is coming. And I think that focus, that sort of constraint, breeds creativity and I didn't have that earlier in my career and I actually think had I, if I could sort of parachute myself into the way I am now as a CEO to that role, boy, I would've done such a better job in hindsight.

[00:17:20] Andrew Michael: Nothing makes you more productive than the last minute, yeah?

[00:17:23] Dan Siroker: Yeah. 

[00:17:24] Andrew Michael: But, yeah. Now, I love that. Now, the focus definitely that's something as well. Like, I see myself in previous startups is that you get, you think everything's important and then which means nothing's important and without this extreme focus. How are you doing that now? It's Limitless. Like, how are you deciding and prioritizing effectively as a team?

[00:17:41] Dan Siroker: Yeah, it goes back to starting with the premise, which is the main thing, should stay the main thing. And that I start by defining what that is. Right now, our main thing is the pendant. So I'm wearing this. This is our wearable AI. It helps augment human intelligence using artificial intelligence, and that's what we're focused on. That alone is probably more than even a team of 20 can do. But that's where we focus, and then specifically, how do we get the product to a, and we have a very clear setup. Here the jobs be done, the things the pendant needs to deliver on. We have a cutline. We say, these are all. Everything above this is what we need to deliver. Everything below this we're just gonna say, hey. not now.

[00:18:13] Dan Siroker: And then, for all of those jobs we done which is kind of the product-facing, user-facing, kind of problems we're trying to solve. We have clear setup engineering priorities, and that's stack ranks with an order with a list, with DRIs for each engineering priority, with a line that we don't go below. So it really starts with that. Sort of writing things down and being clear. Especially, a [real] first company, you gotta write stuff down. You can't just sort of expect this sort of tribal wisdom like, hey, what do we focus on, what are we supposed to do now? You can't assume that that's gonna get there cause even when you write down, you still gotta repeat yourself 100 times for it to really land. So that's how I do it today. And maybe I'm right, I may be wrong, but at least we're making decisions. We're being clear. These are the priorities. And hopefully we got a few of them right and that makes the difference. 

[00:18:52] Andrew Michael: Nice, yeah. I think that was one of the things that surprised me when I first joined Hot Show was how deliberate they were about being a remote company and then how much documentation played a part in ensuring. I was blown away by the onboarding experience. My first week there sort of like, got a full understanding of the entire company, everyone in it and how it operated and so forth. And I think like, in these environments you need to. And I think that's even where products like Limitless can even come in and play a big role and help for the team.

[00:19:18] Andrew Michael: You have the above the line. I'm keen to understand how much of a factor activation and onboarding is in that above the line motion, going back to the start talking about how important it is, and how do you prioritize and think about that initial experience for your users?

[00:19:34] Dan Siroker: Yeah, I mean, it's critical. Like, I just never built a successful product that didn't have some key 'wow moment' in the first 30 seconds. I mean, it's possible, I just don't know how to do it. I don't know how to build a successful product unless you have that. Because people are busy. They don't have time, especially the kind of products I've been building, Optimizely and now, Limitless, which are generally kind of new products or new categories or new reinventions. A/B testing's been around forever, but the idea that a marketer could do without a developer, what you see is what you get editor and real time results, and all that felt very new. 

[00:20:04] Dan Siroker: So you're kind of educating the market on what this thing is. It's not like, hey, here's a new FPNA tool, or here's a new spreadsheet, like everyone uses spreadsheets, or everyone uses PowerPoint. So when you're creating a new category, I do think you need to hook people quick. And that's where it's core to the product, with Optimizely it was, you could come to our homepage and put in any URL you want, you don't have to sign up, you don't have to create an account, and immediately once you put the URL in on the homepage, you can see our visual editor. You can start making changes to your website. Not our canned demo but to your actual website. And so that was a key hook to get people to see, wow, this is actually really easy, I can use this. 

[00:20:42] Dan Siroker: Similarly with Limitless. Today, if you go to limitless.ai, you don't even have to sign up, you can try the product anonymously, you can see transcription, you can see live notes, you can see summaries, you see the value before you actually have to buy a physical device that you put on your body. So I think that's important because you gotta lead people to the lottery. You gotta show the value of the product, and once they see that magic, then the beauty is they pull you. You're not pushing them. Then they start desiring and wanting more. And that's what also led us to do in the pendant is sort of seeing the market pull us toward it.

