Clustea has quietly become a completely different product over the past few weeks
What it does: connect a WordPress (or now Shopify, Ghost, HubSpot, Notion) site, and it researches keywords, writes SEO articles grounded in real competitor data and live web search, and publishes them — on autopilot, every week.
What changed:
Content quality — every article goes through actual research before a word is written: it scrapes the real top-ranking pages for the target keyword, builds a competitor brief, then writes with live web search so every stat is sourced, not guessed. 19 automated quality checks + a repair pass before anything ships. No word count slider anymore either — the pipeline reads what's actually ranking and lets the topic decide the length.
Publishing destinations — went from WordPress-only to 5 platforms: WordPress, Shopify, Ghost, HubSpot, and Notion (useful as a review queue if you want a human to approve before it goes live anywhere).
It watches its own work — a cron checks every published article's Search Console performance, and if one starts losing clicks or ranking worse after 90+ days, it automatically re-researches and republishes it. Not a fixed schedule — it reacts to real decay.
Built for scale — agencies managing multiple client sites can now paste up to 50 keywords across different projects, let them generate in the background, and see everything on one editorial calendar instead of juggling tabs.
The honest test: I run all of this on Clustea's own site too. Went from 416 impressions over 3 months to 1,280 in 7 days after fixing an indexing bug the pipeline itself would've caught if I'd been dogfooding sooner.
Still solo-built, still iterating fast. If you're an agency drowning in content production for client sites, this is exactly who it's for now.
Going from article generation to performance-based maintenance makes this much more defensible than another AI writer. The proof now needs to be more than impressions: show whether refreshed posts recover clicks, rankings, and conversions consistently across several sites before making the automation fully hands-off.
The re-publishing automation based on Search Console decay is the real insight here. Most SEO tools publish once and disappear, but content ranking velocity changes constantly. Building in autonomous feedback loops that actually react to real performance data instead of a fixed schedule is how you move from a tool to a system.
Curious if the 90-day window is based on data you collected, or something to iterate on. That threshold seems critical - too short and you're thrashing, too long and you're leaving money on the table.
The shift toward agencies managing multiple client sites seems like the biggest strategic move here.
What convinced you that agencies are the right customer segment to build around, rather than continuing to serve individual site owners who also need automated SEO?