Ok I get what you’re saying now. To steelman your argument a bit: first, all the credit card interfaces are standardised. I thought you were talking about the payment transfer interfaces, but I’d forgotten that PayPal’s initial product ran on credit card rails. So yes, you’re absolutely right. Second, you’re saying that PayPal isn’t an example of scale advantages so much as an example of a company that is better adapted to a digital shift. Certainly, the credit card companies and the banks could have run a payment transfer service off their existing infrastructure (since PayPal also ran on it), and would have higher scale from the get-go. The key reason they didn’t is that they were not set up to take advantage of the web, since the internet was this new-fangled thing. Finally, if scale were such a huge advantage, we wouldn’t be seeing other players in the market who have overcome the fraud problem today.
All of which are good points.
In response, I have a couple of notes. One thing that we’ve discussed in 7 Powers in Practice and tangentially in Take A Simple Idea and Take It Seriously is that most moats don’t last forever. I think the bar that Helmer uses is around a decade (i.e. if a company resists margin compression for about a decade, that’s probably an existence proof for a moat and considered really, really good) — it was Nick Sleep and Qais Zakaria who asked what moats were the best and/or lasted the longest. And PayPal has certainly had the market to itself for at least that long (they had marvellous pricing power, as Thiel puts it).
So the existence of competitors today — more than 10 years after PayPal’s initial period of competitive advantage + pricing power — is not necessarily evidence that they had no scale advantage.
Which leaves us with the question: why did no competitors beat them in the initial 10 year period? It could be that PayPal’s fraud technology is a cornered resource. But I highly doubt that — technological advantages are rarely if ever a cornered resource, unless protected by patent or government fiat. Besides, members of the initial team took the tech and turned it into Palantir a few years later. Perhaps a more interesting counter-factual is to ask — if PayPal were not the scale leader, could they have beaten the fraud game and won?
I find that difficult to imagine. A lot of PayPal’s anti-fraud tech was built off observing a large number of patterns over hundreds of related accounts. If they were not the scale leader in a highly crowded market, they might not have learnt fast enough to survive. I think there’s a 1-in-10 chance that they could have won had they not have been the scale leader. (To reiterate: most if not all of their competitors from the time did die … though it is difficult to say if they died due to fraud. It’s more likely that fraud was one of the factors, but not the only one — the terrible funding climate in the wake of the dotcom bust probably didn’t help.)
Which leads me to address your second point — yes, traditional players could have beaten PayPal with their existing scale, but they didn’t. Power is relative — against its dotcom competitors, PayPal had a scale advantage. Against traditional players it did not, but it didn’t really have to compete with them since they were positioned differently / less well-adapted to the digital transition, as you put it.
Returning to the general rule of thumb: the notion of about ~10 years for most moats makes intuitive sense — very, very few moats are able to resist competitive arbitrage over the long term, since the market changes, regulation changes, the technology changes, and consumer behaviour changes. One way of looking at the emergence of new competitors is that they occurred due to changes in the environment, and I think that’s broadly true. Certainly, Transferwise, Stripe and Adyen all grew into the payments space by riding different waves. To take Stripe as an example, it started with a credit card payments gateway product, focused on digital merchants, and grew off the back of the burgeoning SaaS+Startup market over the course of the 2010s. That change in market environment meant that it could hit significant scale through this new demand, in order to bootstrap its own fraud detection algorithms. And if I recall correctly — it took them a number of years before they could release Radar.
(Again, correct me if I’m wrong — I have not thought as deeply about the space as an insider might, and am happy to hear perspectives from someone with direct experiences with these companies).
The other thing that is surprising to me is that PayPal is still a player. I was talking to a friend in Stripe about Adyen the other day, ostensibly their direct competitor. To my surprise, he told me that some within the company thought PayPal was the bigger threat — since they had long ago moved away from pure credit card transfers to running their own wallets (meaning that transfers within the PayPal network and to banks had much lower transaction fees). I had assumed that PayPal was no longer a major player, but he disabused me of the notion. Who knew?
I hope this address your questions?
Edited to add: I think the 10 year rule is a rough rule of thumb, but I can’t find a written source for this — I think Helmer may have said something to that effect in a podcast interview. It’s definitely not a rigorous number.