When we launched the Rail 52 and then the 34, there was a ton of skepticism that we'd actually "made" our own rim. I say "made" because we completely outsourced production, but for anyone who might still be skeptical, yes it was 100% our design and sold only through us. The unique mold mark on the original non-tubeless ready rims (remember, they launched in 2013 when road tubeless was barely a thing at all) was meant to prove that, but even that didn't convince anyone.
Being in the business of proprietary stuff is a two-edged sword. Starting with the bad part, your logistics are a pain in the butt. You'll pay ahead of time for production. There's no buffer stock, generally not even any warranty stock (you have to pay for and buy that if you want it), things never in our experience showed up earlier than schedule, and you're just generally very exposed to any issues. Your cash flow is impacted because you pay now for stuff you'll start selling in maybe 60 days, but generally you'll be cash neutral on it in 75 days. That's a lot of forward sales to pay for.
(no more skiing this year)
One of the good things is that you get to capture all the rewards of any promotional effort you put into it. The trips we took to the wind tunnel were worthwhile because there's no "oh that makes that rim attractive to me, I'll go buy it from someone else now" dynamic. On the other hand, no one's sharing the load on promotion - you want to get the word out, you have to get the word out.
When the Rail 52 launched, it was uniquely wide, at least internally but also externally to a lesser extent. By the time we made the second mold for 2016, tubeless ready had become market mandatory, and the 18mm inner width was starting to seem a little narrow. HED had launched the Belgium+, which was the first decisive step beyond 18 for road wheels, and there was pressure to go wider, but we thought that 18 was still appropriate then. And in hindsight I still think that it was the move at the time. Now, 18mm rims are WAY out of vogue. The market moves quickly, and the market has moved quicker lately.
Molds are expensive. Retiring one for design obsolescence rather than lifespan death is painful, and that's a super real risk. Whether you are talking about hubs or rims, alloy or carbon, tooling cost is a bear.
But the one thing that proprietary has continually been able to do is support strong margins. It's a near certainty that we didn't exploit that enough, but that's really not how we do stuff (which isn't to value judge that or act superior, it's just what it is). There is perceived value in proprietary, and people are willing to pay for it.
An important corollary to that is that there's no equivalence. Two shops aren't selling the same thing and easily compared, because of proprietary.
So put that all into a blender and hit go, and what we got was that any rim we might be tempted to make (that would justify itself in audience size, etc) is already available from a very good producer - perhaps/likely a better producer than we'd be able to get to produce for us. So we made the choice to not go proprietary on rims, and it's worked out fairly well for us. Better than the alternative? Impossible to say, but it's working.
Same story with hubs? Stay tuned for the next episode.