Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

SEO › Advanced

Diagnosing a Rankings Drop for a Recipe Publisher

A diagnosis exercise for an invented advertising-funded recipe site that lost much of its search traffic in a fortnight. You rule out tracking, technical blocks and lost links in order, test an algorithm update against an intent shift using page groups and queries, and write a finding, a fix and a recovery measure the evidence can support.
ExerciseAdvanced levelModule 3: Diagnosis, forecasting and the SEO programme5 minUpdated

Why this matters

A sharp traffic loss invites panic: rewrite everything, disavow links, blame the latest update. Each of those destroys the evidence needed to find the real cause, and some make it worse. The guide to sudden rankings drops lists six explanations and insists the quick, provable ones are ruled out first. This exercise adds the advanced step: when two explanations survive, use page groups and queries to tell them apart, and only then decide what to change.

The publisher and its traffic data

Nettlebury Recipe Journal is an invented recipe website funded by advertising, so every lost click is lost revenue. It publishes individual recipes, "best of" roundups that collect its recipes by theme, and ingredient substitution guides.

Illustrative example — not a real result

Nettlebury Recipe Journal: the evidence pack

Totals, fortnight after the drop against the fortnight before: Search Console clicks down 38%; analytics organic sessions down 41%. A confirmed core update began two days before the decline started.

Page groupPagesClicks changeAverage position before and afterImpressions change
Individual recipes1,800−6%4.1 to 4.4−3%
"Best of" roundups420−71%3.2 to 9.8−22%
Substitution guides150−34%2.6 to 2.7−2%
Tag pages900−90%n/a−88%

Change log: new advertising layout with a sticky video player, ten days before the drop; robots rule added to block tag pages, one week before the drop; no template or content changes to roundups or guides.

Crawl and inspection: recipes, roundups and guides return success, are indexable and self-canonical; URL inspection on a sample of twenty roundups shows them indexed. Tag pages blocked as intended; they earned under 1% of clicks.

Links: 25 referring domains lost in the month, out of about 3,400, mostly expired directories, none pointing at roundups. Query sample: for roundup queries, single recipe pages with tested notes and named cooks now rank above Nettlebury; for substitution queries, the results page now shows the answer directly above the organic results.

Ruling causes in and out

Work in order of speed and certainty. First the measurement: if Search Console and analytics disagree, the problem may be the tag rather than the rankings. Then technical blocks, proven or disproved in minutes with URL inspection and the indexing report in GSC. Then links, checked against the pages that actually fell. Only then weigh the slower explanations, a core update and an intent shift, and judge them by shape: an update re-weighing quality hits a type of page wherever it sits; an intent shift changes how a query's SERP looks, often moving clicks without moving position. The Search Console tutorial explains why position and impressions must be read together.

Task

  1. Decide whether the loss is real or a measurement change, with the evidence.
  2. Rule technical blocks in or out, including the robots change in the log.
  3. Rule lost links in or out.
  4. Test the core update against an intent shift, page group by page group.
  5. Write the diagnosis in a short paragraph a non-specialist owner could follow.
  6. Set out the fix, in order, and say what you will not change.
  7. Define how recovery will be measured and over what horizon.

Worked answer

Real. Clicks and sessions fell together, so the tag is not the cause.

Not a technical block. Affected pages are crawlable, indexed and self-canonical. The robots rule removed tag pages, which explains their collapse, but they carried almost no clicks; that change was intended and is not the story. The advertising layout is worth watching for page experience, but it would hit every template, and individual recipes barely moved.

Not links. The domains lost were few, weak and pointed elsewhere.

Two causes, two page groups. Roundups lost position heavily, starting with a confirmed core update, while recipes on the same site held: that is a quality re-evaluation aimed at a page type. Rivals now outranking them show tested, first-hand content, which is what the E-E-A-T guide describes as experience. Substitution guides kept their position but lost clicks, because the results page now answers the question itself: an intent and format shift, not a ranking loss.

Illustrative example — not a real result

Nettlebury: share of the lost clicks by cause

Before the drop, roundups earned about 42% of search clicks, individual recipes about 45%, substitution guides about 12% and tag pages about 1%. Of the 38-point fall, the roundups' 71% loss explains about 30 points, the guides about 4, recipes about 3 and tag pages about 1. The roundups are where recovery effort should go, and the guides' loss should be reported separately because it will not return with better content alone.

Fix. Treat the roundups as a quality programme, not a cosmetic edit: merge overlapping roundups, rebuild the strongest with tested notes, original photos and a named cook, and retire the thinnest. Rank which to rebuild first with the SEO traffic estimator for keyword prioritisation, entering realistic positions. Run the work through the refresh queue in the SEO content operations playbook. For substitution guides, accept fewer clicks on simple answers and add depth that earns the visit: why the swap works and what changes in the dish. Do not disavow links, roll back the robots rule or rewrite the recipes that held.

Recovery measure. Search Console clicks and position by page group against the pre-update baseline, read across the next core updates rather than week to week, with every change annotated; advertising revenue per organic session alongside, because that is what the owner loses.

Common mistakes

  • Rewriting everything at once. No single change can then be judged.
  • Blaming the robots change. It explains the tag pages, not the roundups.
  • Reading a click loss as a ranking loss. The guides held position; the results page changed.
  • Weekly recovery checks after a core update. Quality re-evaluations are judged across updates.

Take what you found into the SEO strategy capstone, or revisit the intermediate SEO level for SEO KPI reporting.

Pages this lesson applies

Terms used on this page

Also related

Sources

  1. Performance report (Search results) — Google Search Console Help
  2. Page Indexing report — Google Search Console Help
  3. URL Inspection Tool — Google Search Console Help

Related