Why Most Apps Hit a Revenue Ceiling (and How to Plan for It) — Patrick Falzon, The App Shop

David Barnard:

Welcome to the Sub Club Podcast, a show dedicated to the best practices for building and growing app businesses. We sit down with the entrepreneurs, investors, and builders behind the most successful apps in the world to learn from their successes and failures. Sub Club is brought to you by RevenueCat. Thousands of the world's best apps trust RevenueCat to power in-app purchases, manage customers, and grow revenue across iOS, Android, and the web. You can learn more at revenuecat.com. Let's get into the show.

Hello, I'm your host, David Barnard and my guest today is Patrick Falzon, co-founder of The App Shop, helping companies of all sizes build, grow, and monetize exceptional apps. On the podcast I talk with Patrick about estimating the revenue potential of an app, crafting an exit strategy, and why LTV is such a terrible metric. Hey Patrick, thanks so much for joining me on the podcast today.

Patrick Falzon:

Thank you for having me. Big fan of the podcast and big, big fan of RevenueCat.

David Barnard:

Yeah, super excited to have you on the podcast. I wanted to kick things off talking about your time at Mosaic. I think that's a good framing. It is kind of where you learned the ropes as far as subscription app businesses go, and Mosaic did a lot of really interesting stuff and built a pretty great business, famously recently got sold off to Bending Spoons. But I wanted to talk a little bit about your experience there, what brought you there and what your role was as context for what we're going to talk about throughout the rest of the podcast episode.

Patrick Falzon:

Yeah, the brief background of me is my career has been roughly split between investing and operating roles, and Mosaic was actually the place where I made that transition. So I got introduced to Mosaic when it was actually just Apalon back in the day, for folks familiar with that brand. Apalon had bid really early to the mobile app ecosystem. Our parent company, it was a company called IAC, had acquired them when everything was still paid downloads. And so Apalon, we owned it when it went through that transition from paid downloads to subscriptions and we saw phenomenal growth and a ton of opportunity on the other side of that. And this was at a time when I was still doing mergers and acquisitions for IAC.

While I was on the team, I got introduced to the Apalon team, they had a big vision. They felt like they really unlocked something special in terms of understanding and being able to capitalize on that in-app subscription opportunity and wanted to go out and build a big portfolio of other apps and other app developers. So I originally came in to help them execute on that from an M&A perspective. First through the acquisitions of iTranslate and Teltech, iTranslate being the developers of iTranslate. Teltech, the developers of Robokiller plus some others, but Robokiller was the flagship product there. And then a number of other smaller deals on the other side of that.

So I got involved in Mosaic and then within a couple years I actually ended up running both iTranslate and Teltech as the general manager of those two businesses while still doing M&A for Mosaic. So I got to see the landscape, I think, from a couple different lenses. One, both from the investing and the operating side over a multi-year period and then also really got to ride that growth of the mobile app industry.

Again, we were pretty early to subscriptions and when there was kind of just infinite growth it felt like in those early days, all the way through to now where I think it's a much more mature ecosystem, everyone's on subscription, there's a lot more defined best practices around how to do it. And then frankly, I also think in some ways growth is maybe harder than it used to be and it takes more execution than it used to. So it's been a really fun journey. And all I would say is at Mosaic we do feel like we accomplished a lot. We felt like we built one of the leading portfolios in the space, had a great exit to Bending Spoons, hope they do well with the portfolio on the other side of it, and excited as for what comes next.

David Barnard:

Yeah, it must have been a fascinating time going from that single kind of portfolio of apps with Apalon to then acquiring and getting to kind of peek inside a lot of app businesses in the M&A role and then actually become an operator over Teltech, which I know Robokiller was one of the kind of flagship apps of the Mosaic portfolio, while still doing additional M&A along the way. So it's like you were both kind of seeing what was going on inside these other businesses that Mosaic might want to buy, but then also implementing those strategies within the existing portfolio companies. So you just have such a unique perspective. And again, that's why I think it's going to be a really fascinating chat today to learn from those experiences, one of the top and one of the early movers in this subscription app space.

So the first thing I wanted to talk about is market sizing. And this is something you must have thought a lot about when you had your M&A hat on was, "How big is this market going to be for this app? How much can we afford to pay? How much do we think this is going to grow?" Because famously Mosaic didn't just buy up a bunch of apps. There's a lot of companies now that'll just buy any app that's doing 10K a month, 100K a month, and it seems like there's not as focused a strategy. But with Mosaic there, it seems like, a lot more thoughtful approach to trying to buy apps that could be big.

And then the other thing I imagine as general manager at Robokiller, then you also had to think about, "What's the opportunity of any new feature we build?" And those are probably two very similar problem spaces to think about, like, "We had this existing app that has a market cap, is this secondary product market fit? Should this be a new app?" And there's so much intertwined with those two, so I want to kind of lump those together. Let's kick it off, how did you think about in acquiring a new app or going into a new line of business how to really understand that potential TAM for that app?

Patrick Falzon:

Yeah, so we did think about it across a couple different lenses. So one was just market size. You are right, we went for depth, so we needed to believe that you could have a product that did tens of millions of dollars of revenue, was kind of a rough benchmark we had for ourselves. The main reason we did that, and I think one of the challenges when you go more downstream than that is you can easily end up with a portfolio of 50 to 100 apps that each do a couple hundred thousand dollars of revenue and your overall P&L looks good.

But it is really complex managing that many apps, one. And two, what it probably means is you're very rarely actually investing in each of those apps. Maybe there's a couple you do, but you have this long tail of kind of zombie products, where they're making revenue for you and maybe that makes financial sense, but it was hard for us to justify that we were truly satisfying those users or providing value on the other side of that. And that's why we tried to shy away from that and really wanted to buy products or build products where we felt like this was an enduring product, an enduring category, we can invest in it and actually provide value to our users on an ongoing basis and be rewarded for that. So we did look for depth.

Now, on the flip side of that, I do think there's a double edged sword on your TAM, where I could also think about you should think about how big do you really want to play in, because it gets into the second lens, which is competition. And typically the bigger markets are going to have more competition and you have to think about how you think about those tradeoffs between those two things, meaning a big market is great only if you can take a substantial share of that market or a specific share of that market. If it is so competitive that you can't actually garner market share, it's not actually valuable to you. So we thought about size and competition as a balance between each other and I think there's a little bit of a Goldilocks approach you can take to those things of markets that are big enough to create enduring products and businesses, but you could still be maybe a relative big fish in a smaller pond and that's not necessarily a terrible place for you to play.

David Barnard:

Do you have some concrete examples of like... I mean let's take Robokiller maybe as an example of something that Mosaic did buy. So what was the thought there on how big that market was and the thesis behind playing in that market? And then maybe an example of a market that was maybe too competitive where you looked at an acquisition and felt like, "This is too big a space to compete in".

Patrick Falzon:

So the Robokiller example, the basic thesis there is it's a broadly applicable product. Everyone gets spam calls or spam texts to a degree. There are certain demographics that care more about solving that problem than others. But we got comfortable that it was a broadly appealing product that could generate, again, tens of millions of dollars of revenue. And then we looked at the competitive landscape and Robokiller was the leader. You had carriers that are somewhat relevant here, but they're not really invested in solving the solution for a variety of reasons that are probably more of a rabbit hole than worth getting into here. So we looked at this is a problem that needs to be solved by independent tools. This is the leading independent tool. There aren't major players entering this space from adjacent categories. So again, good size, good competitive dynamics.

