It is common to speak about hockey stick growth when a company takes off.
I realized recently that there are in fact two pivotal points that trigger this growth. One is commonly referred to as the product-market fit. For the other one I have not seen it identified clearly, and suggest we call it the product-institution fit.
At both points the product establishes itself as a fit for some market, or some ecosystem, so solidly that from there onwards the company is destined to succeed.
The audience that drives the growth in the product-market fit phase is We, The People. It's your average Joe who likes the product, wants to use more of it, and is happy to pay for it. At the product-institution fit phase the audience and the market makers are the corporate / enterprise / governmental entities.
Even though in relative figures the hockey stick is not as noticeable during the product-institution fit stage, the amount of money the company brings to its shareholders "per minute" is dramatically larger at this stage.
Thus, once there, the strategy becomes obvious: tailor to the future huge customers, be those the corporations or the governments. Because this is where the gold is now. This transition does not necessarily mean the product is going to get worse. What it does mean though is that the priorities do change.
For a canonical example, consider Microsoft. Among the end users Windows is infamous for many and one difficulty to be used effectively. At the same time, system administrators respect Windows quite a bit, thanks to its predictable configurability, especially at scale. I believe it would not be an understatement to say that Microsoft cares more about pleasing system administrators with their product compared to its end users. Couple this with their international sales force, and it’s no wonder that various seemingly niche products of Microsoft, such as the MS SQL Server, are $1B+ businesses by themselves. Despite having plenty of open — and free — alternatives!
Apple, on the other hand, sells to consumers. And it shows.
And Google is a behemoth counter-example. Google is notoriously famous for building awesome tech, notoriously famous for building amazing products for nerds, and, at the same time, infamous for being unable to solve large-scale operational problems. Ask me one day about how I tried to pay to increase my storage on a “corporate” account of a corporation of two people that I manage. That part of Google’s offering literally feels like it’s hacked by a bunch of schoolkids overnight. Well, to me this only means that Google can still score big once it finds a way to work with institutions effectively; perhaps the sub-par growth of Google Cloud would make the company realize it has to strengthen its position in this regard.
I was thinking of more recent examples which I’m allowed to talk about publicly, and what comes to mind are Zoom and Slack. Both products have launched literally nothing valuable features-wise in the past ~two years. Being a user of both as the founder of my System Design Meetup, I’d gladly use — and pay for! — the versions of Zoom and Slack frozen as of two years ago.
From the user experience perspective, my feedback on both products is far from unique. Unironically, the market cap of both companies has grown substantially despite the end user experience not improving! One of them even got acquired by Salesforce, for quite a figure.
It would be a longer conversation about what exactly drove the growth of these companies. My bet would be that their valuation reached new heights thanks to one major factor: these companies proving they can adjust their growth strategy to successfully target institutions as customers. Because once an institution is onboarded as a customer, the half-life of the revenue stream from it is literally dozens of years. Not to mention that it is substantially easier to onboard more institutional players once the “zero to one” phase has succeeded.
And, at the same time, I can no longer disable the waiting room, and have to manually admit guests to my meetup events. Even the open off the record ones, where, clearly, I’d rather have people talk to each other, even if they joined some fifteen minutes before I’m even online. But that’s just not where the money is for Zoom.
There are several related points to be made.
One is about regulation. Don’t get me started on how the gig economy companies had to pivot. AirBnb and Uber have undoubtedly made this world a far better place, but it’s unimaginable how much better this world could have been if instead of flirting with the regulators they’d play their own game. With this much money at stake though, it would be rather foolish to expect these companies to not take the huge and safe opportunity of playing in.
(I wrote before how Uber as a corporation actually benefits from its drivers being employees, not contractors, as long as we don’t look at Uber as a separate for-profit entity, but as an organic part of an increasingly socialist Bay Area’s “ruling elite”, when viewed from this angle.)
Where the majority of the population supports various equal pay programs and generous parental leave for both parents, Uber wins a lot more in indirect benefits by playing into this narrative compared to what it could have won by remaining a lean for-profit business. In the latter case it may have even lost to another company that is more institution-friendly and institution-focused.
Another point is about founders. My understanding today is that the “builder founders”, who are great at the zero to one phase, are a different breed compared to the “institution-friendly founders”, who are great at keeping the institutions happy. These two skill sets are just too different, and only a few people in the world can be effective at both roles. For the former, think Larry and Sergey who built Google. For the latter, think Frank Sloothman, who is currently running Snowflake.
That’s probably why Jack has left Twitter and is rebranding Square to stay closer to the crypto world. Because, from the product-market fit perspective, Twitter is 100% done. And Square’s growth, unless steered heavily into some form of crypto-powered disruption, is clearly too little about staying lean and effective, and too much about learning how to work with the global financial institutions. Which is just not what Jack is interested in doing, or so it seems to me.
I may even stretch it further and say that A16Z is so successful specifically because Marc and Ben complement each other so well. At least that’s the impression I get from reading what they have to write, and how they convey their messages.
One more point, this time for software engineers and architects.
When a company enters its product-institution fit phase, forget all the things that nerds cherish. The product is no longer about clever algorithms, sound implementations, faster release cycles, effective resource utilization, etc.
The product is now almost exclusively about being able to quickly adjust to institutional demands. These demands are different from industry to industry, but their intersection would definitely include respecting and following regulations such as GDPR, shifting dashboards to be more friendly towards dedicated administrators and less understandable by your average Joe-s, and overall becoming closer to a yet another service offered via a single sign-on provider such as OKTA.
Because, sadly, making your average Joe’s life miserable when it comes to paying for your Google storage is a negligible loss compared to a huge win of being able to land ten more contracts with various three-letter organizations, or even with overseas governments and their subsidiaries. The latter may even be paid by US taxpayer money, BTW.
And a closing point, for investors.
This gets dangerously close to insider trading, but, seriously, the second-best point for a long-term investment, after the product-market fit, is the product-institution fit.
When a company has steady profits and is growing, forget P/E or the bottom line.
Looking at companies such as Asana or Datadog or GItLab or Amplitude or Roblox, I think the major predictor for their success was or would be how well they were and are capable of onboarding and working with the increasingly institutional “customer” base.
Because whether we, the pro-market people, like it or not, it is a huge gamble with bad odds to play against an institution long-term, no matter how failing or already failed you believe this particular institution to be. For one disruptor that does change the status quo there would be dozens that tried and lost badly. And founders who want to join the three commas club after hitting nine digits net worth are, statistically speaking, quite risk-averse when it comes to their babies.
So, when past the product-market fit stage, invest in and work with the founders who are laser-focused at pleasing those institutional players at all costs, while retaining the stem of their own product and their own company.
Simply put, if GitLab’s growth is based on the interests of the developers who want some new kind of one-click integration, its stock would float around $100 for a while.
At the same time, if it targets being friendlier than its competition when it comes to streamlining large-scale cloud deployments orchestrated by, say, organizations of 1’000+ developers all coding in Java within an IntelliJ IDE — it’s got a bright future ahead. Especially if this large-scale cloud deployment integration is blessed by one of the major clouds.
An obligatory closing disclaimer: I am not even remotely aware of any deals of this kind coming, but, if I do read about such a possibility tomorrow, my portfolio allocation would change quite a bit that very day.
This piece of yours truly was originally published by Alex Fleiss at Rebellion Research.