What We'd Do Differently If We Started Instaustad Today

Building a project always involves a certain amount of learning on the job. If we were to start Instaustad from scratch today, there are several things we would do differently.

Hindsight is a useful tool, not for pretending we had everything figured out from the beginning, but for showing that we are constantly evaluating our work and striving to improve.

What we got wrong initially

First, we would write our editorial policy from day one. As we've mentioned before, we only formalized it after catching ourselves almost publishing unverified advice. Setting those ground rules early would have saved us a lot of internal debate.

Second, we'd focus on clearer error messages from the start. The very first versions of our tools had error messages that were far too technical, lacking the plain-language explanations that our users actually needed. We also would have designed the blog architecture before publishing the first post, rather than having to retrofit it later, and we would have integrated screen reader testing much earlier in our development process.

What we'd keep the same

Despite these lessons, there are core decisions we wouldn't change. We would still refuse to require logins. We would still refuse to ask for private account access. We would still build a platform without user tracking.

These principles define what Instaustad is. The things we'd change are about execution and process; the things we'd keep are about identity. We share this because transparency isn't just about showing our code — it's about sharing how we think, where we've stumbled, and how we plan to do better.