On Capturing The Value You Create

Imagine I come to you, and I offer you a deal. You pay me $1, and I’ll find a way to get you $100. Good deal, right? What about $99 and $100? What about $99.99 and $100? Are those still good deals? Your answer will depend on 1. opportunity cost 2. cost of capital 3. counterparty risk 4. whether it’s accretive or dilutive to the margins of the rest of your business.

SaaS and social network business can look a bit like that first deal, because you don’t get to capture much of the value you create. I used to run a SaaS business where we charged customers 0.1-0.5% of their revenue, in exchange for boosting their revenue 5% (well, hopefully anyway). I’m not sure what the benchmark here is, but capturing 2-10% of the value you create seems fairly standard for many SaaS companies. How much do law firms pay for their Microsoft Office subscriptions, and how much value does Microsoft create for them? Similarly, people on X are always saying “I can’t believe this website is free” - and power users often say variants of “I would pay $10,000 a month for this website.”

My takeaway from that SaaS business was that in our next business, we should go one better. If we were so smart, if we could generate a 5% revenue uplift, then we should do that to our own revenue, in our own company. The learning was, make sure that you capture the value you create!

This is the same logic that leads successful hedge funds (yes, the irony is not lost on me that we were not a successful hedge fund) to stop taking outside money. If they’re charging 2&20 to invest other people’s money, and reliably generate a 30% IRR, then their cost of capital is actually really high; they would be earning 8% on their AUM each year, so they’d be better off just taking out some credit card loans at a 20% APR and investing that to earn a 10% net interest margin. Of course, the big emphasis here is on the ‘reliable’; the 2&20 model gives you downside protection.

What I didn’t fully realise at the time, though, is that there’s often a tradeoff here. SaaS businesses get very high gross margins, but don’t capture all the value they create. ‘Real’ businesses have much lower gross margins, but often do capture more of the value they create. So what should you do - SaaS, or real businesses?

Search ads

Boris Johnson once said that his policy on cake was pro having it and pro eating it; some businesses get to do this. 

Normally, companies compete to capture less of the value they create. Imagine two service providers can both create $100 of revenue for a customer; it costs each them $1 to provide that service; and the customer can only buy the service from one provider. In theory, the two service providers will enter a price war, bidding each other down to a price of $1.01. Effective sales, marketing, and product all help avoid that price war, and hence let service providers charge more.

But in some business, customers bid to let their suppliers capture more of the value they create. A great example is search ads. Skyscanner and Expedia both want to bid on the keyword “cheap flights” - they might each click from a user searching for cheap flights is worth $1. Google (because they aggregate demand) has a lot of people searching for “cheap flights”, and so Skyscanner and Expedia will enter a price war with each other to win that search term, bidding right up to $0.99 - or, if they’re trying to blitzscale, even higher than that. This means that the Skyscanner and Expedia are fighting with each other to give Google more money for the same service. 

In SaaS, suppliers bid to be the one supplier each customer uses. In search ads, customers bid to be the one customer each supplier serves. And hence, search ads models are both high-margin and capture most of the value they create. They’re awesome businesses.

SaaS vs. software-enabled-services

If you’re smart enough to increase your customers’ revenue, and you’re only capturing 2% of the value you create, maybe you should just do their business?

One reason that’s not a good idea is that you might be very good at one aspect of your customer’s business, but not at others; it’s the Adam Smith idea that specialisation is good. The thing is, if you can maximise one key advantage, then you might be able to pay other people - in the form of service providers or actual employees - to suffice elsewhere. So the viability of a software-enabled-services business turns on the magnitude of your advantage (“you need to be 10x better”), and the difficulty of sufficing on everything else. 

A second reason to stick with SaaS is that generating a few basis points of uplift matters more at scale, and as a startup, the easiest way to acquire this leverage is to generate uplift for someone else. I think this is a crucial way to think about the choice between selling SaaS and doing the underlying business yourself. If, given your advantage, you can quickly scale, then you should be in the underlying business.

Examples:

The company where I used to work, Tripledot, achieved stunning revenue growth; it was the fastest growing company in Europe two years in a row. That’s because if you divide 2022 and 2023 revenues (very big) by 2017 and 2018 revenues (very small) you get a big number. Arithmetic, eh! Tripledot’s approach was to build simple games, and do a lot of things a bit better than the competition. If you do all this well, your Solitaire game might generate 5% more revenue per user. Crucially, though, mobile game companies scale using ad networks; and if you can generate more revenue from each game install, then you can bid higher than your competitors. The market is really elastic; if you’re 1% better than your competition, you win every time, and so you scale incredibly rapidly. The mobile game market is particularly elastic, and hence it’s a particularly good space to just back yourself and build the core product. 

That means building Tripledot is miles better than building a SaaS business selling to companies like Tripledot (which was the kind of company we built). But the kind of company you really want to build is a search-ads style business - i.e. an ad network. Fortunately, Applovin has done exactly that. I think Adam Foroughi is one of the most underrated CEOs in the world. To learn more about that business, read here or here or here or here. But the main point is Applovin owns both an ad network (their “software platform”) and a suite of mobile apps. Applovin acquired the studios that make those apps in order to generate first-party data to train their ad network; it’s not quite a loss-leader or a complement, more a way of learning to cater to their customer’s problems by bringing a subset of those customers in-house. And that was a good idea, because the software platform - which works according to search-ads economics - is a way better business than the apps. 

Another way to scale aggressively is if you’re in a space that’s ripe for a rollup. This is basically the approach that Metropolis are following: most parking garages aren’t very efficient, and so Metropolis hope that they can expand margins by adding tech. There are a lot of parking garages, and the market is fairly fragmented, which means that Metropolis can execute a roll-up by paying an above-market EBITDA multiple for each parking garage justified by their margin-expansion capabilities.

In short, you need a really good plan for either organic or inorganic growth.

I’ll add a post-script here soon…