Query Search Console data older than Google's 16-month limit

The scenario

You want to compare this quarter's organic performance to the same quarter two years ago. Standard Search Console returns nothing past 16 months. Google purged it. You either (a) pulled monthly CSVs every month for years and now grep through them, (b) pay an enterprise tool $300+/month for the same archival, or (c) accept the data is gone. Option (d): GSC PAP's archive.

The prompt

Use the gsc-pap MCP. For https://example.com, compare clicks and impressions for Q1 2024 vs Q1 2026 using search_analytics_archive (Google's live API doesn't go that far back). Group by query, return the top 30 by Q1 2026 clicks, with Q1 2024 numbers and the YoY delta and pct_change for each.

What happens

1. 1

   AI checks archive coverage with archive_status

   Before running a long-range query the AI calls `archive_status(site_url)` to confirm the archive actually has the date range you want. If a site was added recently the archive may not stretch back to your earliest target date. `earliest_date` and `latest_date` per site tell you what's available.

2. 2

   AI calls search_analytics_archive for both windows

   Two calls: one for Q1 2024, one for Q1 2026. Each returns an aggregated table by query. The AI joins on the query string in-memory and computes deltas + percentage change.

3. 3

   AI surfaces the storyline

   Output is the diff table plus a brief narrative: which query categories grew fastest, which collapsed (often a sign of intent shift or AI Overview eating clicks), which net-new queries appeared. With historical data the AI can give you context Google's 16-month UI cannot.

Outcome

If your archive has a couple of years of coverage, you get YoY analysis Google literally cannot show you. Most sites discover at least one query category that has quietly halved in the last 24 months while CTR for new categories doubled, the kind of slow trend invisible in 16-month-window reports.

Where to take it next

• ·Trigger a backfill from the dashboard (/account/archive) to grab the maximum 16 months Google still has on the live API; everything after that point we keep forever.
• ·Pair with `historical_position_trend` for a query-level deep dive across the same long window.
• ·Combine with `correlate_gsc_to_ga4` (where GA4 history allows) for a full-funnel YoY view.