Hey, I’m Łukasz. I’m building PostSider, a social media publishing tool, and I use it almost every day to promote my other app, Local Waifu.
The biggest thing this taught me is that some features I thought were important really aren’t. And some boring, basic things matter much more once you actually depend on them.
For example, I don’t really need a built-in AI image generator in my scheduling tool. I already use Higgsfield or whatever other tool works best for what I need at the time. That changes constantly. The same goes for AI models: I want to choose what I use, rather than build my whole workflow around one model.
My actual workflow is pretty simple.
I use Hermes through Telegram. I can be walking or at the gym, send it an idea, and talk through a post. It generates a draft and sends it back. When I’m happy with it, I have it schedule the post through PostSider for the day I want.
The topic, context, and kind of content I need keep changing. I want the publishing part to work with that.
PostSider isn’t some huge, revolutionary invention. It publishes content to social media. You connect your accounts through a simple panel, log in, and use it. You don’t have to figure out Meta’s developer platform or build separate API integrations yourself. There’s analytics and other functionality too, but publishing is the main job.
For me, it fits into the system I already use. That turned out to be much more useful than trying to make it do everything.
Using it daily also changed how I see bugs. When something broke, I was often the first person to find out, because I needed it for my own work. Suddenly, fixing that annoying issue mattered more than adding another feature.
Before other customers started using it, it already had a heavy user: me. Someone who used it every day, ran into problems, and needed those problems fixed.
That’s what dogfooding has taught me so far. I understand much better what needs to work, what’s useful, and what I can leave out.
I still need feedback from other users. But building something I actually want to use has given me a much clearer starting point than guessing what an imaginary customer might need.
Anyone else using their own product daily to run or promote another business? What did you end up fixing or removing because of it?
Dogfooding turning "nice-to-have AI features" into "please just publish reliably" is such a useful correction. Being your own heaviest user also changes bug triage — the annoying edge case stops being backlog filler when it blocks your other product's launch day. Curious which single boring reliability fix paid for itself fastest once you were using PostSider every day.
For me, the biggest priority is reliability. Every feature, integration, connection, and publishing flow should just work and stay stable. I’d rather have fewer features that work 100% of the time than a lot of features that randomly break.
Your workflow makes the publishing layer valuable without owning content generation. Have external users independently shown the same preference, or is that still mostly your own usage?
For now I have positive replay, and more sense. If you go and see what is avaliable on the market, problem is no making a content, problem is delivery. Last month was higsfied popular, now is motion graphic from ChatGPT/ Claude next maybe something else, yet social platform integration stay the same, meta don't change his API policy.
We’re already connected by email — I sent you a note there on Tuesday. Reply there when you get a chance and we can continue.