A couple of weeks ago I dropped into a thread by Mosh Lee, about how ATproto has too many projects and too few businesses.
Yes! A few months ago I volunteered to help teams with ops or strategy, and kept this pinned to my profile until recently. A few founders took me up on it.
mu.social/profile/rica…
A recurrent issue was precisely what @mosh.bsky.social mentions: is this a project or a company?
— Ricardo J. Méndez (@ricardo.bsky.social) July 30, 2026 at 10:04 AM
[image or embed]
I hadn’t written about this mental framework before because, truth be told, it seemed sort of obvious. However, posting about it got me thinking that I hadn’t articulated it clearly before. Furthermore, it had one glaring issue: splitting it between project and company wasn’t specific enough, and conflated a few things.
We actually have two axes to consider:
- Project vs. Product (“who is this for?”); and
- Startup vs. Business (“where does your money come from?”)
Project vs. Product
As I mentioned above, Project vs. Product is easy to separate by asking yourself: Who is this for? Who is this about?
The more the answer is “for me”, the closer you are to the project end of the axis.
Products are reusable, they are easy to pick up by others, and they make trade-offs that you wouldn’t have done if they were just for yourself because they come from your user’s requests (or what you expect your user persona to need).
Perhaps more tellingly, if you are an engineer, products involve boring stuff that you would more likely not do if they were a personal project: documentation, some user hand-holding, smooth concept introductions, UX polish, clear error messages, instrumentation, dashboards that help you keep up not only with the project’s internals, but with the usage that users are actually giving it out there in the wild.
Products have a chance at surprising you, because users are unlikely to behave the way you do. Products need to adapt to others.
Startup vs. Business
The difference between a startup and a business is one of commercial focus, so the closer you are to the project side of things, the less likely either is to apply to what you are building.
Mind you: a startup is not just a business that’s young, small, or still burning cash — plenty of established businesses manage to set money on fire.
What sets them apart is the question of who you foresee up-front, as you start, will be giving you money for the next few years, and why.
The best definition I’ve seen is the one from Steve Blank, who launched the lean startup movement. He defined a startup as a temporary organization formed to search for a repeatable and scaleable business model.
If you start a new, small kind of burger joint, you are not a startup.
Sure, you might be doing something innovative, something nobody else is doing in the market (high-tech burger delivery, just-in-time supply ordering, lab-grown meat, vertically-integrated supply chains… whatever it is). But the burger joint business model is a known one. You know who is supposed to give you money (hungry customers) and why (hamburgers).
All the other innovations are things that may affect your cost structure or profit margin, but the business model itself is a known quantity.
Bluesky, for instance, is currently very much a startup. This is not about ATproto, portable identity, being decentralizable, or how they are approaching the development community — all of these are the equivalent of the burger joint innovation I mentioned above.
The key reason is that Bluesky does not have a business model right now.
So the money answer changes. You can’t be sure if users will give you money or why, because Mundus Sine Caesaribus apparel is unlikely to be repeatable or scaleable, so it comes from investors, with an expectation of outsized returns.
Counterpoint: if Bluesky was from the start planning to sell ads, then it would just be a business running on novel rails, because the ad-driven social media business model is a known quantity.
Some examples
- Bluesky: product startup, as discussed above.
- Graze.social: product startup — it’s built for others to use, but was in search of a business model when it got Sherlocked.
- HappyView: started as a project, and while the documentation seems to be improving, I wouldn’t call it a product yet.
- Marque: product business. Notice that your product doesn’t need to be something that you distribute beyond running your own website, but it does need to consider how an average user would interact with it.
Why are we talking about this?
Because you need to pick a path, and the path you choose narrows down who you talk to and the stories you get to tell — and, as Mosh mentioned, investors are interested in the space, but the space doesn’t seem to be ready for them.
Here's the deal: Most ATProto projects are projects. Investors need businesses — or at least teams who can show they're gonna be able to build a business.
— Mosh Lee (@mosh.bsky.social) July 29, 2026 at 10:11 PM
Investors won’t fund a project. There is a certain hassle and overhead that comes from polishing a product, and you need to know if you want to pay that price. If you are unsure, you shouldn’t waste your time or the investors’.
It’s fine if you don’t: plenty of things that started as personal projects ended up changing the world (e.g. the Linux kernel or the World Wide Web). But that means if you want to do it long term, you are committing to investing your own time and focus on it, in exchange for the uncertain potential of grants or donations.
And then, if you do decide you want investor money, you need to decide between startup or business, because they are different types of investor, with different concerns and horizons (and they will expect different validation points from you).
Fine, so what do I do now?
You need to make a call! Not everything needs to be a product, a business, or even a startup. Not everything needs to scale massively, but you do want it to scale at least to the number of people you want to serve. Not to mention, you may find that a minimum addressable audience is what makes the business work in the first place.
Personally, I’d argue that unless you find yourself in a crazy bull market, or have some unique angle for pulling in investors, chances are that you do not want to be a startup: businesses are a lot easier to reason about, validate, and pivot. You are also a lot more likely to spend your time where it matters — thinking about your users — instead of schmoozing investors and trying to hit the magic growth numbers you’ll need for your next life-saving funding round.
Meanwhile, my offer above still stands: if you need help driving your business, ops, or strategy, hit me up!
Putting my time where my #atmosphereconf ecosystem mouth is: if any early-stage ATProto team needs help honing their ops or strategy, ping me.
If we are aligned, we can find a way to make it work.
— Ricardo J. Méndez (@ricardo.bsky.social) March 30, 2026 at 4:14 PM
[image or embed]
Further reading: if you’ve gotten this far, you may be interested in my complementary piece on ATmosphere applications, separating them by survival mode.