I think one you can think about in the other direction categories that we never got in ourselves. Two examples that come to mind are VPNs. VPN is a massive category, but there's also already a lot of people making a lot of money there. So if you want to enter that category, you have to be prepared to spend a lot of money on user acquisition to be able to garner market share because everyone else is spending a lot of money on user acquisition and if you don't play that game there is no game to play. So that was one we shied away from because we said the uniquenomics are not going to make sense for us if you have to be so reliant on aggressive paid marketing. Slightly different lens, but similar rationale is we never got into personal budgeting or finance, for a similar but slightly different reason in the sense that we thought the adjacent competitors in banks and more traditional financial institutions made that challenging long-term.

Similarly, when you go out to acquire users, if you think about what you're actually acquiring them for, you're acquiring them for financial management or budgeting or their money and how should they manage that? Those are the same desires and objectives and keywords to be very tactical that a bank would potentially want to acquire that customer for. And the bank is going to have a much, much higher LTV than you are ever going to have on your budgeting app. So you will always lose that battle because they'll always be willing to pay more for that customer than you. So again, that was one we looked at and said, "There's a lot of interesting things in the category, a lot of interesting features that make those products attractive, but we don't think we can effectively compete there in the long-term because it is such a big and attractive market," because the LTVs can be so high if you actually provide the financial services, which is not something we were ever going to do.

David Barnard:

Yeah, no, that makes a ton of sense. And then you had a follow up, I interrupted you, but I appreciate the concrete examples. And I think again, super helpful for folks to think about as they're thinking about growing their business and those same competitive dynamics. I mean I know folks in the meditation space and the language learning space and it's like, yeah, those are big spaces, but it's important to understand those competitive dynamics and understand the challenges you're going to face along the way. But yeah, you had more to say.

Patrick Falzon:

One point on that I'd say is in those spaces I think there is an argument for if you're bringing something truly new and innovative to the table, the whole are you 10x better than the current solutions? Not to say you shouldn't try, but that just being 10x better doesn't solve all of those dynamics, meaning you still have to get your product in the hands of users and let them see that you're 10x better. So you still need to have a strategy for how are you going to do that user acquisition if everyone's already doing paid and they're already gobbling up that top of the funnel. And then the third lens I was going to add is we thought a lot about are we tapping into existing demand or do we have to go out and build a new product category? Not to say one is always going to be better than the other. They're just different and they have different capital requirements.

So the way we thought about it was if you're tapping into existing demand, you're probably going to have more competition up front, but it's typically a lower capital investment because you can start doing things like performance marketing right away or ASO right away and start getting users in in a relatively profitable or rational way. If you have to go out and build a new category, people don't know this product exists, you're the first to market, you have to build demand for it, you're probably running at a loss for some period of time while you go out and do that. So you need to factor that into your thinking of, "Can I do that and am I willing to do that?" So I think particularly for independent app developers, one thing I talk to developers a lot about is, "That's a really great idea and that's really exciting. Do you have the resources to see it through and scale it on the other side of that though?" I think that's how you should think about it.

David Barnard:

Yeah. And then how did you apply that thinking, how do you today apply that thinking when it comes to building new features? I mean, this is something where folks maybe aren't always strategic enough, it's like, "Oh, we'll build this new feature because it's cool." But then I imagine there's some thought that should go into, "Is this a secondary product market fit? Is there a keyword that this feature is going to allow us to compete on that we're not competing on currently? Or what was the kind of strategic thinking behind expanding feature sets and secondary product market fit and then even where maybe spinning up a new app was a better solution than adding features to an existing app?

Patrick Falzon:

Yeah, so I think you hit on all the main points. The one thing I'd say upfront is the biggest mistake I see, to your point on building features because they're cool, is teams build what they think should be in the product and what they think would be cool to be in the product. And at no point do they talk to the users and try and understand, "What do your users want? And are you building what they want and need?" It's really easy to just keep building what you think is cool and then those features never get used and they get lost in the shuffle and you end up with kind of a crowded product on the other side of that. So strongly recommend talk to your users, get feedback from them around what they want, because ultimately that's what matters. You're not building a product for yourself, you're building a product for other people.

Within that though, we did think about it in terms of, so when we're building a new feature, is it a feature or a product? My general bias in the mobile space is towards slimmer products and being okay having multiple products that serve adjacent use cases. I think the mobile form factor has a lot of benefits, but it also has some limitations. And I think having a product that, for example, has eight different features on a smartphone app and a six inch screen, it's very hard to get users to understand all of those features and effectively switch between them and use them. And I think maybe having two or three products that have two or three features each could actually deliver more value and you may even get more financial value out of it because you could better segment users into the product that actually suits their needs.

So we did think a lot about that and we generally erred on the side of less features and being more willing to just build a net new product if we thought there was demand there. Then when we were actually doing features, we went through that analysis and said this should be a feature in the product, two ways we thought about it there, which is, "Are we trying to increase ARPU or are we trying to increase retention?" And I would say all of your features should fit into one of those two buckets because I'm not sure what the other rationale for releasing a new feature is if you're not going to increase ARPU or retention.

David Barnard:

That's a great lens.

Patrick Falzon:

So then you have to think about, "What am I really trying to achieve with that?" So we had some where we released and it was an upsell and we said, "This is an adjacent use case. It helps solve the same problem." One example is in Robokiller we integrated data broker removal service and that was an upsell, so we sold you into the base subscription and then we said, "Do you also want this on the other side?" It was really successful, a really good way to increase ARPU and it made sense having it as a feature within the product. There are other things you're going to release or features you're going to release where it's more about retention and engagement. And I think the strategies for how you develop those features and how you market those to your user bases are very different. So having that clarity up front, is this a retention oriented or an ARPU oriented feature, I do think is very important.

David Barnard:

In thinking about that kind of market sizing both for features that you're adding onto and new apps or new lines of business, one of the things I know comes up a lot in the industry is how many apps seem to get stuck in this 10 to 30 million a year revenue threshold. How did you think about hitting that ceiling, why some apps hit the ceiling and then why some apps don't and how to push through that.

Patrick Falzon:

Maybe not the answer everyone's going to want to hear here. I tend to think most apps will hit that ceiling and it's actually very hard to break through it. So the way we tended to think about it and the way I still think about it today is more of accepting that as a constraint for the mobile business, and then how do you architect around it or how do you build a business given that constraint? And I think there's a couple of reasons why that's just a real constraint. One is if you think about user acquisition, at a very basic level you're going to get organic users, people just coming to the app store searching for things and they find your app as a result of it and download it, and then you're going to do paid user acquisition and go out and pull users into your app.

On the organic side, it's hard for any individual developer to materially move the amount of searches happening in the app store. That's a pretty static figure that's relatively out of your control. So all you can do is increase your ranking for those keywords. And at some point you're going to hit enough top three rankings, you you're going to get your rankings and they're going to stop being able to move and you're going to almost maximize your share of that organic pie.

Now not everyone always hits that point, but there is a cap on the organic side for most products. And then you get to the paid side. And I think the challenge there is most people similarly hit a ceiling on the paid side because mobile LTVs just aren't in the grand scheme of things, if you go to other markets like B2B SaaS, LTVs in mobile are pretty low. Most people are operating under $100 subscriber LTV. So there's only so much you can then pay to acquire a subscriber. And then basically what happens is you're going to keep spending and at some point your cap is going to reach that maximum and you can't spend more unpaid without going unprofitable so you decide to stop spending. And then when those two things happen, that's when you kind of hit a ceiling on your growth and you can get, call it, single digit growth on the other side of that through LTV improvements, testing, conversion rate improvements, things like that. But you're not going to be doubling in size on the other side of that.

