Your organic traffic fell off a cliff and the Slack channel is on fire. Before anyone touches anything, take a breath: most traffic drops are diagnosable, most are fixable, and a surprising number aren't real drops at all. The expensive mistakes happen in the first hour — panicking, changing things blindly, and blaming the algorithm for what was actually a broken redirect.
This is a triage playbook. Work it in order, write down what you find at each step, and resist the urge to "fix" anything until the diagnosis is done.
Step zero: is the drop even real?
Before diagnosing, rule out the false alarms. They account for a depressing share of "emergencies":
- Check the date range. Comparing a week that included a bank holiday against the week before will always look like a disaster. Compare year-on-year for seasonal businesses, and week-on-week for the same weekdays.
- Check tracking. A removed analytics tag, a consent-banner change that now blocks measurement, or a GA4 filter someone added "temporarily" all look exactly like traffic loss. Verify in a second source (Search Console) before doing anything else.
- Check Search Console, not just analytics. If GA4 shows a drop but Search Console clicks are steady, your measurement broke, not your rankings.
- Check scope. Is it the whole site, one section, or one page? A site-wide drop and a single-page drop are completely different investigations.
The five suspects, in order of likelihood
1. Something technical broke
The most common real cause. A deployment that added noindex to staging templates and pushed them live. A migration with missing redirects. A robots.txt edit. A canonical pointing at the wrong place. A server upgrade that slowed the site to a crawl. Check: Search Console's Page Indexing report for sudden spikes in "Excluded", the URL Inspection tool on affected pages, and your deployment log against the date the drop started.
2. A Google update rolled out
Core updates, spam updates and helpful-content-style updates roll through several times a year. If the drop date lines up with a confirmed update and the affected pages are thin, templated or heavily AI-generated, this is your prime suspect. The fix is never technical — it's content quality, reviewed honestly page by page.
3. The SERP changed around you
Sometimes you didn't fall; the page did. Google adds a featured snippet, a "People also ask" box, shopping results or an AI overview above your listing, and clicks drain away even though your ranking never moved. Compare impressions vs clicks in Search Console: impressions steady + clicks down = a SERP-layout problem, not a ranking problem.
4. A competitor took your lunch
One strong new page can displace you on your money keywords. Check your top 10 keywords: are the same competitors there, or has someone new appeared? If a rival published something genuinely better — fresher data, real expertise, better structure — no technical fix will win it back. You need better content.
5. Demand itself moved
Seasonality, news cycles, a product going out of fashion. Search demand for "garden offices" doesn't look the same in January as in May. Google Trends is your friend here: if interest in the whole topic fell, that's market reality, not an SEO failure. Budget accordingly instead of chasing rankings that can't convert.
The triage sequence (do this in order)
- Pin the date. When exactly did it start? One day = technical or update. Gradual = competition or content decay.
- Scope it. Whole site, section, or page? Whole site = technical/update. Section = template or category issue. Page = content or competitor.
- Check Search Console. Indexing errors, manual actions, security issues — the free answers first.
- Check the changelog. Site changes, content changes, agency changes, tracking changes. Most drops have a human fingerprint.
- Check the SERP. Search your top keywords. New features? New competitors? Different intent?
- Only then, plan the fix. One hypothesis, one change, one measurement period.
Communicating the drop to stakeholders
The technical diagnosis is half the job; the other half is explaining it to people who just see a red graph. Three rules:
- Lead with the cause, not the data. "A deployment on Tuesday added noindex to our service templates; we've reverted it and submitted the pages for reindexing" beats ten charts. Non-specialists need the story first, the evidence second.
- Give a realistic timeline. Technical fix: days to weeks. Content rebuild: months. Algorithm hit: no promises, here's the plan. Nothing destroys trust faster than a missed "we'll be back in two weeks".
- Separate traffic from revenue. A 30% traffic drop on informational content with a 2% enquiry impact is a different conversation from a 10% drop concentrated on money pages. Always report the business impact alongside the traffic number — it stops panicked budget cuts to the thing that's actually working.
What recovery actually looks like
Honesty matters here: technical fixes recover fast (days to weeks once Google recrawls). Content-quality recoveries are slow (weeks to months, and only if the improvements are real). Algorithm-update hits don't "recover" at all in the old sense — you rebuild on better foundations. Anyone promising a 30-day recovery from a core update is selling you something.
And write the incident up. A one-page note — what dropped, when, what caused it, what fixed it — turns a crisis into institutional knowledge. The next drop gets diagnosed in an hour instead of a day.
The post-mortem template (turn crisis into knowledge)
After every traffic incident — real or false alarm — write it up while it's fresh. One page: what happened (the numbers, the dates), what caused it (the evidence, not the first theory), what fixed it (and how long recovery took), what we'd do differently (the monitoring or process change that catches it earlier next time). File it where the next person will find it. Teams that do this diagnose repeat incidents in minutes; teams that don't relitigate the same mysteries every year. The template is also your defence against the most dangerous post-drop sentence in business: "let's just change everything and see".
And one final habit: after recovery, don't declare victory too early. Watch the affected pages for a full month — algorithm recoveries in particular can wobble before stabilising, and mistaking a dead-cat bounce for recovery leads to premature celebration followed by a second, more embarrassing incident report.
Traffic drops are stressful, and an outside pair of eyes helps — we diagnose these weekly. Our SEO audit service pinpoints the cause with evidence, not hunches. If a competitor is the problem, our SEO competitor analysis guide shows you how to size up what they did. Get a free consultation.
