Built a tool to automate my dev marketing. No real traction yet. What am I missing?

This is very good tool for developers, Honestly "get them to try" is usually a selection problem wearing an activation costume. If you incentivize the trial itself — discounts, "just try it" — you pull people motivated by the incentive, not the product, and they bounce. What's worked better for me is making the thing they unlock something they already wanted, so finishing the step self-selects for real interest. (Building in this space right now — happy to share what I'm seeing if useful.)

The fact that you explicitly blocked em dashes and "Excited to announce" openers tells me you actually get the problem, not just the symptom. Most AI marketing tools just crank out more slop, faster.

One thing I have noticed about the blank-screen problem you mentioned: part of it is a writing bottleneck, not a content bottleneck. When I cannot figure out how to tweet about a bug fix, it is usually because typing feels too formal for something small. The fix for me was dictation. I hold a hotkey, speak the post out loud like I am telling a friend what I shipped, and the text lands in the composer. It comes out more natural than anything I would type.

I built DictaFlow for this exact use case. Hold-to-talk dictation that works in any app, including social media composers. It is a different approach to the same problem PushToPost solves: getting from "I shipped something" to "here is a real post about it" without the friction of staring at a cursor.

Good luck with the launch. The anti-slop scoring system is the right instinct.

The "killed the em dashes" detail made me laugh, literally got called out on the same thing in another conversation today. Real one.

Honest take on your category question: I don't think the category is dead, but the ICP framing might be inverted. Devs who push code are often the worst customers for social marketing tools, because if they cared about marketing they'd already be doing it, and if they don't, the last thing they want is a tool that generates more output they then have to manage. The ones who'd actually buy this are non-dev or barely-dev founders who happen to ship code: solo indie hackers, technical PMs, agency owners. That probably means different copy on the landing, and different proof.

On the trial bar: in 2026 with AI everywhere, I think the threshold moved from "try the tool with a signup" to "show me one generated tweet from a public repo in 10 seconds, no email required". The friction of signing up just to see if it sounds human is probably what's killing the first interaction more than the trial length itself.

The category is fine. The framing is the problem.
“Turn GitHub pushes into posts” sounds like a utility. But the real pain you’re solving is that technical founders ship meaningful work every week and the market never sees it — because translating commits into public momentum is slow and inconsistent.
I ran into the exact same framing problem with my page. Our first description was “monitors 6 platforms for mentions of your product.” Accurate, but nobody cared. When we changed it to “someone just asked for a product like yours — will you answer first?” signups jumped. Same product, completely different emotional response.
On your specific questions: the 3-day trial is fine, but I’d remove friction elsewhere first. What does the first 60 seconds look like after signup? If someone has to connect a repo, configure settings, and wait for a push before seeing any output — most will leave before the trial starts.
The biggest unlock might be showing a real example output before signup. Not a marketing screenshot — an actual post generated from a real commit, with the commit message visible next to it. That builds trust faster than any headline

I'd split this into two jobs: release notes for existing users, and public marketing for strangers. A commit can produce the first pretty reliably because the audience already cares about the product. The second needs a stronger filter: why this change matters, who it helps, and whether it proves momentum worth sharing.

If I were testing this, I'd make the first onboarding path generate three outputs from one repo: internal changelog, customer-facing release note, and founder post. Then let users approve/edit before anything publishes. That makes it feel less like auto-posting and more like a shipping communication workflow.

I don’t think dev marketing is a bad category. I think the current framing makes it sound smaller than the real problem.

The real pain is not “turn GitHub pushes into posts.” It is that technical founders ship meaningful progress, but the market never sees it because translating product work into clear public momentum is annoying and inconsistent.

That is a much stronger angle than automated posts.

The naming is worth pressure-testing too. PushToPost is clear, but it frames the product as a push-to-social utility. If this grows into changelogs, launch updates, founder marketing, SEO pages, release history, and repo-based distribution, the name may box it into a narrow automation feature before users understand the bigger workflow.

Xevoa .com would fit the broader direction better: a clean dev/workflow automation brand that can carry product updates, release intelligence, and technical marketing without sounding like another AI posting bot.

Since you are already seeing that generic AI marketing language kills trust, the brand has to work harder in your favor. The product should feel like it helps founders communicate real shipping momentum, not just auto-generate content from commits.