Drinking from the Firehose #141: 💵 Substitute goods. 💵


Fred Wilson of USV wrote a blog post that's seen a LOT of circulation over the weekend. It got me thinking about the relationship between margins and valuation, as well as the implications for consumer companies.

In the post, Fred draws a distinction between how late stage private markets look at a new set of lower margin, technology-mediated services businesses (Uber, WeWork, Peloton, etc) and "classic" high margin software businesses (Cloudflare, Zoom, Datadog, etc). As a group, the latter have higher revenue multiples than the former. The implication is that the two quantities are correlated. And, ceteris paribus, they are.

Gross margins are an input to the eventual EBITDA margin of the business. EBITDA drives the bulk of free cash flow, which ultimately drives the discounted cash flow analysis that determines what the stock price (and, therefore, the multiple) should be.

The reason we pay so much attention to gross margin at the time of IPO is that very few tech companies go public while profitable. Even those who do are far away from achieving their long-term EBITDA margins. That means that public market investors need to understand how an unprofitable, fast growing business generates operating leverage and eventually produces reliable cash flows. The higher the gross margin, the faster it should hit profitability, since gross profit is presumably growing faster than operating expenses.

Once a company arrives at its long-term EBITDA margin, it settles into a long-term growth rate. That rate is also a significant determinate of stock price, and again multiple. The ability to sustain that growth while still producing profits is where the magic happens.

That's why focusing on gross margin in a vacuum can be a bit misleading. Plenty of companies (software or not) exist with lower gross margin profiles that, nonetheless, generate meaningful EBITDA and still grow at double digit rates:

  • Apple* ($1T): 38% GM, 31% EBITDA, 16% growth

  • Netflix* ($124B): 37% GM, 11% EBITDA, 35% growth

  • Amazon* ($884B): 40% GM, 12% EBITDA, 31% growth

  • Etsy ($6B): 69% GM, 18% EBITDA, 41% growth

  • Grubhub* ($5B): 42% GM, 10% EBITDA, 50% growth

  • Chewy ($10B): 20% GM, -5% EBITDA (but narrowing), 69% growth

  • Canada Goose ($5B): 62% GM, 26% EBITDA, 41% growth

  • Lululemon ($25B): 55% GM, 25% EBITDA, 24% growth

  • Nike ($148B): 45% GM, 14% EBITDA, 8% growth

It's no coincidence that every name on this list is a consumer company. Consumers purchase physical goods and services, often mediated by technology. Most consumer goods are priced elastically because nearly every one has a substitute. More substitutes generally imply lower gross margins.

In comparison, enterprise software businesses tend to have fewer substitutes at scale due to high customer switching costs and/or the lock-in effects of developer platforms. How many CRMs are there outside Salesforce? Productivity software outside Microsoft? In addition, a new class of consumer-like enterprise companies (e.g. Zoom, Slack, Dropbox) have user networks that generate lock-in. Gaining critical mass before substitutes come along allows them to secure high gross margins.

Consumer companies need to develop moats to sustain margins in other ways: layering user networks of their own, building platforms with multiple sides, decreasing unit costs with scale, and/or building an ecosystem around a brand. All of these protect against the dilutive effects of substitute goods.

The companies that have seen the toughest IPO performance are consumer companies who have not insulated themselves against substitutes. Uber and Lyft are locked in a death match over driver incentives in the U.S., driving margins down to difficult levels and requiring billions to subsidize losses. Governance issues aside, WeWork competes with every commercial landlord on the planet for a fundamentally similar product.

To the contrary, I am bullish on both Peloton and Spotify* because I believe fewer substitutes exist for these services, especially at their scale. Relatively few users are unsubscribing from these services. Both have artificially low margins due to the unattractive structure of the music licensing business. Both have paths to grow margins in the future.

I'm glad the tech investment community is discussing margins now, especially for consumer companies. More attention should be paid to economic moats and the ability to drive operating leverage at scale. The best way I've internalized these is by thinking of substitute goods. If none exist and the market is large enough, I tend to believe companies will eventually find their way to healthy long-term profit margins and sustainable growth.



A digital shoo-in.

Nike declared its intention to own more of its customer base directly a few years ago. Its path to victory involves the same digital disruption that we see from brands that are born on the internet.

This article lists a few pieces of evidence of Nike's evolution to a digital brand. First, its discount rate has declined to a 3-year low, a function of owning its own distribution. Second, it has invested heavily in its Run Club app, which has nearly 1 million ratings on iOS thanks to its tight integration with Apple (Tim Cook is on Nike's board). Third, Nike is staffing up its owned full-time retail operations at a rapid clip before the holidays, while part-time hiring remains flat.


Your TV watches you (tweetstorm).

A Princeton professor highlighted multiple academic papers that detail how popular cloud TV apps like Roku track what you're watching. Research shows that many channels have multiple trackers which can gain access to your MAC address (a unique ID for your device) and often your email address.

The rationale behind this tracking is that it is occasionally useful. For instance, "automatic content recognition" services can detect what you're watching in real-time in order to display more info on that show/episode to the user. The price you pay is that your TV is watching you, while you're watching it.


Amazon cares.

Amazon, which has been dipping its toe into the healthcare market, announced its Amazon Care virtual clinic for internal use this week.

With nearly 675,000 employees, Amazon is tiny compared to the country's biggest health plans (50m covered lives at United, 40m at Anthem, 22m at Aetna), but its population is large enough to get real signal out of health initiatives before rolling them out to external customers. I expect Amazon to use this telemedicine service to create a new channel for the prescription medicine business it entered via its Pillpack acquisition.


Forever blowing bubbles.

Astronomers have discovered mysterious "filaments" clustered around the supermassive black hole at the center of our Milky Way galaxy. Even stranger, these filaments are encapsulated inside a pair of giant "bubbles" that measure 1,400 light-years top-to-bottom.


Baby on board.

Japan Airlines quietly introduced a controversial new feature to its seating chart. During the booking process, you can now see where children under 2 are seated. Only a matter of time before this feature becomes an "upgrade" and costs more!