[00:21:10] Andre Michael:  Nice. And the pendant as well is like very, cause it just triggered a thought now as well the onboarding experience. So far they said the pendant works really well in the real world. A lot of the times like in remote environments, we speak and I'm going to put our features probably not a great question for the show, but. How does the product itself deal with a conversation like this now? How was it able to understand what’s being said from our part?  

[00:21:32] Dan Siroker: Yeah, so it is, it's designed to help you, and for everywhere you go. You know, we know some users just wanna stop. They're using all kinds of online recording meeting tools. [unclear] something I want a device that they, where they can set it and forget it.   And whether it's a casual conversation with your wife before you go to Trader Joe’s where she runs you go get the milk. Or it’s a, you know, it’s a cute thing your kids ad or it’s action item for a meeting, you know, It's all your life, and you want to remember those moments in your life. You want to remember your to-dos. You want to remember the moments you care about, and that's what we help you do. 

[00:22:01] Dan Siroker: For a conversation like this, you know, it's a little bit more challenging than you could imagine if we just had, like if we used a standard, by the way, back in 2020, we started, we were just a vanilla sort of meeting bot. We joined a meeting, we recorded. That's how we started, and at the time was pretty novel, now everyone does it. So what we realized was there's some really hard problems that we need to solve to making a wearable replace those kinds of experiences. 

[00:22:22] Dan Siroker: A big one is, voice fingerprinting and speaker diarization, knowing who I am, knowing who you are. And in a meeting or a conversation like this, capturing what you say, capturing what I say, separating that out, being able to identify by the fingerprint of your voice who you are, that's critical to summaries, to note taking, to action items. When I say I'm gonna do something, that's an action item for me. When you say you're gonna do something, that's not an action item for me. So those are the kinds of really hard problems we've focused on. And then there's a whole class of problems we've focused on around privacy.

[00:22:51] Dan Siroker: You know, how do we build a product that's not just good for the person wearing it, but good for you, the person I'm meeting with, good for society at large. I can talk at length about all the things we're doing there, but rest assured our goal and what our focus is on is making sure your privacy is not coming, your convenience is not coming at the cost of privacy and privacy is not coming at the cost of convenience. You have both. 

[00:23:11] Andrew Michael: Yeah, I can definitely see these endless challenges when it comes to building a product like Limitless, but...When it does work, I think it is gonna be something like truly magical. That really brings a lot of value to a lot of people's lives. I'm keen then to hear a little bit about the journey now when it comes to churn and retention and what you're seeing, it's still early days, but how things have been going so far, like what have been some of the challenges on that front?

[00:23:35:] Dan Siroker: Yeah, I think we've learned a lot. You know, the very first thing when we launched Rewind, which is the predecessor of Limitless, kind of inspired what we're working on. We were kind of blown away with the number. We sort of put something out there that we thought was kind of rough around the edges. Shows you, by the way, a minimal viable product if you're solving a problem. Doesn't have to be perfect. We put it out there. We got like 300,000 people on our waiting list. Way more than we, I think our goal was like 1,000. And right away we realized, wow, there's a ton of, we're tapping into a market need that we didn't even realize was as big as it is. 

[00:24:02] Dan Siroker: And so what we did was we did some experiments. The first thing we did was a pricing experiment. And we actually, before even letting people use the product, we actually had them put in a credit card and you know, do a just standard 30-day free trial, sort of prepay cause we had just huge flood of you, we wanted to sort of focus on the people who might eventually pay, that had sort of like you can kinda predict, how am I go, that had a classic. Like, huge influx but you know, just not everyone's gonna convert after 30 days. So like you end up sort of, you sacrifice that first retention in exchange for higher top-line revenue to start. 

[00:24:34] Dan Siroker: Then we realized let’s move to a model and then we realize from that evolution that you know from a customer experience, it's worse. To sort of over-promise on under-liver. I’d much rather create so much value for somebody that they like, they’re dying to pull out their credit card to pay then they pulled out their credit card and they kinda had like a subpar experience or maybe didn’t use it [unclear] to make sure the main reason people would convert after 30 days, they just forgot to use it. Which is a bit ironic, so the product is helping you remember but before after they use it, and so then we realized, let's just wait until people get that magic moment. 

