A sudden fall in a rank-tracking report can make a business owner wonder whether their website has disappeared from Google. New research raises a reason to check the report itself before commissioning a rewrite. For organisations in Bath and the South West, the useful question is whether the alert matches what is happening to relevant searches, website visits and enquiries.
What the new research found
Rank-tracking provider Nozzle published its study on 10 October, examining 27.5 million automatically collected search results pages across 49 countries. Its observations cover 20 August to 9 October 2026.
Its test flagged results where fewer than half of a query’s established baseline websites appeared and those websites subsequently returned. In the UK sample, 3% of judged results pages since 13 September met its definition. That figure describes the provider’s sample, rather than all UK Google searches.
Nozzle calls these results “poisoned” and attributes them to efforts against automated scraping. That explanation is an interpretation, rather than a demonstrated motive or Google confirmation. The published method does not describe simultaneous matching searches by ordinary users. It therefore cannot establish what your customers saw. Returning websites also form part of the classification, so recovery is not independent proof of manipulation.
Ask what the alert actually measures
If your reporting service flags a fall, ask for the affected search phrases, pages, collection times, devices and target locations. A report configured for national desktop searches answers a different question from one configured for mobile searches around Bath. Keep those settings consistent when comparing snapshots.
Ask the provider whether it detected a collection problem, changed its reporting method or marked any observations as uncertain. Request the underlying result snapshots where available. A chart showing your site’s position alone gives less context than a saved results page showing which other websites appeared alongside it.
This is particularly useful when an alert reports an abrupt disappearance followed by a return. Keep the original observations rather than silently replacing them. A corrected report should make clear which figures changed and why, so next month’s comparison does not mix different versions of the evidence.
There is no need to run hundreds of searches yourself. Google explains that results vary with the searcher’s time, location, device and recent history. One manual search can add context, but it cannot settle whether every potential customer sees the same position.
Compare the relevant evidence
Use Google Search Console’s Performance report to examine clicks and impressions for the pages and search phrases affected. Check the appropriate country and device, then compare equivalent periods. The whole website’s totals can conceal a loss on one important service page, especially if unrelated pages are gaining traffic.
Average position is an average of the highest result from your site in the selected data. It is not a fixed place reserved for your business. A rank tracker and Search Console can disagree because they measure different sets of observations; disagreement is a prompt to investigate, rather than proof that either number explains everything.
Check data timing as well. Google’s data guidance says the newest figures can remain preliminary and change. Avoid comparing a partly collected day with a complete one, and record the dates used before drawing a conclusion.
Then look at the business side. For a local venue, the useful companion evidence might be wedding enquiries; for a trades firm, quote requests. These are suggested checks, not outcomes demonstrated by Nozzle. Confirm that forms and tracking still work before treating fewer recorded enquiries as fewer real enquiries.
Investigate a real loss without rushing a rewrite
A tracker anomaly is not a reason to ignore persistent losses. Google’s traffic-drop guidance describes several possible causes, including technical faults, changing demand and ranking changes. Small position fluctuations can reverse without intervention; an already effective page does not automatically need radical changes.
If relevant clicks fall repeatedly, ask your website maintainer to check the affected pages and indexing evidence. A broken page or an unintended indexing restriction calls for a specific repair. Rewriting good content would not resolve that fault. Keep a record of changes and how their effect will be checked.
For a broader SEO review, start with an explanation of the loss and its evidence. If a genuine content weakness emerges, targeted website editing can address it. The new study gives businesses a useful extra question to ask about their reports; it does not diagnose an individual website or promise that its rankings will recover.