I think that's a very real constraint for a lot of apps and a lot of mobile first products. Though, again, the way we think about it is that's why you see a lot of portfolios where it's, okay, if you accept that as a constraint, the other way to grow is to have multiple products. And if you have three $30 million products, that's just potentially as good and maybe even better because of the diversification benefit of just having one single $90 million revenue product. So I do think it's a real constraint. I don't think it's a constraint on the ability to build great mobile businesses though, and I actually think it could be an opportunity if thought about the right way.

David Barnard:

Yeah, but important thing to think about, especially for those who've raised funding and kind of increased the bar of success is that to understand this eventuality is important. And one of the more interesting ideas on that paid front that I've heard Eric Seufert, gosh, probably like six or eight years ago, talked about how this idea that if there is ultimately a limited ideal customer profile for your app in a limited pool of people, especially willing to pay whatever higher price, which a lot of apps are kind of tending toward now, there is a limited pool of people who care enough to subscribe to an app that does XYZ. And the more you spend and the more you're willing to stretch your CAC, the faster you saturate that pool and you still end up at that same eventuality.

So if today you're growing at 200% and it just feels like you can't spend enough and you're just going so fast, you may just be rushing faster toward that same cliff, not necessarily kind of escaping the laws of thermodynamics, and the speed at which you're currently going isn't necessarily sustainable just because you seem to be accelerating faster than other apps. As you've said, and I'll reiterate, I don't think this is universally true, there are going to be breakout products like a Duolingo, but it's like those are the exceptions that prove the rule that there are just so many spaces that don't have a big enough TAM, don't have enough willing consumers to pay these higher prices that make the return on ad spend work, that you are going to find some ceiling. And for some apps it actually may be lower than that $10-30 million threshold. For some apps it may be a little higher and then you're going to have those that break through. But understanding those dynamics I think is super important in informing strategy, how much you're willing to spend, how much you're willing to stretch out your payback periods, how fast you're willing to invest that revenue and things along those lines.

Patrick Falzon:

And I actually think on that point, one thing to think through is if you actually think about a lot of those exceptions like a Duolingo, even a Spotify is a inherently mobile first product, a common thread I see least is they have big free user bases, or at least for a large part of their life had very big free user bases. I say this as someone who has always prioritized limited free user free usage or free user bases. I'm in favor of even hard paywalls in some examples. But you have to also accept when you're doing that, you are playing into the, "I'm probably capping this 10, 20, 30 million of revenue." There are not very many paid only subscription products outside of the streaming world that really gain traction without that substantial free user base that drives adoption and community and brand.

David Barnard:

Yeah, that's a really good point. How do you think retention plays into all this? Because for an app that does actually have ridiculously good retention, you can still build a great business layering those cohorts year after year after year, where even if you kind of max out the number of users you can acquire, if you're actually retaining a much higher percentage, you can still continue growing at high rates. But that's a challenge.

Patrick Falzon:

Yeah, I do think that's the path forward and I think that's what you should be focused on is one of your most important signals in your business is how do your subscribers retain and what's that average life cycle? And you want it to be higher and longer as much as possible. But even there, I think there's two challenges, coming back to why things cap out. One, it comes back to LTV. Your retention is very obviously an important input to your LTV. And in the mobile app space, great retention is often on a blended basis across all your durations 50% of subscribers every year is like you're doing pretty well if you're doing that. But that also means 50% of your cohort is leaving at the end of that year. So that's, one, going to impact your LTV. If you think about B2B SaaS, they're at sometimes over 100% retention because they're actually upselling. That's a very different world they're living in from an LTV perspective.

And then two, again, if your incoming new users is flat, so your cohort size is flat and you're turning out half of the cohorts every year, you can only grow so much. Now higher retention is better. These are still recurring businesses, and as you get into years 2, 3, 4, 5 for a lot of these cohorts, it's actually much higher than 50% retention is what you're seeing. So you definitely get a stacking benefit as you drag these things out. I do think one way to optimize that is to your earlier point is don't blow the doors off marketing right away and have one giant cohort that's going to slowly adrift over time, be patient and bring in more evenly sized cohorts over a multi-year period and you can actually potentially get much better recurring revenue and a much more evenly spaced out renewal base across various points on that renewal curve so you can get really sticky basis of revenue that are good relative sizes to those new users that are coming in.

David Barnard:

And then part of this whole topic around hitting these caps and the reason folks think about these things is that especially if you've taken any amount of funding, but even if you haven't, I think a lot of people are thinking about the eventuality of a potential exit. So then how an potential exit come into play with these challenges in scale, what do you think the range of potential outcomes exist for subscription apps today?

Patrick Falzon:

So the outliers like Duolingo, you can IPO, that's an awesome strategy if you can execute on it. And setting that one aside, because I do think that's the exception, not the norm, typically what we'd see is there's three buyers of these assets if you want an exit. One is you could sell, there are many portfolio plays in this space, people who are rolling up diversified portfolios of apps, they're always willing to take a look at things. I think they're a very willing seller. Probably your most realistic seller. Their business is buying and selling apps, so they're probably going to be more regimented on valuation than some of your other buyers because their whole thing is they have to get good ROI in every single deal for their business model to work, so they're going to be pretty strict on their valuations. Second is you can sell to private equity. There are some small scale private equity players who are interested in this space.

I think the key there though is it's hard to sell them a single app. So if you're just a single app developer, you're probably looking at the app roll of portfolios. If you've built a portfolio of apps, you can be able to hand off almost more of like a business and then a product to those private equity players and you can say, "Hey, I have these three apps and I have a consolidated tech stack behind it and this is a great platform that you can go out and add more products to." That's a story and a narrative that can get you into the lower end of private equity and then get an exit there. And then lastly, for certain apps, I think strategic buyers are very real. You saw Under Armour, what has that been, five, 10 years ago at this point, did a bunch of acquisitions in this space. I think you can see similar things where there's a strategic who has a strong rationale for wanting to buy your product.

That's probably for any given product, there's going to be a handful of companies that you'll know will make sense for that. But you can have an eye towards trying to establish relationships or making yourself visible to them to set up a potential exit. And then the last one I always advise people is also I don't think everyone has to exit per se. Again, part of the benefit is these are recurring businesses. You can build up a sticky renewal base and maybe the right thing for you is to actually just at some point shift a little bit more into maintenance mode where if what you're trying to do is spend more time on other things, you do that, but you still take your cashflow from maintaining a product and continuing to provide the value that those users thought they were getting when they purchased the product. And then I think in the mobile app space, you could do that with a reasonable time commitment that still allows you to go do other things.

David Barnard:

Yeah, I wanted to step back to the strategic buyer. It seems like we saw a lot more of this early on. I don't feel like I've seen as much of this recently, although maybe I'm just not paying enough attention to the M&A space, but have you seen any kind of strategic acquisitions that you think are interesting? Or do you see specific categories and companies that might make sense for these kind of strategic purchases?