[00:25:00] Dan Sirorker: And then we sort of to find the pricing around that, we looked at, so we rewind then next evolution was basically a usage based indefinite free trial. Like you get to used and still today you can use it and it only wasn't until you reached certain thresholds of usage that you would actually end up paying. And that we looked at the data very carefully. We said, where is that magic moment? When do we see through people's behavior? Not what they say they'll do, but they actually do. When do they become sort of asymptotically retained indefinitely? And that's where we put the pay gates.

[00:25:28] Dan Siroker: And then similarly with Limitless, we're going through that similar exercise where we look at, so when do people get to that magic moment? When they're sufficiently enough activated. And that's when we sort of maximize the retention by charging them then.

[00:25:39] Andrew Michael: Nice. I think with Limitless, it's a little bit different cause they've actually invested in the product itself to begin with, so you have a little bit more of a stronger hook there, but what was the motivation as well then to launch Limitless off the back of Rewind if you had seen that initial traction, and what drove you in the direction of a wearable? 

[00:25:56] Andre Michael: Yeah, it was not an obvious choice, by the way. I mean, this is not, and there's certainly a lot of diehard Rewind fans who hate the fact that we're no longer focusing on the Mac app and we're building something that's more general. And so…It was inspired really by the market, you know, kinda learn the promise people are having, we started to see more and more the people wanted the convenience of accessing their data from any device. 

[00:26:16] Dan Siroker: You know, Rewind was a Mac App your data was stored locally on your Mac which is great for a lot of reasons, but your kinda limited by your ability to access the data and of course we wanted you to be limitless. So that sort of said, Okay, let’s how we do find an architecture that can deliver on some of the same things people love about that local approach, but in the cloud? And so that’s what it started inspired that thread of work.

[00:26:36] Dan Siroker: And we also realized that the foundation models are just gonna get better and better. How do we build a business that banks on and sort of relies on the underlying foundation models to get better? And unlike most companies or some startups which would get sort of fearful of a new open AI announcement or a new anthropic announcement, how do we make it so that we're joyful no matter what happens? And how do we avoid being steamrolled? And that's another sort of premise behind the shift which is how do we leverage the best cloud-based large-angle models, not a local large-angle, but the cloud-based models. So that our product just gets better on its own. Gets cheaper on its own, gets better on its own. 

[00:27:10] Dan Siroker: It actually does get, you know, when GPT-4.0 came out, literally our product, our costs, halved. We just, all of a sudden, overnight, it was like a magical, it's like the same quality, a little bit better actually in some ways, and our costs halved. Like that's where we want to be, and the analogy I draw is sort of to 1996 and an internet company. Like if you were an internet company in 1996 or 1997 or 1998, all you had to do was make sure that you were on the internet. And already by the fact that more people would get on the internet and people's bandwidth was getting faster, your product would get better and you'd get more users. And we're doing the same thing with Limitless. 

[00:27:39] Andres Michael: Yeah. No, I love that. And I love like sort of the focus on the assumption that these are just gonna get better because I think there's a lot of like companies and startups out there that aren't running under that assumption still, and they're still concerns that the data is not perfect or it hallucinates or it like and instead of just like seeing the path and as you say, sort of like the internet back then, the dial-up speeds and where we are today, I think we're definitely on this trajectory and just week by week you're seeing improvements everywhere. 

[00:28:06] Andrew Michael: So like running with that assumption. And then the second thing is like how to not be a victim of this wave though, as well, I think, because there's a lot of startups that build off like use cases that are pretty much what chat GPT is going to just be a rapper. Whereas like the value I think to product like Limitless really is in the data that is stored in the context. And I think like, for me, I think the next wave is in this understanding of context is king. I feel like when it comes to the product and having that context is where you can create these personalized AI experiences. 

