Diagnose a traffic drop

Watch · 2:42

• 0:00The shape of the drop
• 0:33The workflow
• 2:11Gotchas

In shortTo diagnose a traffic drop, first confirm it's real and find where it happened: search only, or every channel. Then use the right tool for the shape of the drop: the Anomaly engine for a sudden one-day fall, Algorithm impact for a drop after a Google update, the URL ranking analyzer to see which pages lost clicks, and Content decay radar for slow declines. Finish with a fix list from Site audit, Core Web Vitals and content refreshes.

Start with the shape of the drop

"Traffic is down" can mean five different problems. A sudden fall across every channel is usually tracking. A sudden fall in Google traffic only is often an algorithm update or a technical error such as an accidental noindex. A slow slide over months is content decay. Steady impressions with falling clicks can mean the results page changed, for example an AI Overview now answers the question. The tools are cheap individually, so the skill is picking the right one first.

The workflow

1. Confirm the drop and where it happenedOpen Analytics → Search Console → GSC overview, which compares the current period with the previous one, and Analytics → Google Analytics 4 → GA4 overview. If GA4 fell but Search Console clicks didn't, the problem isn't Google. Cost: free. Lessons: Connect Search Console, Connect GA4.
2. For a sudden one-day fall, run the Anomaly engineOpen Analytics → Google Analytics 4 → Anomaly engine and click Detect Anomaly. It compares the latest day with the same weekday in the last two weeks and splits a drop over 15% by source, country and device. Check the date: the latest day is usually today, which isn't over yet. Cost: per analysis. Lesson: Anomaly engine.
3. Rule out spam leavingIf sessions fell but engagement rate rose, a spam source may have stopped. Open Analytics → Google Analytics 4 → Spam purger to see which sources had near-zero engagement. Losing spam isn't a real loss. Cost: per analysis. Lesson: Spam purger.
4. Line the drop up with a Google updateOpen Analytics → Search Console → Algorithm impact, choose the update date and click Analyze Impact. It lists pages that gained or lost more than 10% of clicks around that date. The "after" window always runs from the update date to today, whatever Days After says, and raw totals are compared, so windows of different lengths can look like gains or losses; read the per-page changes, not the headline. Cost: per analysis. Lesson: Algorithm impact.
5. Find exactly which pages lost clicksOpen Analytics → Search Console → URL ranking analyzer, set the period before and the period after the drop, and read the Clicks Delta column. A handful of pages usually explain most of it. Cost: per analysis. Lesson: URL ranking analyzer.
6. For slow declines, run Content decay radarOpen Analytics → Search Console → Content decay radar and click Analyze Decay. It lists pages with a steady downward trend, most urgent first. It only includes pages averaging more than 100 clicks a week, so small sites may see none. Cost: per analysis. Lesson: Content decay radar.
7. If clicks fell but impressions didn't, check the results pageOpen Analytics → Search Console → CTR opportunities to find queries with high impressions and low CTR. For the biggest ones, add the query as a prompt and run it on the Google AI Overview engine in AI search → Brand visibility → Prompt research to see whether an AI Overview appears and whether it cites you. Those Brand visibility runs always use US English results; for another country, use the AI Overview batch check at /dashboard/brand-visibility/ai-overview, which takes a country and language per query (it isn't in the sidebar; see SERP analysis in batches). Cost: CTR analysis per run; the prompt run per engine; the AI Overview check per query. Lessons: CTR opportunities, Run prompts across AI engines.
8. Turn findings into a fix listFor technical causes, crawl with SEO → Audits → Site audit and check affected templates in SEO → Audits → Core Web Vitals. For decayed or out-ranked pages, refresh them with sharper answers and current facts, using the AI content writer if helpful. Cost: per page crawled, per URL scanned, per article. Lessons: Site audit, Core Web Vitals, Write your first article.

Gotchas