Patrick Falzon:

Yeah, I do think you can. So actually the budgeting space is one example. Rocket Mortgage bought, I think it was Truebill at the time, and now it's Rocket Money. Because for them, that's actually, to the reason why we never found the category attractive, it's a great customer acquisition funnel for them of getting people into their budgeting tool and then selling them actual financial products on the other side of that. So I think financial services is definitely one where you could continue to see that. I think privacy security is another area where you have some big consumer brands like the Norton's of the world and the McAfee's and things like that, where they could still be interested in maybe increasing their mobile presence or their mobile brand. Those are the two big ones that stand out to me.

And I also think health and wellness, although to your point, I think that's one where we've seen that fade recently. And I generally agree with your classification here, which is if you look back 10 or 15 years, mobile was this exciting new platform and was like desktop first mobile. I think we're kind of on the other side of that where mobile is more of a standard and just kind of baked into everything everyone does, and most people have some degree of mobile and desktop presence. So I do think you had a rush early on where it was every big company felt like they had to have a mobile play and they were using M&A as a way to accomplish that. And now I think it actually is more tightly tied to product fit and, "Does this product make sense? Does it generate us more revenue? Does it help lower our customer acquisition cost?" Not just mobile for the sake of mobile, which you got more of in 15 years ago.

David Barnard:

Yeah, and two recent acquisitions that I think point to that is both Under Armour and The Weather Channel selling to Francisco Partners, a private equity firm. Where both of those, I would have guessed, had you asked me three years ago who would buy something like that... I mean famously IBM actually bought The Weather Channel, and from what I understand, they still maintain the B2B side of the weather business and then Under Armour having bought MyFitnessPal-

Patrick Falzon:

And a few others.

David Barnard:

And a few others. And then they spun it out and now it's landed at private equity, it's like MyFitnessPal kind of being a darling of the health and wellness space. It's like if somebody was going to get a strategic acquisition by a bigger player, that's one that I would've guessed would and they didn't. So those are I think two data points showing maybe the strategic acquisition space is limited. But then one other data point on the other side is Ziff Davis acquired Lose It! recently. That is in the health and wellness space. They're a big media conglomerate. So it's like there are opportunities, but yeah, it seems very different in 2024 than it has been.

Patrick Falzon:

Yeah, it's going to be a very small universe. Typically when we would do... I also worked on sales out of our portfolio at IEC. When we'd make buyers list, for most companies, your strategic buyers are like five to 10 potential, and then you have a list of 100 financial buyers who could potentially be interested in it. And I think that's going to stay true here where it's a very small population of potential strategics, and then a much larger population of more financially motivated buyers.

David Barnard:

Yeah, I think few people listening to this podcast don't have some inkling in the back of their minds, like an acquisition or some kind of bigger liquidity event would be an interesting prospect in the future. And so understanding those dynamics I think is important.

Patrick Falzon:

The other thing I'd advise is if that is what you're optimizing for or if that's what you seriously want the end goal to be, build towards that goal, don't just assume it's going to happen. Be thoughtful around, "Who is a potential buyer? What can I do to make this more interesting to that buyer?" Or even if you haven't built a thing yet and you want to build a thing to get acquired, have that be part of your selection criteria of what you're going to build is, "What does the acquisition pool look like for this thing? And therefore, do I want to build it or not that?" Both from a product perspective. And then I'd also say from an operational perspective, the cleaner and more professionalized your setup is the easier it's going to be to find a buyer for that. Buyers don't love messiness. They're already taking on a lot of risk by doing an upfront payment for an uncertain outcome on the other side. Anything you could do to reduce that messiness for them is great.

David Barnard:

Yeah. Any other advice on opening up those potential doors? Strategic or PE? I mean, imagine if you're building toward a private equity exit and you think that's kind of the optimal return, then building something that's throwing off cash and being more careful, more disciplined with spend, being more disciplined with expenses, like showing that profitability or the opportunity for profitability is something that is going to be way more valued by private equity. Again, maybe you shouldn't be building exclusively this direction because we know it's such a limited pool of potential outcomes, but if you are hoping for a strategic acquisition, would you then maybe be trying to build more of a brand or have a unique take or any advice on that spectrum of building toward a hopeful exit?

Patrick Falzon:

Yeah, I do think, particularly if you're going to be going to a financial buyer, having a track record that they can understand and project out and say, "If I just do more of that, this is going to go well for me," is important. So for example, you've been running unprofitably your entire life and then you go to them and say, "But if you buy this, it's going to become profitable." That's just a harder hill for them to get over as opposed to saying, "Look, it's already been profitable for the last three years, just keep doing XYZ and this is going to go well for you. That's a much easier narrative to sell them.

So I do think being to a point where you are either already profitable or your uniquenomics are clearly going to drive profitability in the future, even if right now it's not immediately profitable, that is great for the financial buyers. It does probably matter a bit less for the strategic buyers. And I do think brand and maybe even just the size of your user base matters more for strategics. I think for example, strategics will always place more value on free users and community than a financial buyer will. So there are differences there in terms of what they're going to value.

David Barnard:

Yeah. Well, this has been fascinating. I mean, picking your brain, having been in this space for so long and looking at so many apps over such a long period of time, both, I didn't even think about that, you would be working on selling parts of IAC's portfolio while also potentially acquiring others. So you've been both on the buy side and the sales side of this in a really interesting way and across many years of ups and downs of the industry. So yeah, that was really fascinating. Thanks for sharing all those insights.

So I did want to dive into the more tactical side of things that you learned running as a GM of iTranslate and Robokiller. The first one there is how to think about user acquisition, LTV, CAC, paybacks. I mean, it's just such a big hairy mess. And I mean it's actually something we're thinking a lot about at RevenueCat. We now do have some LTV predictions. We're kind of, teaser, going to be adding those to our experiments feature. But it's tricky and we've actually been really careful internally and had a lot of thought around how we build these products because it's really easy to just throw a PLTV stat out there, "But what does it actually mean? Can I trust it? Can I buy against it?" Is a really important question. So yeah, I'd love to hear your thoughts around how to think about these metrics.

Patrick Falzon:

LTV is hard. Particularly if you're in the early days, what I advise people to do is think more about your range of possible outcomes because the reality is you have very limited data. That data is not going to tell you a perfect story on the other side of it and say, "With 100% confidence, this cohort's LTV is going to be X." You're getting a directional estimate at best. So the way we do it in the early days of a product and what I advise people we work with to do is run a range of scenarios and say, "Plus or minus 20% in maybe 5% increments, what are my possible LTVs?"

Assign some probabilities to them, come up with your best estimate. Start making decisions and acting on that, and then just go back and check it every month or two months as you're getting that retention data in. Start validating. What you could hopefully start doing is if you start with, "Hey, I have six possible LTVs, then I'm giving some weighting to each of them," then you can get it down to four LTVs over time and then two, then three or however that goes, and eventually get to a more precise LTV estimate that you're going to have. But you're going to need maturity just from a scale perspective to get to a relatively precise LTV, you need multiple years of data realistically to be able to reliably predict that. So that's okay in my mind. You just have to again know where you're at in that cycle. Are you in the less certain and therefore you need to be more flexible and dynamic, or are you in the more certain side of your cycle where you're more mature and you can have more of a singular target of how you think about things?