[00:28:37] Dan Siroker: Exactly, yeah. We view the context as our mode. That's why we did the pendant. How do we capture more context? How do we store it in a way that people trust? How do we walk the line of what I describe as the personalized AI privacy paradox? which is this idea that like, yeah, if you want to personalize the AI, you want it to have more context, you want more data, but the more data you have, then you're like, wow, I'm a little bit worried about data protection, which then makes it, you want to build systems to make it less useful and sort of capture less data. 

[00:29:02] Dan Siroker: And so we're trying to walk this tightrope between making it as useful as possible with the data and trusting that the privacy of your data is gonna be paramount. And that's kind of what we think is gonna be key to building a company that defines sort of personalized AI and captures as much context as is useful to make your personalized AI as useful as possible.

[00:29:20] Andre Michael: Very nice. When it comes to the pendant itself, then how I've, because that's like another new layer as well for you. So you sort of in B2B source where you could just subscribe, unsubscribe when you wanted, then you went to sort of a more B2C, but still B2B to some degree with Rewind and then now with the pendant, you don't only have like the B2C element, but you also have a hardware elements involved. How does that influence like churn and retention specifically? And my assumption will be that you see, higher retention rates just due to the initial investments and that up front capital being played but that’s something you see in your data or something that you've noticed? 

[00:29:56] Dan Siroker: Yeah, we are, we still don’t know yet on the data cause tomorrow actually is gonna be the very first handoff to a customer of the pendant. So we’ve had a long list of pre-orders. We were very excited. We've been building it in our [unclear] scaling production in our factory and tomorrow is literally, I’m flying in the morning to San Francisco to handle our, the very first pendant to our first customer. We’ll share a video on that later. 

[00:30:15] Dan Siroker: And so we don't know. But the premise and sort of assumption is that the energy and sort of the mental energy to actually put something on your body, it needs to be valuable enough to you that you’re gonna actually remember to unplug it from the charger, charge it, put it on. And that’s, that already in itself is like a huge [unclear] but also if we've delivered and something that people find valuable, that will also make them very well retained.  

[00:30:39] Dan Siroker: So our focus is kind of on how we build the best possible product that you're gonna wanna wear every day and you can’t imagine living a life without it. Certainly that's the experience I've had so far. And then we think, we can actually almost sell the devices at a loss. You know, it's kind of a Razor-razorblade model where the device itself is just the hook to maintain retention, not the main revenue driver. And that’s where the beauty of subscription model comes in which is you know, then people are customers indefinitely because of the value they get, not because of this up front cost of the actual physical device. 

[00:31:09] Andrew Michael: Yeah, well. So you must be nervous then tomorrow is the day. I definitely think like the challenges that you mentioned as well, it's introducing a new behavior. And I think behaviors are incredibly hard to change. To some degree that we are familiar with like bringing devices with us and charging and things, but it's another one to include. So it'll be interesting to see like how, I think it reminds me of previous guests, Nir Eyal with the Hooked Model, talking through these things and one of the hardest. 

[00:31:35] Andrew Michael: But I think like, as you say, like, If you can deliver that value immediately and it becomes clear and apparent and specially for your audience that you mentioned as well, it will be very interesting to see how this tracks. And I love to probably catch up in like 6 months time to see how it's been running   

[00:31:50] Dan Siroker: Yeah, I'm actually not nervous, you know, I think we've got  a lot of bugs to work, work through and it's still a hard complex product and getting it to work just right is gonna take a while. But I think we've done a good job of focusing, making it  clear with the job’s to be done are, and you know we’re excited to see and learn from our, you know, first customers, how they use it and how we can improve. And we’re patient, you know we wanna make sure-

[00:32:11] Dan Siroker: There’s a lot of companies out there recently and I won't name names. Who’ve built harbor products that I think did a really bad job of over promising and under delivering. We're tryna’ just deliver on our promises. You know not over promise, be clear of what we can do, what we expect to do, we can’t do everything, it doesn’t have lasers, you know. It doesn’t have all the fancy things that maybe other companies thought it needed. We’re just trying to do a few things exceptionally well, And if we do that well, I think we'll earn the right to keep a customer. And that’ll tell others and we’ll grow from there.  

