The problem
A core update took 70% of organic clicks in four months — 231K down to 69.2K — with impressions falling 60% and average position sliding from 23.2 to 32. Recoveries like this usually stall because the response gets split across separate vendors: one for technical, one for content, one for links. Each optimises its own slice and nobody owns the interaction between them.
What I did
- Five workstreams, one owner. Technical and on-page SEO, semantic optimisation, topical authority, internal linking, and third-party mentions — run together rather than sequenced.
The outcome
Average position recovered from the post-decline low of 32 to 8.2. That is the number to judge this by: it measures whether the site earned its standing back, independent of how traffic happened to be distributed across the period.
The decline
| Metric | Mar – Jul 2025 | Jul – Nov 2025 | Change |
|---|---|---|---|
| Clicks | 231K | 69.2K | −70% |
| Impressions | 27.5M | 11M | −60% |
| Avg. position | 23.2 | 32 | −8.8 |
Search Console. Note that Search Console lists the earlier window first, which reads like the current period — it is not.
The recovery
| Metric | Post-decline low | 12 months to May 2026 |
|---|---|---|
| Avg. position | 32 | 8.2 |
| Clicks | 69.2K over 4 months | 1.01M |
| Impressions | 11M over 4 months | 153M |
Search Console. The 12-month totals span the decline as well as the recovery, so they understate the recovered run-rate — read the position figure, not the totals.