David Barnard:

And maybe this is one of those things you need to pick a point in time and say, "What's my LTV going to be at one year? What's my LTV going to be at two years?" And kind of more cash flow prediction because LTV is such a loaded term that I don't even like in the subscription app industry in that I've seen with my own app, I launched my dinky little side project weather app, launched it seven years ago, and still to this day I've retained like 10, 15% of the original cohort who bought it seven years ago. And so trying to estimate that into the future and then saying, "My seven-year LTV is going to be X," you can't actually calculate the average lifetime value of a user because it can go decades into the future.

I mean, there's products like Netflix. I subscribed to Netflix as a DVD customer in 2000, what, four or something? 2001? And then I transitioned from DVDs to their streaming service, and I've been a monthly paid customer for close to 20 years now. I think that's where the rest of the industry is going. Like a weather app, whether it's on AR glasses 10 years from now, I'm probably going to have users that have 20-year lifetimes, 30-year lifetimes. Even my simple little dinky side project weather app. I'm ranting here, but it's like I don't like LTV as a term because of that. Because nobody really... You can't actually estimate an average lifetime value of a user, but you need some metric to buy against for cashflow. And that's where...

Actually this is something Dan, he is a colleague of mine at RevenueCat, formerly a colleague of yours at Teltech, at Robokiller, and he brought to our charts this called Realized LTV. The way I think about it is like day X realized LTV because in the RevenueCat charts you can actually pick, "How much revenue have I made on this cohort at 30 days, at 90 days, at a year?" And there's a little drop-down that you can select it. And to me that's just such a more helpful way to think about quote-unquote, "lifetime" value is more like pick different time points and do those calculations toward, "What's the range of expected revenue at 12 months, at 24 months at 36 months?" Because it just feels like that's so much more helpful than thinking of it as this, "This is all the money I'm ever going to make from this cohort and here's the line I'm drawing in the sand."

Patrick Falzon:

Totally agree. A couple thoughts there. One, we always did cap our LTVs. So to your point, maybe calling them lifetimes is not actually a accurate way of framing that, but we typically would cap them at a certain number of years for two reasons. One is you have to get paid back at some point, so you have to care about your payback period. If you're buying against a 20 year LTV, maybe you're not even making a profit year 15, and that's not a super functional business model to be managing. And then two is, not to get too finance nerd on people, there's a time value of money. So a dollar in 20 years is worth less than a dollar today. So even if you're still having some of your cohort renewing and paying you in 20 years, if you discount that back to what is that actually worth to me today, it may be very minimal and therefore not actually worth baking into your LTV.

So we did cap it. I would advise people to cap it. And then also to your point, I think coming back to where are you at in your maturity life cycle, you also need to think about payback periods. So typically what we say is when you're earlier in your life cycle, you can't be waiting 12 months to make a profit on a cohort. You need to have tighter payback windows because you just don't have a lot of cash on hand to spend. One year later in your life and you have more money in the bank, you could potentially think about if maybe you started a two year LTV, maybe you start expanding that to a three or a four year LTV and being okay having a 12 plus month payback.

Again, I think you can play with those levers as time goes on, but I would definitely advise capping your LTV. It just will make your life simpler and easier and it's at worst you're being conservative and that's not a terrible thing to do. And that I would think about payback periods as almost like a separate and distinct metric from your LTV to CAC or your ROAS.

David Barnard:

But then how does that work out practically? In what time frames would you cap your LTV? Would you look at a particular app and say, "Based on the cash flow of this app, we need ROAS at day 90 and so we're going to predict..." I mean, would you call that a predicted... Again, it's like you're not predicting lifetime value, so you're predicting cash flow within 90 days, not an actual lifetime. How did you do those specific kind of projections? "What's our cash flow going to be at nine months, 90 days, at nine months at 12 months?" How did you think about that tactically?

Patrick Falzon:

Yeah, so we did have an in-house proprietary system forecasting LTVs. Again, we benefited from being more mature, so we had a ton of data on all of our products, where if we got an initial renewal point or initial two renewal points from a given cohort, we could pretty reliably predict out then what the next 20 would look like from there. And also what you see is typically in those first couple of renewals is actually where you see the bulk of your churn and then it starts flatlining after that.

So we were able to pretty reliably predict those and then we manage on LTV to CAC predominantly. I think of them as two separate ways. I think you either have a primary filter of, "What's my payback period?" And basically I'm at a stage in my company's life cycle where I need to make my money back either in the first month or in the first three months, and therefore that's the primary thing I'm optimizing for. And then secondarily, I'm looking at LTV to CAC to understand what that looks like. Because what you may find is, "Okay, I need to get paid back same month," but then what I see is I have a crazy high LTV to CAC. So once I solve my cash liquidity problem and I can extend that payback period, I now know that I can spend a lot more money and still have a healthy LTV to CAC. So then you can think about what do you need to do from a business perspective to unlock growth, which is find liquidity somehow.

David Barnard:

Right.

Patrick Falzon:

Alternatively, later in life, you're going to be probably more focused on LTV to CAC and say, "Okay, I have a threshold. It has to be above a certain LTV to CAC," and your payback period is almost more of a sanity check on this side. "Okay, If I'm targeting a 2 or 3x LTV to CAC," I want to know how long it's taking to get me paid back. But that's probably not the thing that's driving your business decisions. It's just more of a nice to know on the side.

David Barnard:

Yeah. Did you have any particular moments where specific new ad spend or even specific new ad creatives or new prices or any specific changes that you saw break your model? Did you have moments where like, "Oh, we ramped up spend on Meta and everything looked great, LTVs were amazing," and then you do the three month look back and you're like, "Oh crap, we were 25% under what we predicted. This cohort sucked. We shouldn't keep acquiring these users in that way," do you have any specific examples like that?

Patrick Falzon:

A lot is the short answer. So it's definitely a thing. A couple of quick examples. So on price testing, that was always an issue. We generally skewed either annual or monthly subscriptions or some combination of the two. You can pretty easily get a sense of what's going to happen to your monthly subscriber retention when you're running a price test if you're willing to wait a little bit of time. You're obviously not going to have data on what happens to your annual subscribers retention though when you're making that decision. No one's reasonably waiting over a year to make a call on a price test. So you have to make some simplifying assumptions there.

We typically what we would do is look at the impact on monthly subscriber retention and say, in theory, the annual retention impact should approximate what we're seeing in monthly subscribers. That was not always true though. And in particular, if one is clearly your primary plan and one is your secondary plan, so it's like a 70-30 split between those durations, they won't actually always work the same. So we often end up with price tests, particularly if they were overweight annual subscriptions. Sometimes we'd make calls and then you'd see those cohorts come in a year plus later and you'd be like, "Yeah, that maybe wasn't the right call, unfortunately," or you're very pleasantly surprised and it was a better call than you even anticipated.

I don't have a great solution for avoiding that. What I typically advise people is you're going to look at your A/B test results and just sensitize it to say, "If retention comes in 10% worse, does it change my answer of if I should roll this out or not? If it comes in 20% worse, does it change my answer?" And then make a decision based on your relative comfort level there. If you only have a 5% buffer on retention before this price test winner is no longer a winner, maybe don't roll that one out or maybe be more cautious about how you roll that one out. If retention can go down by 50% and that thing's still a winner, close your eyes and roll it out. You're going to be fine with any reasonable outcomes there.

So I would say sensitize price test is a big learning we've had, particularly on your annual subscriptions. And then new marketing channels is definitely a thing. I think a couple implications of that are when you're ramping up on or you're launching a new channel or you're ramping up on a new channel, be a little conservative. Don't go full bore. My experience is actual performance almost always contracts from that initial predicted performance when you're doing a new channel. Because a combination of retention is probably going to come in a little bit worse than you're expecting because that's just life, and those first handful of users you acquire are almost always the cheapest users you're going to acquire.