[00:32:38] Andre Michael: Nice. Going back to Minimum Viable Product and [unclear] few things exceptionally well. Great. I see we’re running up in time, so, I wanna make sure I ask you one question [unclear] but I also have one additional question that I am intrigued and interested on. Is that, fundraising, you took a totally unconventional approach to you almost like open sourced it, went out on LinkedIn. Where did the motivation come from to do something like that? And obviously, like, I know it was incredibly successful because I saw the Docs and saw the following from your perspective, but what motivated you or inspired you to actually go out that way? 

[00:33:14] Dan Siroker: Yeah, so, you know, I'm just a deeply nonconformist person. So anytime there's a chance to do something differently, I almost always try to embrace it. In this case, it was really actually, it didn’t root, it wasn’t rooted that we needed the money, we still haven’t actually spent our seed round  or we started spending it when we raise [unclear] but it rooted in this realization that we really believe for us to be successful, customer need to trust us. Like that is just paramount, because if they don’t trust us, then they’re not gonna be willing to use our product to capture more and more of their day and you know, it's not going to work. And we started to think, well, what are the ways to build trust? The best way in my opinion to build trust is transparency.

[00:33:46] Dan Siroker: So we decided, and we still continue to do this. I put out my 360 reviews, I put out Transparently to the World. This was just putting us our pitch deck out. Hey, I'm getting a lot of people investors want to meet. Instead of me having to meet with all these investors, here's just on Twitter I posted my pitch, the pitch I gave, the standard pitch I gave, seven minutes. It went totally viral, and we got a lot of interest. We got you know, thousands of folks interested in investing, a few hundred I think confirmed offers. 

[00:34:10] Dans Siroker: And then you know, then I was like shoot, guess I should raise a money. So I know that I really did my homework. I really narrowed it down to like the absolute best of the best of that group. You know, I was able to actually include a lot of people as part of the [unclear] I think that's what you're talking about, which is a lot of early customers, interest people. People don't typically invest in startups, they were like hey, this is great. They're accredited investors, they wanted to join and be part of the journey. That’s been great as well. So it worked out great and I highly recommend it. I do think the sort of like if you’re founder listening to this and thinking to yourself, I should just raise money how everyone does.

[00:34:40] Dan Soriker: Let me just draw this analogy to you today the way people raise money is the way dating was before dating apps. You would just go to the local bar and hope there's somebody that you could fall in love with and get married just happens to be at the bar with you. Or you could cast a white net, you know, put yourself out there. Maybe you're maybe a little more public on the dating sites, but you actually end up getting somebody who's probably much more likely to be compatible than somebody who just happens to be at the bar on the Friday night that you're at the bar. And I think most investors like that. You know, they liked the idea that you're gonna come up and down Sand Hill Road and kiss their ring and give them the...exclusivity and the narrow market that they like. 

[00:35:17] Dan Siroker: But you as a founder, you don’t, you shouldn't. There's lots of great investors out there. You may not even think to want to invest in your company unless you cast a white net. So be a little vulnerable, be a little open, put yourself out there and I highly recommend it. I do think it's only, just seeing the success we had and now I think a couple others are doing it, people will be shocked at the way that investors and founders raise money today. Founders will be kicking themselves, like why did I do this so silly backroom deal with three or four people up and down Sand Hill Road?

[00:35:42] Andrew Michael: No, I love it as well. Again, like I've said, I love it a few times on this podcast. It's been great. But, the, because also notion, I think typically like founders don't realize like if you're building a good business, how much power you actually do have when it comes to it, because the number of really good businesses out there is very, very few and like investors meets hundreds and hundreds of people every year and invest in on your handful, it's because not because they only choose to invest in a handful, but just because there's only a handful of good businesses. 

[00:36:08] Andrew Michael: So if you do have something that is magical, like some really good [peregrine] experience. Like this is a fantastic way to create former [unclear] and really get them to compete with one another. So like I was amazed when you sort of, it piqued my interest completely. I was also curious to see how many actually users you acquired as a result of a just one people, like, oh who is Dan? What is he doing? Why? Maybe I should check out Limitless too.   

