There goes another lap around the sun

👋 Hey — Egemen here.

This year really hammered something home for me. Building was never the hard part. Most of us can build (honestly we always could). Now with AI, that gap is basically gone. You can ship something decent very quickly, and almost everyone can do it.

Twelve months ago, GPT-4o felt like the ceiling to me.

Back then it honestly felt like we were close to “good enough” and the conversation would finally move away from models and toward products. That obviously did not happen. Another year passed and the pace only picked up. Everything got faster, cheaper, and easier to use.

The problem shows up right after that. I think once building stops being painful, we stop being careful.

I see teams fall into this all the time, myself included. You start with a clear idea, then you realize how easy it is to add things. One more feature here. Another use case there. It all feels logical in the moment. Nobody is doing anything “wrong.” But slowly the product gets heavier and harder to explain, and nobody can point to the exact moment it went off track.

That is called featuritis (a fancier term of scope creep).

At first it feels like progress but then there is a point where adding more actually makes the product worse. More screens, more decisions, more things to learn. Users feel it immediately, even if the team does not.

When capability is unlimited, restraint becomes the real skill. Users do not care how much your system can do. They care how quickly they can get value without thinking too hard.

The KISS principle keeps showing up for a reason. Keep it simple, stupid. It sounds blunt, but it is incredibly practical. Simple products are easier to adopt. Simple messages are easier to sell. Simple decisions compound instead of slowing you down.

Speaking of kisses, here’s my favorite KISS song:

I have seen this play out firsthand (not the plot of the song).

The moments where things actually started working were almost always after we removed something. We cut features, tightened flows, and stopped trying to serve everyone. The product felt clearer. The team felt more confident. Progress picked up again.

This is also why I am not convinced founders need more advice.

We are surrounded by advice already. What we usually need is a mirror. Something that forces us to look at what we are building and ask whether it is actually helping or just adding noise.

Looking ahead, I think 2026 is going to be wild for AI and unforgiving for founders.

I guess it’ll come from doing less, on purpose. From making clear calls. From keeping things simple even when it feels uncomfortable. That is the part I think really matters now.

Happy new year y’all!

Do you like SCALABLE?