And eventually it's going to start getting more expensive and eventually it's going to start contracting on you. So patience is important there. And give it a little bit to play out before you just say, "I want to pour gas on the fire." I do see sometimes people are a little too quick to, in their minds, pour gas on a fire that may not be real. So I do think you have to think about that. But it's I think just something you have to factor into your decision-making. And it's just a reality of the world we live in.

David Barnard:

Yeah. How much did you think about the impact and how much did you see the impact of price on those longer retention periods? So I feel like it's a bit of a trend, and I've talked about it a lot on the podcast, I actually talked about it on the episode just before this one, that these much higher subscription prices do work really well to get that payback quickly, like this shift toward annual subscriptions. And then more specifically, the shift toward higher priced annual subscriptions is closing the loop on being able to get that seven-day return on ad spend for some apps that seem to be doing really well and finding the right creatives and stuff like that. But then just inherently, the higher your price, there's going to be some impact on churn. And how do you think about balancing those things? And what did you see in the data as you looked at some of these bigger cohorts in the Mosaic portfolio?

Patrick Falzon:

Retention as a general rule decreases as price increases, so you should think of those as inversely correlated. Basic logic holds, if someone's paying more for something, they're more sensitive to the value they're getting out of it. So we did see that on the whole, it is definitely not a one-to-one relationship, meaning you're going to not get LTV wins of the time, assuming conversion rates line up. So we typically see if you increase price 10%, maybe retention comes down by a couple percent, for example.

And then it's also not one size fits all. So some products saw it more than others, and particularly I always advocate for being really honest with yourself around what your product does and the value it provides. I think there are a lot of very successful and very profitable and great products in the mobile space that provide a somewhat thin layer of value. And I don't say that with any negativity associated with it. They provide a distinct value, but it's thin, and you built a great business around that. But be realistic about that and know that there's then probably a lower ceiling on how much you can charge for that value.

There are other products that are like Duolingo or language learning is potentially an example there where it's like there's a really deep corpus of content underlying them. You're truly teaching someone a skill, a thing that they pay in the real world maybe thousands of dollars to go somewhere to learn. You have more price power there. And so be honest with yourself about how much value you're providing. There aren't necessarily wrong answers there. The most important thing is just be honest with yourself.

David Barnard:

Yeah. And were there specific leading indicators that you would look at to understand how price... Especially on annuals, because again, so much of the industry is shifting toward annual to get those earlier paybacks. But then it lengthens the feedback loop. Were you feeding into the payback model and the retention model and the A/B test results, early turning off auto-renew, engagement data, what were the factors that helped you get a window into the future performance when the real data was going to be 12 months out?

Patrick Falzon:

Yeah, so a couple of things. So like I said, if you have a shorter term subscription, like a monthly subscription, seeing the impact on that retention from a similar price increase and being able to extrapolate it out. Similarly, if you've raised price before and you said, "Okay, last time I raised an annual price 10%, this is what happened to retention," using that as a indicator of what could happen in the future works. Also, if you have a free trial, trial to paid conversion is typically going to be a leading indicator of retention on the other side. So if you're seeing a drop in trial to paid, you're probably going to see a drop in retention. Not necessarily one-to-one again, but it's a leading indicator.

We did not see a lot of success using early cancellation rates as a predictor of churn. We did a lot of testing around it. Only a fraction of users actually cancel early. What you see is you get a lot of cancellations right around the renewal point when they're getting those emails letting them know that it's coming up. So we did not end up... We used it. It was not a particularly valuable data point for us. I think engagement can be a valuable data point if you have the clarity and the tracking and the good data around those features.

So if you have a key action the user's going to take in your app or a key feature that they need to use, being able to track that and making sure people are still doing that and still doing that on a recurring basis, I think can be a good indicator, is definitely good indicator. That one gets trickier to feed back into something like an LTV algorithm though. So that's I think the challenge there. Oftentimes what I see is that's almost used as a thing that sits on the side where you have your LTV algorithm and then you're looking at your cohorts and are they doing that action and then you're making some in your head like mental adjustment to what your LTV algorithm is saying because it's just hard to feed those back in.

David Barnard:

No, that makes a lot of sense. So it's like, "If the LTV is saying this but we see usage retention differing from past cohorts by 10%, then we should discount that LTV 10%." I mean something I've talked about before in a talk I gave and we have talked about on the podcast, but I'd love your insight on, are there specific things that are going to be the most meaningful to track to get that sense? Do you need to pick one specific core user flow, one specific key value that the app delivers? I mean I guess every app's going to have their own thing. Any tips on picking that user action that's going to most strongly correlate with retention, which then is going to help more strongly correlate with LTV and your financial projections?

Patrick Falzon:

Yes, we tried to have one or two max for every product. I think if you have seven, then you're creating noise, then the actual signal's getting lost in that noise. So really think about the one or two key things. So for example, Robokiller, it was actually less about what was the user doing? It was more about are they actually getting spam calls and are we blocking calls on their behalf and trying to understand-

David Barnard:

The passive usage.

Patrick Falzon:

Yeah, "Are we providing the value that the user came to us for?" And that may not be dependent on an action they've taken. So that's the lens I would suggest to applying is it's more about the, is your product providing the value? That may be because we have another app called TapeACall, which is the call recorder. The value there is recording a call, that has to be a user action. So there it is a user action. Did they actually record a call? But again, the Robokiller example, it's not actually a user action you're looking for, it's just a thing that happens in their environment and you're confirming that they are getting spam calls, we are blocking them on their behalf and therefore they're getting value out of it.

And actually even there, we had two metrics of are they getting spam calls and are we blocking them? And then also did we get one wrong? And for us, that was actually the big flag for us of when we needed to do something proactive about retention, but if we saw we got something wrong for our user, that was highly correlated with churn. So then we needed to remediate that in some way, shape or form.

David Barnard:

Probably a good exercise for any app that hasn't already done this is do some user surveys, do some user interviews, pump all your reviews into ChatGPT and figure out why are people actually paying for the app? What really is the value? And I like that one. The Robokiller example is so good, is that people aren't paying to interact with the app, they're paying to not interact with the app. They're paying for the app to sit in the background and block those calls.

Patrick Falzon:

At times an inverse correlation between how much you used Robokiller and how happy you were with the product. Where it's like if you were using it a lot, you actually probably were a less satisfied user because you were trying to do all these tweaks and coming in to check things, "Did we really get it right?" Our happiest users kind of forgot we were installed on their phone in a good way.

David Barnard:

Even in that case, this is maybe a rabbit hole to the current conversation, but were there ways in which the user did see Robokiller taking action? Would it pop up a notification like, "Robokiller blocked this call," or you still have to kind of reinforce the value that you're delivering like, "Hey, we blocked 50 calls for you this week alone", and reminding people that how much value they're actually getting delivered when it's just sitting there in the background passively delivering value?

Patrick Falzon:

It was a challenge for us for a couple of technical reasons, namely around Apple limitations around the data you do and don't have on a user level and then also our own privacy layer we had put on top of that. But think of creative ways around that. So one thing we did is we generally de-anonymized like all data because we thought that was the right thing to do. So instead of giving you necessarily personalized stats, we would give you stats across the whole user base of, "We've blocked this many millions of calls this month for our users." So still being able to demonstrate we are doing things, we are providing value even if we can't perfectly tie that to you as an individual user.

