The building-in-public system that actually stuck for me (after a lot of failed attempts)
I've tried every "just be consistent" framework and failed all of them.
What finally worked was making the content a byproduct of the building, not a separate job:
Step 1. Decision log. Every time I make a real product decision, I write one sentence about why. Not a post. Just a log. Takes 30 seconds. This is the content backlog.
Step 2. One post per week from the log. I pick the decision that felt most uncertain at the time. Those are always the most honest posts, and honest posts perform better in this community than polished ones.
Step 3. Replies are reactions, not performances. I stopped trying to write the "perfect" reply. I just react genuinely to what someone said and add one thing they might not have considered. That's it. 30 seconds per reply.
The last piece was fixing the AI problem. I was using generic prompts and everything came out sounding the same. Switched to training on my own tweet archive — replies actually sound like me now. Built that into XreplyAI, which I run as a BYOK tool.
Milestone this week: paused Google Ads to clean up the keyword match pool, shipped a couple of account fixes, and got back to actually writing instead of just running ads.
What's the system that's stuck for you? Especially curious if anyone's found a way to make replies less of a time sink.
https://xreplyai.com?utm_source=indiehackers&utm_medium=social&utm_campaign=edusales-2026-05-26
I really like the idea of making the content a byproduct of actually building.
I'm trying to do something similar with my own project. Instead of planning content separately, I'm starting to document the problems I'm running into while building and growing the product.
I've already noticed that the things that teach me the most are usually the small problems I didn't expect—not the features I originally planned.
I'm going to try keeping a simple decision log like you described. That could make building in public feel much more natural.