How to Read a Ranking Drop After a Google Core Update

Every time Google announces a core update, a certain kind of panic starts. Ranking trackers wobble, forums fill up, and somebody tells a local business owner they need to rewrite every page by Friday. Almost all of that is noise, and Google has told us so in writing.

This is my reading of what Google actually says about core updates, applied to a London local business, with the two 2026 updates as the working example. I am separating what Google has published from what I think, because the two get blurred constantly in this industry and it costs people money.

What core updates did Google run in 2026?

Google listed two core updates in 2026 so far: the March 2026 core update and the May 2026 core update. The Search Status Dashboard records the March update starting on 27 March and running for 12 days, and the May update starting on 21 May and completing on 2 June. Google also listed a March spam update, a Discover update in February, and further spam updates in June, August and September.

Source: Google Search Status Dashboard, Ranking history. Times on that dashboard are US Pacific. The separation matters because a spam update and a core update are different instruments. A spam update enforces Google's spam policies. A core update re-weighs how helpful and reliable content is across the whole web.

What does Google say a core update actually is?

Google describes core updates as significant, broad changes to its search algorithms and systems that do not target specific sites or individual pages. It happens several times a year, Google announces it, and in its own words most sites do not need to worry and may not even notice one happened.

The analogy in Google's documentation is the one I find most useful. Imagine you wrote a list of your twenty favourite restaurants in 2019. Since then, new restaurants have opened, some of the old ones have got better, and your friend now wants somewhere dog friendly. The list changes. The restaurants that dropped are not necessarily bad. There are simply other restaurants that now make the twenty.

That framing should change how you feel about a drop. A fall after a core update is not a penalty and it is not a verdict that your business is poor. It is a relative result: somebody else's page now answers the question better, or Google's idea of what a good answer looks like has shifted.

When should I check whether my site was affected?

Wait at least a full week after the update finishes rolling out, then compare that week with the week before the rollout started. That is Google's own instruction. The reason is simple: during a rollout rankings move around, and judging the result halfway through is like reading an election on the first count.

1. Confirm the update has finished. Check the Search Status Dashboard for the end date, not a blog post or a tracker's opinion.
2. Compare the right dates. In Search Console, set the week after completion against the week before the start. Not the day before, not the month before.
3. Look at your top pages and queries. Ask whether the drop is small or large.
4. Split by search type. Web, Images, Video and News move independently, so a drop in one tells you nothing about the others.

For the May 2026 update that meant waiting until the week after 2 June before judging anything. If you were reading your numbers on 25 May, you were reading noise.

How big a drop is a real problem?

Google draws the line at position 2 to 4 being small and position 4 to 29 being large. Those are its own worked examples. For a small drop it recommends no drastic action and actually advises avoiding changes to content that is already performing well. A large, sustained drop across the site as a whole is the signal to do a deeper assessment.

Is the Map Pack affected by a core update?

Google's core update guidance is about web search rankings, and it says nothing specific about local results. The Map Pack is driven by the Business Profile and the three factors Google publishes for local ranking: relevance, distance and prominence. A drop in the pack and a drop in organic results are different problems with different fixes.

You will find plenty of articles online claiming to know exactly what a given core update changed for local pack rankings. Google did not publish that, and anyone claiming precision is inferring it from tracking data. Treat those claims as hypotheses. What you can say with confidence is that local results sit alongside organic results on the same page, so a change in either can change what a searcher sees.

If your pack position fell, I would start with the usual suspects before blaming an update: a changed primary category, a new competitor nearby, a review flow that stalled, or a profile edit that triggered re-verification. I have written up the common reasons a profile loses Map Pack position separately, and those causes explain far more drops than any update does.

What should a London business check first?

Start with the pages that were most affected, then ask honestly whether another page on the web does a better job for the searcher. Google's self-assessment questions are about whether your content is helpful, reliable and written for people first. They are uncomfortable on purpose.

• Is the page the same as everyone else's? A page about boiler servicing in London that says what every boiler page says is a commodity page. I cover what to do about that in the post on non-commodity local content.
• Does it show first-hand knowledge? Named engineers, real jobs, real postcodes, real photos.
• Is the information current? Opening hours, prices, service areas, team members.
• Is the business clearly identifiable? A real address, a real phone number and a named person. This is the entity layer I describe in what entity building actually is.
• Would a customer feel they got what they came for? Google's whole test reduces to this.

A useful trick from Google's guidance: ask people who have no connection to your site to run the assessment. You cannot be objective about your own homepage. A friend who has never seen it can be.

Why is the average position number misleading?

Average position blends every query a site appears for, so it can fall sharply while the searches that matter commercially hold steady, and it can look healthy while they collapse. Two of the case studies on this site show the effect from opposite directions.

Dr Kish Aesthetic Center carries an average position of 43.5, which looks dreadful, alongside a 2.5% click rate, the best of the six projects published here. Thousands of irrelevant impressions drag the average down while the small set of commercial terms rank at the top. Brow Queen LA is the opposite shape: over a million impressions, an average position of 18.5 and a 0.4% click rate. The big number is mostly visibility on page two.

After a core update, this is why you segment before you conclude. Look at your ten most valuable queries individually. A fall in the average that comes entirely from long-tail impressions you never wanted is not a problem. A fall on your three money terms is.

What should I not do after a core update?

Do not make quick-fix changes, and do not delete content unless it cannot be salvaged. Google lists both explicitly. Its example of a quick fix is removing a page element because you heard it was bad for SEO. Its view of deletion is that it is a last resort, and that if you are considering removing entire sections of a site, that is probably a sign they were made for search engines first.

• Do not rewrite pages that held position. You are changing a variable that was working.
• Do not buy links in a panic. A drop in a relative ranking is not fixed by a manufactured prominence signal.
• Do not rush an AI rewrite of the whole site. Google's guidance on generative AI content is that it must still meet the spam policies. Mass rewriting is the fastest way to turn a quality question into a spam question.
• Do not change categories, names and addresses on your Business Profile hoping to shake something loose. Edits can trigger re-verification and you will not know which change mattered.

How long does recovery take?

Some changes can show within days, but Google says it can take several months for its systems to confirm a site is now consistently producing helpful content. If you have worked on the site for a few months and seen nothing, Google says that can mean waiting until the next core update.

It also says you do not strictly have to wait for a major update, because smaller core updates run continuously and are not announced. In practice this means improvements can pay off at any time, which is a better way to think about it than treating each announced update as an exam you either pass or fail.

My own timeline advice for a London local business is deliberately dull. Month one: do nothing except measure properly and confirm the drop is real. Months two and three: improve the weakest of your affected pages for readers, not for a tracker. Month four onwards: judge the result against the same window a year earlier where you can, because London local demand is seasonal and a straight before and after comparison lies to you.