David Barnard:

All right, so we've kind of been dancing around the acquisition side of things in talking about LTV and measuring your LTV and figuring out how your users are engaging with the value of your product. But then you do have to go out and acquire those users. So how did you think about marketing and customer acquisition costs and keeping that in check, especially in relation to LTV?

Patrick Falzon:

Yeah, we did think about the world in organic versus paid were two lenses we put on it. And then I think you can think about your CAC then both from a blended perspective, so basically accounting for your organic traction, and a paid only perspective. And I would strongly recommend that you look at both of those metrics because there are apps that you'll see that are unprofitable if you look at them on a paid only basis but profitable on a blended basis, which is great and that may be the right decision for your app. But that probably means you don't have a lot of growth ahead of you. The more you spend on paid, the more negative that's going to go and it's just going to drag down your margins. So that's why it's important to understand it because it influences what you can do going forward.

David Barnard:

But looking at how paid impacts the blended, because for some apps there is a little bit of a network effect where if you spend more, it's also going to juice your organic. So if you have any kind of network effect, spending more maybe does balance out. But to the point we've been making throughout this whole podcast is that's probably the exception, and for most apps you're not going to have enough of a network effect. Some of the things that do come into play here though is that as you spend more and bring in more users and get more reviews, you do push yourself up the search results. And so that's an important aspect. But to your point earlier, it's like you also cap that out. It's like if you can't spend more than a million dollars a month, maybe you peek at number three result for the most valuable keyword and you just can't push it beyond that. So kind of watching the interplay between paid and blended is probably how you determine where that's going to land.

Patrick Falzon:

And I do think that's a very real effect in the mobile app ecosystem. I think it has diminishing returns though. So I think early in your life there is a very strong correlation between paid traction and organic growth, where as you drive more paid downloads you can absolutely increase your organic installs along the way. At some point you will start to see that flat line though and you'll get diminishing returns and for every paid install you're getting you're getting a smaller and smaller fraction of an incremental organic install as a result. So again, understanding that trade-off and where you are on that curve is important. Because then what you're thinking about is, "Is all that matters my LTV to CAC on that incremental dollar of paid spend? Or should I actually be discounting that or maybe applying a premium to that LTV to CAC because I know I'm getting some incremental organic benefit on the other side?"

So I do think that's very important. I also think, to the organic side of things, like ASO is very critical. I think it is... Based on some of the data I've seen, it's like over 50% of installs come through search and go to one of the top three rankings on search. So it is really critical that you, if you're going to launch a product, understand what are the relevant keywords for your product, understand what is your ability to rank on those keywords. This comes back to when you're thinking about a category in the competition. If you say, "I need to rank on this keyword," and the first three keywords are like $100 million companies with massive budgets, that's probably a concern that you need to think through and think through how you're going to handle that.

David Barnard:

So you're saying I shouldn't build a weather app.

Patrick Falzon:

Well, it seems to be a category that's done well regardless.

David Barnard:

Yeah.

Patrick Falzon:

But I do think that that's real where it's you have to be really thoughtful about what are the keywords that matter and then measure where you're at on those keywords. Because what you want to understand is, one, what's your growth opportunity and how realistic is you can achieve that. But then also that tells you where you are on the curve of diminishing returns on your organic lift because of paid spend. If you're already ranking one, two, three on the keywords you care about, there's only so much more upside you're going to get from your more paid spend other than just the economics on that paid spend. So I think that being keyword aware is very important.

David Barnard:

I think this goes to a kind of broader strategy though, is that I think any app needs to have some leverage in user acquisition. I've talked about this a lot and we're talking about it through these two specific lenses, ASO and paid spend. I do think there's maybe 10 other potential lenses to look at it through AllTrails being a really famous example, it's like they didn't have to rely as much on Apple search, although they got featured a ton, they probably did get a ton of Apple search, but they had a user acquisition advantage in SEO on the web. If you go search for any trail and they're going to show up.

I think there are advantages you can build outside of ASO and paid spend, but you need to understand what that advantage is and what you're building toward. And for a lot of apps, and especially again like you said at the very beginning of the podcast, if you're trying to build a new category that people aren't searching for, it is a really hard uphill battle because you got to figure out something outside of the app store that's going to drive people, whether it's like TikTok organic, whether it's SEO, whether it's some advantage in paid spend that you're building something unique, but it is hard to go build a real business in the app store around something that people aren't searching for. And I think a lot of people discount just how much some of these bigger apps benefited from the Calms, the Headspaces, the MyFitnessPals, Lose It! famously. I mean they were at the top result for decade plus, one of the earliest calorie counting apps and rode that ASO wave into this massive business that they are today.

I think that's discounted a lot or not kind of fully understood when you're trying to compete in that calorie counting space or other spaces, the incumbents have this advantage in that search that is nearly impossible to break in 2024. So what can you bring to the space and leverage in user acquisition outside of ASO or paid if that's going to be really hard to compete in? Or do you have some advantage in paid because ASO is going to be really tough to crack in these bigger categories?

Patrick Falzon:

Agree with all of that. So we treated things like content and SEO as organic in our view. So ASO was kind of one leg of the organic stool. But totally agree, you couldn't get real user acquisition through other leg. So content was definitely one that worked for us. But our best example was for Robokiller we were the biggest app in the space and in the US so therefore we had a lot of data on spam calls and texts in the US, and we started publishing reports and working with reporters to basically help them report on the issue to say, "Here's how many spam calls Americans are getting every month, here's how it's growing." We worked with the FCC and the FTC to get them data too because they didn't have great data on all of this.

So we found that we had this proprietary data set, people were interested in talking about it and it wasn't a direct, "We talked to this reporter and we're going to get a bunch of users." But when you start accumulating multiple Wall Street Journal articles that reference you and multiple Financial Times articles that reference you and you're getting placements on cable news and things like that, that starts building a brand and organic halo and you could absolutely drive a significant user acquisition through that. So we saw a lot of success with that. To the extent you can replicate it, definitely do that.

And then community is another one that I think could be very impactful if you could build a community. I think it's really hard, but I think it can be very impactful. And I think maybe not the most traditional sense of it, but Duolingo, I think that was a big part of their early growth was they made it largely free. It was very gamified. It became this thing that you talked about you were doing, you did it visibly on the subway to work, people saw you doing it and they created this real brand awareness by building a very large community of unmonetized users and then they figured out the monetization in a much more meaningful way on the other side of that. So I do think when you think about breakout stories, I actually think the most common trend between them is they have found some very meaningful channel of organic acquisition outside of app store search and that's really what's helped them break through that ceiling we were talking about before.

David Barnard:

Yeah, no, it totally makes sense. So kind of took us down multiple rabbit holes. But back to brass tacks tactics on how to think about customer acquisition costs. What are the specific things you're looking at there? So we talked about kind of blended and paid CAC, but then what are some of the formulas and other things you look at there?

Patrick Falzon:

I think it's important to look at it across your funnel. There's a couple of different intersection points you could think about things. Generally what we see is you really thinking about a cost per install, a cost per free trial or cost per subscriber. I would pick one of those as your primary metric and that should tie to your monetization strategy. So for example, if you're a free app that's mostly advertising driven, you probably want to think about a CPI, because you don't have a free trial or subscribers, so therefore those aren't relevant metrics to you. If you don't use free trials, cost per free trial isn't super relevant. But if you just go into direct subscriptions, think about that.