[00:36:32] Dan Siroker: Yeah, Yeah, it was massive. I mean It was win all front it helped build transparency, build trust, got a ton more users, got some great new investors. It was quick too, those are the things, it was really quick. I went from putting this deck out there, it’s like sign sign term and issue, I think a month which is unheard of, including, you know, in that time, a week after, I put that in my calendar, but I was like I had I don't know, 60 or something back-to-back meetings with like, really great investors who just, and the nice things is they’re all now, they're moving their schedules around. You’re not sitting there start doing a pitch every often but at the end of the week, I got really good, because you're just doing these back-to-back. So, that was another thing that was super efficient. 

[00:37:07] Andrew Michael: Get them to pay for the meeting as well, I think-  

[00:37:09] Dan Siroker: I didn’t do that, that was more recent, That was more of an experiment. I don’t think that was such a good idea because I think that, what I did was I was again, so we still need the raise money but I was still getting a lot of inbound interest. And a lot of inbound the interest was just like some sort of associate VC Firm who just need some quota. So I thought, hey, let's just donate some money to charity. If you wanna meet with me here’s my Calendly link you gotta give a hundred bucks. 

[00:37:29] Dan Siroker: I think that rubbed some people the wrong way. So, I'm not gonna do that again, but it was a experiment, we’ll see. I rubbed them the wrong way cause I think it sort of sent the message that I think I'm somehow better than them or whatever, even though we weren't giving the money to charity. But you know, investors have sensitive egos. So you gotta be sensitive to that. 

[00:37;43] Andrew Michael: For sure. Great. Last question then from our side is what's one thing that you know today about churn and retention that you wish you knew when you got started with your career? 

[00:37:52: Dan Siroker: Boy, there's so many things in terms of lessons that I would share or things that I wish I known early in my career. I wish I had recognized it's gonna sound super boring. I had reached, reached, I had recognized how simple it is and that it ultimately always comes down to did they use it? You know, we did this big sophisticated analysis that optimized the un-churn and retention. And it comes down to like, we had this like 90-20 rule or whatever, basically like, you know, you could boil it down to if a user didn't use your product, they're not gonna renew. If they did, they will. And so I think that key focus on usage, which is very measurable, I think something I wish I had been more, you know, simple and main thing about. 

[00:38:31] Dan Siroker: Which is just make sure you understand what it takes for a user to use your product. When they use it, they'll renew. When they don't, they won't. And that was, it seems like too simple of a lesson because everyone, like you said earlier, everyone's looking for like, what's the seven friends in five days or whatever, the magic formula. It comes down to like, did they use your product? You know, when they signed up for it, they signed up for something, they paid you, and now if they're using it, they're likely thinking they're getting nothing they paid for. And then it's as simple as that. Get your users to use your product. 

[00:38:56] Andrew Michael: Yeah, I know, absolutely. I think that's something I say as well, quite a bit on the show, is that like, it really is that simple. They came to you with a problem. If they're not getting value, they're going to turn like value comes through usage. And so we overcomplicate it, I think a lot in different spaces and times, but then it's been an absolute pleasure chatting with you today. Is there any final thoughts you want to leave the listen with anything they should be aware of or how can they keep up to speed with your work? 

[00:39:20:] Dan  Siroker: Yeah, I'd love for a thank you so much for having me. Yeah. If you're interested in trying Limitless, go to limitless.ai. You can follow me on Twitter/X @disroker. You know, my only parting advice is just remember that the main thing is the main thing should say the main thing. Hopefully that's helpful. 

[00:39:35] Andrew Micheal: I love it. So, well, thanks so much for joining in. I wish you a bit of luck now with your launch tomorrow and looking forward to following the journey. 

[00:39:41] Dan Siroker: Thank you so much. 

[00:39:43] Andrew Michael: Cheers.

[00:39:44] Andrew Michael: And that's a wrap for the show today with me Andrew Michael. I really hope you enjoyed it and you were able to pull out something valuable for your business. To keep up to date with churn.fm and be notified about new episodes, blog posts and more, subscribe to our mailing list by visiting churn.fm. Also don't forget to subscribe to our show on iTunes, Google Play or wherever you listen to your podcasts. If you have any feedback, good or bad, I would love to hear from you. And you can provide your blunt, direct feedback by sending it to Andrew@churn.fm. Lastly, but most importantly, if you enjoyed this episode, please share it and leave a review as it really helps get the word out and grow the community. Thanks again for listening. See you again next week.


People on this episode