Detect SEO decay. Fix it before noon.
Real-time GSC ingest, SERP comparison, and content-decay detection — all running on the same event bus as your attribution and automation. When a top-10 page slips, the refresh is in your editor's queue before the second cup of coffee.
Most SEO tools surface issues. They don't fix them.
Weekly reports. Manual triage. By the time someone notices, the campaign has already lost a position.
Weekly reports, daily reality.
Most SEO platforms run nightly batches and email you a Monday digest. By Wednesday the position is already gone.
"Page slipped." Now what?
Reports flag the issue but don't act on it. You read the dashboard, copy the URL into a Google Doc, file a ticket, wait two weeks.
AI Overview moved the goalposts.
Position-1 organic now has AI Overview above it. Most tools don't track AI Overview presence, so you can't optimize for it.
Detect. Decide. Dispatch. Audit.
Every GSC delta and SERP change flows through the same four-step pipeline.
GSC delta arrives
Per-property quota-aware ingest. CTR, position, impressions, CWV, AI Overview presence — every metric tagged with shop + page + traceId.
Decay or opportunity?
9 detectors run in parallel: rank decay, CTR drop, AI Overview gain/loss, schema-fit, content-gap, internal-link orphan, CWV regression, broken canonical, soft-404.
Refresh queued
If detector fires + confidence > threshold, an AI refresh draft is generated and queued in the CMS approval flow. Editor approves, page republishes.
Logged + traced
Plain-language audit row written: 'Refreshed /blog/X because decay-detector flagged -34% CTR sustained 7 days.' 365-day retention.
Five sources. One detector. Four outputs.
The exact runtime topology. Hover any node to inspect.
Three concrete moves customers made.
Anonymised but representative. Pulled from active deployments.
Detected, queued, fixed before lunch
A SaaS-comparison site lost a top-10 page Tuesday morning. Sumeru detected the rank drop within 60 seconds of GSC reflecting the change, queued an AI refresh draft by 9:42 AM, the editor approved it at 11, and the page was republished by 11:14 — recovered position by Friday.
Catches what Looker dashboards miss
An $18M GMV brand replaced a weekly 'SEO health' dashboard with the live decay detector. Average catches per week jumped from 3 (manual review) to 47 (real-time detection). Most are small drops fixed before they cascade.
AI Overview opportunity surfacing
Detector flags pages newly eligible for AI Overview placement — based on query class, content depth, and schema fit. Customer in the home-goods space saw 11× citation rate in AI Overview within 3 weeks of acting on the recommendations.
What's available where.
| Capability | Starter | Pro | Agency | Enterprise |
|---|---|---|---|---|
| GSC ingest (queries · pages · CTR · CWV) | ✓ | ✓ | ✓ | ✓ |
| Real-time decay detection | — | ✓ | ✓ | ✓ |
| AI Overview opportunity flagging | — | ✓ | ✓ | ✓ |
| Content gap detection (vs competitor SERP) | — | — | ✓ | ✓ |
| Auto-queue refresh drafts | — | ✓ | ✓ | ✓ |
| Schema-fit + structured data audit | — | ✓ | ✓ | ✓ |
| Internal-link graph optimization | — | — | ✓ | ✓ |
| P95 detection latency | 10m | <60s | <60s | <30s |
SEO Intelligence speaks to the rest of the runtime.
Detectors fire events, automation handlers act on them, audit log persists every move.
See decay detection on your own catalog.
We connect to your GSC, run the detectors against your top 100 pages, and walk you through the live results. You leave with a refresh queue, in writing.