A Year of Shipping at Adalo
Deep linking, X-Ray, and Studio looked like three unrelated features. Looking back, they were the same idea told three times.
If you'd asked me a year ago, I'd have told you we shipped three unrelated things. Deep linking in the spring. X-Ray in the summer. Studio a few weeks ago. Different teams, different problems, different quarters.
Looking back, they're the same idea told three times: a no-code tool earns trust by being honest about the real thing underneath it.
Deep linking — the app is not an island
We started by connecting apps to the world outside themselves. Deep linking was our third most-requested feature, and it's the least glamorous thing we shipped all year — a link that opens the exact screen instead of dumping you on the home page.
Nobody demos it. Everybody needs it. It was the first admission of the year: the apps people build on us are real, and real apps have to connect to sharing, notifications, and everything else that lives beyond their own four walls.
X-Ray — the cost is not hidden, just invisible
Then we turned the honesty inward. X-Ray scans an app and surfaces what the builder couldn't see — the nested list quietly multiplying database queries, the screen carrying too much, the orphaned screens inflating the bundle.
None of that was hidden on purpose. It was hidden because "you don't have to think about it" slowly curdled into "you can never see it." X-Ray was us drawing the line between those two, and deciding a good tool hides the mechanism but exposes the consequences.
Every one of these features was really about the same move: stop pretending the hard part isn't there, and instead give people a clear view of it.
Studio — the AI is not a black box
And then Studio. The temptation with AI was to make it a magic text box: describe an app, receive an app, don't ask how. We built the opposite. Ada generates a real, multi-screen app onto a visual canvas, and then hands you the controls — you direct it by pointing at what's on the screen.
Same principle, one more time. The AI does the cold start, but it doesn't get to be a mystery you argue with. You can see the thing, touch the thing, and change the thing directly.
The thread
Three features, one belief:
- Deep linking — treat the app as a real thing that has to live in the real world.
- X-Ray — treat the builder as someone who deserves to see what their choices cost.
- Studio — treat AI as a collaborator you can direct, not an oracle you obey.
No-code gets caricatured as hiding reality from people. The best version of it does the reverse. It removes the busywork and hands you a clearer view of the thing you're actually building — the app, its cost, the machine helping you make it.
That's the year I'm proudest of. Not three features. One argument, made three times, in code.