On September 29, I sent George Robson, Partner at Sequoia, my article on how AI companies will get acquired. His reply got me thinking because it’s a very different approach from what I’ve been seeing.
What I have been seeing is that distribution is becoming the moat. Rollups are buying for distribution. Legacy industries are being bought together because what is happening now is the “last-mile delivery of AI”.
And so I was a bit shook that they had decided against rollups and still stuck with the strategy that better product = more customers switching to you.
What George says is that he still believes the better product wins. In his words, “most people aren’t capable of building good products and then packaging it and selling it, even with infinite intelligence” and “better product = right to win.”
The question is where we draw the line.
If we play this forward and all companies eventually have the same automations, and there is no longer much alpha a company can pull from the product itself, what will actually keep a customer?
Thought Too Short
I think this is thought too short.
Even though the better product wins in the short term, at some point there will be another, different better product. That product might even be forced through regulation, the state, or some other external event.
When we extrapolate the trendlines, better product becomes diminishing alpha. It’s an edge, but not a permanent one.
I still think it’s a very rare event that an actual better product - where the better product itself is the event - makes companies switch.
George wrote back with Revolut, the banking app, where it’s currently very hard to switch. I think that’s temporary. In the future, I think people will have a very easy time switching from one bank to another because the bank basically has no system of record anymore.
What was more interesting to me was this idea of an event happening outside of the company’s force that makes customers switch: retirement, a takeover, succession, etc.
The most common one is succession. The owner runs the firm on QuickBooks for 20 years, then retires, and whoever takes over picks the new system.
So the switch happens at the handover.
And the handover is exactly the moment a rollup buys the firm.
The Meat
When I read “packaging and selling,” it reminded me of selling to banks at 16, because I saw over and over and over again that the person who is better at selling will have more customers, will have more feedback, and therefore will have a better product.
And that’s what I’m seeing with rollups as well, just in an extended version.
The rollups that buy together the distribution have more insight into where the market is going through the trends and predictions they can read out of it. And therefore, they will develop the better products, the better positioning, the better standing.
When everyone ends up with the same AI, my bet is that it will be way easier to build a good product. Still hard to build a very good product, but easier to build a good one.
George’s bet is that even though AI makes everything easier and building has become a commodity, it’s still going to be hard to build exceptionally good products.
I actually agree with part of that.
I think the floor is going to get higher. It will become easier to make a mediocre product very good, while it will still be hard to make an exceptionally good product. But the cliff between those two will become smaller.
And that means fewer people will switch.
The Seat
We are pricing different moments.
George is pricing the moment where the company enters the market and starts becoming number one.
I am pricing the moment where 99% of companies end up: being acquired because of their distribution rather than forever holding the edge.
Octrace, my first software company, taught me that by scraping public data, you don’t have an alpha in investments. Everything we could see, everyone could see.
The advantage belongs to whoever stands where the data is born.
The Bet
So my bet is that in the future, we will see more and more companies being acquired solely for their distribution.
Therefore, in a time where you can buy your distribution in the form of rolling up industries, I would buy it now.
There is one trap.
If you run the rollup only for margin -cut the staff, standardize everything, let the small clients churn - you destroy the record you bought it for.
So the first question I would ask any platform is:
What do you standardize after the acquisition?
