The poor unit economics of Zara and Adidas

Unit economics are the lifeblood of technology companies . If you don’t get more cash from each customer than it costs to serve that customer (or an equivalent “unit”) the business hardly makes economic sense.

Before the rise of subscription business the profitability analysis was rather simple: If you were a furniture artisan you made money on a chair when it sold for more than it cost to manufacture. Your gross margin told you exactly how much was left for overheads after paying the direct costs. Selling a chair for $100 that cost $60 to make meant you ended up with $40 left for salaries, rent and whatnot. Life was simple.

Rise of subscription businesses made this equation more difficult as now that $100 is actually divided among payments split over multiple months. Since then, many “rules of thumbs” have emerged such as the need to have >3x LTV / CAC in order to have a successful business.

Rules of thumbs are of course great for simplifying the world, but taking them for granted risks becoming too blind to see the forest from the trees.

Take an example: Some of the best companies in the world, like H&M, Starbucks, NYT or McDonalds enjoy a meager 20-50% gross margins. Yet such margins are absolutely frowned upon for a software company where the standard is 70%+. “Terrible unit economics”, they say. Why exactly is 3x LTV / CAC great but the moment you go down to 2x LTV / CAC things get difficult?

Gross margins of 20-50% are hardly acceptable within software

But here’s the thing: what if I told you that 2x LTV / CAC could be equivalent to ~40-50% gross margin i.e. the same margins that Adidas, Mcdonalds and H&M enjoy?

Gross margin on a product should include the direct costs related to serving that product and there is hardly a more direct cost than the cost of acquiring the customer. More so, there is nothing preventing you from being creative and taking that CAC from the LTV / CAC denominator and pushing it to the nominator to get a “fully-loaded gross margin” thus defining Gross Profit as Revenue less COGS less CAC.

Let me exemplify below what I mean with 2x LTV / CAC = 40% gross margins. A company that is selling a $33 / month subscription to a customer that churns in 3 months will make the same amount of cash as a company that sells $100 perpetual license. This is simple arithmetics and a 101 of subscription business model - so far nothing new.

Dividing a lump sum into many smaller payments creates no extra value by itself

If the gross margins are the same, customer acquisition costs the same, it makes no difference whatsoever whether you sell the product as a perpetual license or as a subscription model. In fact, if nobody told you this was a software company this could actually be H&M in disguise. At short enough lifetimes, when time value of money is negligible, the difference between subscription models and one-off / license sales is nonexistent and it might even be conceptually easier to think of the software company as a one doing one-off sales.

Needs volume to work

There is one common denominator between H&M, Starbucks, Zara and others and it’s that they are what the industry calls fast moving consumer goods (“FMCG”) companies that largely rely on volume. In order to be viable businesses with their gross margins Starbucks is estimated to sell almost 3Bn cups of coffee and Zara to produce over 450 million products per year .

Indeed, in order for this 2x LTV / CAC or 40-50% gross margin software company to also be viable it needs volume - a lot of volume. Volume either through (a) acquiring a lot of customers or reactivating old customers periodically or (b) shoving out new products consistently. Most of the time it is the former, while the latter is more difficult in practice but used by e.g. many hypercasual games.

Of course, acquiring new customers is also hard and for an enterprise software with 6 month lifecycles, $100K acquisition costs and 2.5m potential customer universe in the entire world the economics become difficult.

But for consumer companies with potentially >6bn of customers and low customer acquisition costs this is more doable - let’s take a look at one example next.

Tinder does it right

To give an example of this let’s look at Tinder. In a way it should be a great example of a subscription app with a low lifetime built into the design (people find a match!). Yet it should thrive, because it addresses an extremely large global market that creates new customer volume and (perhaps unfortunately) sees a high number of reactivations for churned customers as some of the matches weren’t meant to be. Bottom line: it sees a lot of customer volume.

Using a combination of information from Tinder’s Q4-21 report , and data from this excellent Lazard Consumer Subscription Software report I’ve built a simple model that shows both the subscription unit economics, and the pro-forma perpetual license / fully-loaded gross margin economics.

Tinder effectively has 45% gross margins once its subscription with 6 month lifetime is converted into a lump sum sale

We see that using the rough inputs above we get ~2.7x LTV / CAC which is below the magical 3x threshold that makes most practitioners shiver. Yet by “converting” that subscription revenue to perpetual license revenue it’s actually a rather decent 45% fully-loaded gross profit margin vastly above Starbucks’s 29% and not too far off from McDonald’s 50%. Furthermore, all this results in 23% net income which is quite close to Match Group’s actual 29% net margins, giving me further comfort that the numbers above are probably not too much off.

So there we have Match with seemingly “poor” unit economics but regardless a thriving company: look at the below growth and profitability year-over-year (sidenote: all analysis so far also excludes Match’s other product portfolio such as Hinge, OKCupid, OurTime…).

Despite <3x LTV / CAC Match Group (Tinder) is growing >20% YoY with almost 40% operating margins

Sales cycle makes it difficult in B2B

As mentioned earlier, in order for low unit economics to “work” you need high volume of sales which is more typical to B2C than B2B.

While you could make it work with B2B as well by getting an army of sales people to hit the phones, it gets vastly more expensive due to longer sales cycles. When you get a lead it can take months until that lead goes through the whole sales funnel to convert into a hopefully thick contract.

All that time enticing the customer the B2B company is not generating any revenue but incurring a lot of sales-related expenses. Thus a B2B company will always end up with lower margins at the same LTV / CAC simply because the timing of expenses and revenues are further away unlike with B2C where CAC and all revenue is generated effectively simultaneously. Indeed, because each contract takes time the unit economics in B2B are ever more important as exemplified by this citation from David Vandergrifte :

... In our professional opinions, LTV and CAC are all the more important to get right for enterprise startups precisely because the number of contracts is so small…

See below as a visual example of acquiring one customer. The B2C one roughly follows the Tinder example, while the B2B one assumes a 6 month sales cycle and $60K CAC that results in $120K 12 month contract (i.e. 2x LTV / CAC) after which the customer churns.

Sales cycle is a key reason why B2B companies burn more despite having the same LTV CAC

The difference in margins is palpable. The B2C one is able to roll with ~40% margins while the B2B is rocking at 10% for the calendar year.

The usual argument against the different margins includes arguments about B2B needing an expensive enterprise sales force. While that is true, it should be netted off by vastly larger contracts. Yet in practice it doesn’t and the reason tends to come down to the sales cycle.

To be clear, this does not mean B2B software is a bad investment. Because of how the math works it will eventually reach the same margins, but just needs more time to reach that point due to the built-in sales cycle. Plus the quality of the revenue is often higher given B2B is easier to upsell and more recurring in nature as B2B is tied to budgets and dependent on the opinions of many people in the organisation. More so, on a free cash flow basis B2B will likely generate higher margins given large amounts prepaid contracts.

Finally, see below some illustrations from B2C and B2B margin differences from beforementioned Lazard’s consumer subscription software benchmarking with Jamin Ball’s fantastic B2B software benchmarks layered on top. Only a handful of B2C companies have negative profit margins whereas effectively all B2B companies do. Although this is only a subset of B2B companies, you'd see the results hold when looking at the complete B2B software list .

The margin difference is palpable between B2C and B2B

That's all this time folks

By Joel-Oskar Raisanen

Sep 20th 2024