If you use free trials, it may benefit you to think about a cost per free trial because because that's a more dynamic quicker metric. You don't have to wait seven days for them to convert out of the free trial to figure out what your cost per subscriber is. So just think about it on a cost per trial basis. Obviously you have to make sure that your LTV matches whatever you're using as the denominator for your cost. But I think about where in the funnel do you want to measure your acquisition cost is important. And like I said, then you think about it as paid and blended. And then even within paid, think about channel specific and acquisition costs. So understanding is Meta more expensive than Google or TikTok, for example, and then understanding where can you pull levers and increase your spend on the other side of those.

And then I think more broadly tactically where a lot of paid spend has trended in the mobile space is largely black box algorithms. If you think of Google and Meta are the two biggest platforms you're able to spend on, ASA as well, but both Google and Meta basically set a bid and a budget and that's the degree of control you have over those platforms. So it becomes a huge lever, and I'm certainly not the first person to say this, creative is a very big lever and you need to be really focused on it. So a couple of things we say there is always be testing creatives. You should always have some... If you think about the creatives in your campaign and there's different structures how you do this, but there should be some percent that are experimental, that are unproven and you're just always putting new ones out there to see what resonates. And then two, one tactical tip I give people a lot is rotate designers on your project.

Unsurprisingly, everyone has, no matter what your profession is, your own distinct style, certain engineers have a certain way that they code things and the way their mind works. Finance guy background, my financial models had a very consistent theme and flow to them in terms of how my brain worked and how that translated into a model. All designers have a pretty unique and personal theme or the things that they like to lean into. And if you keep using the same designer, what you find is you get in this trap of incremental improvements where what you're really doing either changing 5% or 10% of a thing to see if it works a little bit better. And sometimes it's actually better just to bring in a new designer and just be like, "Hey, this is the product. Go wild with it." Obviously you got to have brand guidelines around it and it's got to be relatively consistent, but actually try something wholesale new and see how that performs.

So we tried to get a rotation of we'd actually for our creative teams rotate them across products. So you'd work on this product for a couple of months and then you'd rotate to another product for a couple of months. And that way got us more diversity of the types of creatives that were flowing through there. And then so we could pick the winners out and start understanding, "Okay, well this style is resonating, let's do more of that style," as opposed to the style almost being like a fixed thing because it's just the designer style.

David Barnard:

Now this is a huge can of worms, and we do need to wrap up. So I'll ask for the top tips and the two to three minute answer, but you got to ride the wave of ATT at Mosaic. So we've been talking CAC and measuring your paid spend. The question people will be shouting back is like, "Okay, cool, those calculations are great," but any top tips on actually putting pen to paper and measuring these things, or keyboard to spreadsheet, as it were? How do you actually meaningfully measure both paid spend return and organic return?

Patrick Falzon:

The quickest answer is run multiple analyses and look at the aggregate of those analyses. So the way I think about in the post ATT world is like you still have MMPs, they're still going to give you some degree of a breakdown of your acquisition by channels. Take that at face value, understand what that says. Then also do a blended analysis and say, "What does my blended look like? Am I still okay with blended?" Similarly I just do paid versus organic and say there's maybe some mix up between channels, "So on a holistic paid basis what the MMP is telling me, do I feel good about that?" And so aggregating across those. I don't think anyone can rely on one singular dashboard. I yearn for the days of pre-ATT where I had a Looker report that I fully trusted and I just knew that this specific geo campaign on Google was doing this. But those days are gone and you have to accept that they're gone in my view at least. So you do that and then also think about operationally how and when you launch and/or scale things.

So what I mean by that is if you say, "Okay, we don't do any paid spend now and we want to launch on Google, Meta and TikTok," don't do all through it once. Just launch on Google and then understand, "Okay, all we've changed is we're now spending on Google. What's the impact of that?" Get Google to a steady state and then say, "Okay, now I'm going to launch on Meta," and now you have a new baseline and all you've done is change one thing. If you launch all three at once, it's going to be very hard to understand the relative performance of each and what's going on there. So always think about doing things very incrementally. And same thing if you've already launched and you want to start changing bids and budgets, change one bid, give it a month, see what happens, then change another bid. And you have to give yourself time to establish baselines that you can do that incrementality analysis of, "I made a change and now what am I seeing in my aggregate numbers as a result of that?"

David Barnard:

Yeah, that's super helpful and probably the most helpful thing is for folks to just hear it's hard.

Patrick Falzon:

Yeah, it's hard. You're always kind of guessing. And at the end of the day, I ground myself in blended performance, which is... Because your P&L is inherently a blended mix of everything you're doing, and at the end of the day, your P&L is what has to make sense, what has to work. So that should be your north star and your grounding light of, "Does the blended make sense?" And if the answer is no, you got to go figure that out. And then as long as the blended is making sense, then attempt to better understand the dynamics under there. But always be grounded in the blended has to make sense.

David Barnard:

Yeah. Well, I think that's a great place to wrap up. I know people would be hoping for a magic answer. I feel like I get asked this all the time, "How do I measure? Which M&P do I use?" And the answer is like, it's just hard and everybody's got to figure it out.

Patrick Falzon:

Tell Apple to reverse their change.

David Barnard:

Well, I've been pushing hard on improving scan. I think Apple... Well, what is it called now? I don't even remember the new acronym for it. But I don't think Apple even has to reverse ATT to get back to better. They just need to better incentivize the industry to use whatever it is now, not SKI network, but whatever it is. We could get to a much better place on reporting if they would provide the tools and be a little less precious about anonymity. Because at scale, some tiny amount of potential for user data leaking, the way they built their tools is very ivory tower, zero personal data can leak.

That's just not how the real world works. Because if data on two or five users is potentially able to be de-anonymized at scale, it doesn't matter. Nobody's going to bother de-anonymizing those handful of users when there's millions of data points flowing through the system. I'm still hopeful. I mean obviously Apple's never going to put the ATT genie back in the bottle. So my hope is actually that they just continue iterating on those tools to make it better and more deterministic over time by improving those systems, which I think there's tons of head room for them to do. They've just been very slow to do it.

Patrick Falzon:

It's also a great opportunity for the marketing channels themselves to start differentiating. Here recently, Meta is starting to pull away in terms of relative performance because they've invested a lot more in how to operate in a post-ATT world. So I think what you could start seeing is some marketing platforms invest more in that than others and that's where more of the dollars start flowing to.

David Barnard:

All right, well this has been so much fu., but as we're wrapping up, anything you wanted to share? I know you're working on some new projects and anything you wanted to share about those?

Patrick Falzon:

Yeah, I've started a new venture, it's called The App Shop. It's myself and some old colleagues actually from the Mosaic Group. The gist of what we're doing is taking all of our lessons and experiences and learnings from all of those years building Mosaic and offering it up to other companies playing in the space. We love the mobile app ecosystem. We want to see it grow. We're largely doing consulting and agency services, can help with marketing monetization and product development. If anyone's interested, we'd love for you to reach out, our website's theappshop.io.

David Barnard:

Awesome. Well, Patrick, thanks so much for joining me. It was so much fun.

Patrick Falzon:

Thank you.

David Barnard:

Thanks so much for listening. If you have a minute, please leave a review in your favorite podcast player. You can also stop by chat.subclub.com to join our private community.