Grow search traffic with Claude.
Month17 archives your Search Console data forever and serves it to Claude, Cursor, or any MCP client. With years of data to work from, your AI finds the pages worth rewriting and the rankings worth improving.
Free while we build. No card, ever, until you say so.
Representative chart · traffic grows when you fix the pages your history points at.
Search Console data is an asset. Month17 puts your AI to work growing your traffic.
01 / How it works
Connected in five minutes.
Connect GSC
OAuth into Google, pick your properties. Month17 backfills the full 16 months Google still has, then syncs daily, forever.
↓ backfilling 16 months - 483 days saved
Point your AI at it
Paste one config block into Claude or Cursor. That's the whole setup.
"url": "https://month17.com/mcp/"
}
Ask, ship, repeat
Ask about quick wins or decaying pages the way you'd ask a colleague. Ship what it finds and check back on the numbers.
⏺ quick_wins → 14 candidates
02 / In practice
Diagnose a traffic dip in one conversation.
>Traffic dipped since March. Algo update, seasonality, or decay?
⏺month17 · core_update_impact (2024 → today)
⎿ March 2026 core update: −4% · within normal variance
⏺month17 · seasonal_patterns (36 months)
⎿ March–May dip recurs in all 3 archived years
⏺month17 · content_decay (min 3 months decline)
⎿ 2 pages in sustained decline · both fixable
⏺ Not the update. The dip is seasonal: same weeks, same depth, in both archived years before this one. Two pages are genuinely decaying; refresh those and ignore the rest.
✓ answered in one prompt · with 36 months of evidence
>Which pages should I rewrite this quarter?
⏺month17 · content_decay (min 3 months decline)
⎿ 7 pages in sustained decline
⏺month17 · compare_periods (each page · now vs 2 years ago)
⎿ 4 lost rankings · 3 lost only CTR
⏺ Rewrite these four first, starting with /pricing-guide: position 6 → 14, clicks −61% over two years. The three CTR-only pages need new titles, not rewrites; don't spend writing weeks there.
✓ a prioritized rewrite list · with the reason attached to each page
Every call above is one of 22 tools your AI can use →
03 / Data retention
Claude's advice is only as good as the data it can see.
Google deletes Search Console data once it turns 16 months old. Any MCP server reading the live API is stuck with that window. Month17 keeps its own copy, which grows for as long as it runs.
You can't compare year-over-year with 16 months of data, read seasonality without a few full cycles, or judge a core update without the months before it. Everything interesting starts past Google's deletion line.
04 / Pricing
Free right now. On purpose.
Early access is free: connect, backfill, let the history accumulate while we build and improve the product. By the time paid plans arrive, your archive is already safely stored for months. Free while we build. No card, ever, until you say so.
- Permanent storage, full 16-month backfill on connect
- Daily sync from Google's API, under your own OAuth
- 22 MCP tools for Claude, Cursor, any MCP client
It's free, connect your GSC
Later: $10 per month.
05 / Questions
Asked often, answered straight.
Do I need Claude Code?
No. Anything that speaks MCP works: Claude Code, Claude Desktop, Cursor, Windsurf. If you can paste one config block, you're set. There's no dashboard to learn because your AI tool is the interface.
Why free? What's the catch?
No catch: the product gets more valuable the longer your archive runs, so we'd rather you start now. Paid plans come later at $10 per month, and your data leaves with you if you ever want out.
Can you recover data Google already deleted?
No, and nobody can. Once a date leaves the 16-month window it's gone from Google's side. Month17 backfills everything Search Console still has on the day you connect, which is why connecting early matters more than anything else on this page.
How is this different from the free GSC MCP servers?
Same protocol, different substrate. Free servers query the live API and inherit its 16-month window and row caps. Month17 queries a permanent archive that grows daily, hosted, with analyses (year_over_year, seasonal_patterns, content_decay) that only work on multi-year data.
Is my client data safe?
OAuth tokens are encrypted at rest, access is read-only, everything runs on EU servers, and your data is never used beyond your own account. Revoking access is a normal Google operation your client controls. Details in the privacy policy.
The archive you start today is the unfair advantage in two years.
Free during early access. Full backfill on day one. Queryable by your AI from minute five.