Blog · Algorithm Watch

Google rolled out the March 2026 core update over 11 days starting March 4. We tracked it across 40+ client GBPs in California spanning HVAC, dental, med spa, plumbing, legal, and multi-location systems. By March 22 we had a clear picture of what got hit, what didn't, and why. This is that picture.

The TL;DR: the update targeted "thin local pages" — service pages with minimal content, no real local context, and weak entity signals. Map-pack rankings shifted hardest in markets where competitors had stronger entity signals (knowledge-graph presence, Wikidata entries, branded citations). Organic local rankings shifted hardest where service pages were under 300 words or had no real local-context content.

What Got Hit Hardest

Three patterns came up consistently across the 40+ client base:

Pattern 1 — Thin service pages. Clients with service-area pages under 300 words lost 10-30% organic local position within the first week. These were the pages where we'd already flagged content depth as a Phase 2 priority — the update accelerated the urgency. By end of update, the affected pages either needed depth expansion or de-indexation.

Pattern 2 — Weak entity signals. Clients whose entity profile was thin in Google's knowledge graph (no Wikidata entry, weak brand-mention footprint, limited cross-platform consistency) saw map-pack position erosion. Not always dramatic, but consistent — 1-3 position drops in markets where competitors had stronger entity profiles.

Pattern 3 — Stale GBP cadence. Clients we'd onboarded recently — before the GBP cadence had time to compound — were more exposed than long-tenure clients. The update seemed to weight "GBP signals over the last 90 days" more heavily than prior updates. Recency matters more than it used to.

What Held Up

Clients with deep service-area pages (1,200+ words, real local context, schema markup, internal linking), strong entity profiles, and consistent 90-day GBP cadence held their positions. A few moved up — picked up positions vacated by competitors who got hit harder.

This is the pattern we've been preparing for over the last 18 months. The "thin local content + weak entity + sporadic GBP" pattern was always going to be exposed eventually. Most agencies' clients sit squarely in that pattern. Ours, increasingly, don't.

What We Changed

Three adjustments we rolled across the client base within 7 days of the update completing:

Adjustment 1 — Content depth audit. Every service-area page under 800 words got flagged for depth expansion. Where the keyword-target volume justified it, we expanded. Where it didn't, we consolidated or noindex'd.

Adjustment 2 — Entity profile push. Accelerated the entity-optimization work — Wikidata entries, knowledge-panel claims, structured-data audits across the board. The signals the update rewarded are signals we can build deliberately.

Adjustment 3 — GBP cadence floor. Where we'd been running 1-2 GBP posts/week on some accounts, we moved everyone to 2-3/week minimum. Photo cadence too. The "alive business" signal weight is up, and we're matching it.

What This Means For You

If your local rankings dropped between March 4-22 and you don't know why — this is probably it. The exposure pattern is documented. The fix is documented. The work is straightforward, not mysterious. See Local SEO Recovery for how we handle update casualties.

If your local rankings didn't drop, that's the signal that your foundation is strong. Don't get complacent — the next update will target the next pattern. But for now: keep doing what's working, watch for the next pattern, and bank the position you have.

The agencies that read the updates beat the ones that don't, every single cycle. Most agencies still don't read them.

We'll be tracking the next update from the moment it announces. Subscribe to the blog if you want the patterns documented in real time — and the playbook adjustments we make.