i shipped 20 single-purpose AI writing generators in 48 hours after a generic one flopped. 0 sales yet — roast my distribution plan?
ok so this is gonna sound dumb but bear with me.
24h ago i had a generic ai text humanizer up. 0 sales. spent another 6 hours polishing it. still 0. clearly the offer was the problem, not the polish.
so i scraped a github list of 373 indie microsaas and stared at what was actually getting sold. the pattern is hyper-niche: ai wedding toast, ai cover letter, ai eulogy. specific deadline-bound pain, single $9 transaction.
shipped 9 single-purpose generators in 2 hours that first afternoon on a factory backend. extended to 20 over 2 days. each one is its own URL (aitells.vercel.app/eulogy, /wedding-vows, /apology-letter, /salary-negotiation, /obituary etc) targeting one search intent.
each niche page took ~12 minutes after the factory was built. one /api/generate endpoint, one reusable client component, thin page.tsx per niche.
48h in. $0 revenue. still believe SEO compounds in 4-8 weeks and chatgpt/perplexity start citing single-purpose URLs once indexed.
stuff i'm doing while i wait:
- indexnow pinged bing/yandex
- 2 dev.to articles published (one with the full factory architecture)
- sitemap submitted
- PR to awesome-ai-tools just opened
- linkedin + twitter posts up
- show HN submitted
stuff i'm probably not doing well enough:
- partnerships with people who already have audience (wedding planners, career coaches, funeral homes)
- paid micro-ads to validate fit before betting on seo
- video demo / tiktok content
- price test (maybe $9 is wrong for a brand new site with 0 reviews)
would love honest feedback. is this the right approach or should i pivot completely? aitells.vercel.app if you want to look. all 20 generators have free preview, $9 unlocks all of them.
specifically curious: for folks who've shipped niche AI wrappers, what was your indexation timeline + what worked in week 1-4 before SEO kicked in?
I had a similar slog and the first couple days were crickets for me too. What helped was picking one niche and going deep instead of spreading myself across a bunch. I also got way more traction by adding tiny free samples so people could test the vibe before paying. Partnerships only moved after I reached out with a super specific angle rather than a generic pitch.
The distribution problem is real, but I think the bigger issue is trust.
Single-purpose pages make sense for SEO, but aitells.vercel.app makes each generator feel like a quick experiment, not something people should pay for in sensitive moments like eulogies, apologies, salary negotiation, or wedding vows.
The niche strategy is sharp. The brand layer is the weak part.
If this becomes a serious suite of high-intent writing tools, a cleaner .com like Beryxa.com would probably carry more trust than a Vercel subdomain and generic AI wrapper framing.