Google Core Update: What It Is and How to Check Whether It Hit You

After a core update there is one question: did it hit me? I wanted to know what that looks like for my own site if I apply exactly the method Google itself recommends. So I analyzed the Search Console data for the three most recent core updates: December 2025, March 2026 and May 2026. The result is further down. In short: each update had a large drop for one or two search queries.

This guide explains what a core update is, lists all core updates from 2023 with their duration and shows Google’s check method step by step. Each update has its own article, which I link at the relevant points.

Google describes core updates in its documentation like this: “Several times a year, Google makes significant, broad changes to our search algorithms and systems.” According to Google, these changes are broad in nature and don’t target specific sites or individual pages. Verbatim: “don’t target specific sites or individual web pages”.

Google explains this with an image. You have a list of your 20 favorite restaurants from 2019. There are new restaurants now. Some old ones you have reassessed. The list changes. According to Google, restaurants that move down aren’t necessarily bad: “restaurants that move down aren’t necessarily “bad””. Others have simply earned a place in the top 20.

For most sites, according to Google, this isn’t an issue: “most sites don’t need to worry about core updates and may not even realize one has happened.” The large core updates are announced on the list of ranking updates in the Search Status Dashboard. On top of that there are smaller core updates that, according to its own documentation, Google doesn’t announce because they aren’t widely noticeable.

In March 2024, the former helpful content system became part of the core systems. According to Google’s overview of ranking systems, it “became part of our core ranking systems”. The last separate helpful content update in the dashboard dates from September 2023. The core systems process many different ranking factors.

The table shows all core updates from 2023 with start and duration as listed in the Search Status Dashboard. The dashboard lists start and end with a time in US/Pacific. The days are given here.

The shortest update took just over 6 days, the longest 45 days. The median is 16 days. 2023 and 2024 had four core updates each, 2025 three and 2026 two so far.

For August 2026 the dashboard only lists the spam update from 18 August, no core update. Spam updates and core updates are different things.

In its documentation Google describes a check in a few steps. This is how I apply it:

1. Wait for completion: The Search Status Dashboard lists the start and duration of each update. While it is rolling out, fluctuations are to be expected.
2. Compare the right periods: Google recommends “waiting at least a full week after a core update completes before analyzing your site in Search Console”. Then you compare that week with a week before the start.
3. Review top pages and queries: In the Search Console performance report with the compare function, once by pages, once by queries.
4. Tell small from large: Google gives two examples. A drop from position 2 to 4 is small: “There’s no need to take drastic action”. A drop from position 4 to 29 is large and deserves a closer look.
5. Look at search types separately: According to Google, Web, Images, Videos and News can develop differently.

For each of the three updates, I compared the seven days before the start day with seven full days. The week after begins on the first full day that is at least seven days after the end. Start and end come from the dashboard in US/Pacific, and according to Google Help, Search Console also counts its days in Pacific Time. The data comes from the Search Console API (web search, property https://www.seo-kreativ.de/), retrieved on 24 September 2026.

Totals across the whole site say little. New posts were added in every period. The mix of queries changes as well. That’s why, as Google recommends, I looked at individual queries. I counted all queries with at least ten impressions in both weeks. I counted a drop as large if a query was on page 1 before and beyond position 20 after. This threshold is my own, based on Google’s example of 4 to 29. Precisely: an average position of at most 10.0 before and above 20.0 after. A change of at least one position counts as better or worse, anything smaller as stable. The classes are mutually exclusive: a large loss is not counted again as worse. I evaluated the query dimension across all countries and devices. The position is the average that Search Console reports per query across a week’s impressions, not a fixed position for all users, devices and locations.

One pattern stands out. Among the large losers were news queries themselves, such as “google core update 2025” and “march core update”. On the page level, after the March update the post on the March spam update had the largest click losses. After the May update, it was the news post on the March update, from 30 to 5 clicks per week. In my view, that is consistent with fading interest in these topics. But clicks can also drop at the same position, for example when demand falls. Whether demand, rankings or both are behind it cannot be separated with this data.

The measurement has limits. It covers a single small site with a few hundred URLs. The March spam update fell into the week before the March core update. It began on 24.03.2026 and ended after 19.5 hours. It may have affected that comparison week. The week after the December update falls shortly after the holidays. And the result depends on the window: shifting it by one day put the number of large drops for the May update at 0 or at 1. So what transfers is not the result but the approach. If your impressions drop without a core update running, my article on diagnosing collapsed impressions helps.

If the whole site sees large drops over a longer period, Google refers to its self-assessment questions. One of them reads: “Does the content provide original information, reporting, research, or analysis?” Google recommends staying as objective as possible: “Take a close look at your site as a whole, and try to be objective.” According to Google, an assessment by people who aren’t affiliated with the site can also help.

Next come the pages that lost the most. Google suggests comparing them with pages that do a better job of helping the searcher. What I avoid:

• Quick fixes: Google explicitly advises against “quick fix” changes, such as removing a page element because someone heard it was bad for SEO.
• Premature deletion: “Deleting content is a last resort.” If you want to delete entire sections, according to Google you probably created them for search engines rather than people.
• Touching good pages: For small drops, Google recommends not changing content that is already performing well.

For checking individual pages, I wrote a self-test for interchangeable content. For the whole inventory, a systematic content audit followed by pruning is the right tool. And for questions of trust and experience, E-E-A-T is worth a look.

Google writes that some changes take effect in a few days. But it could take “several months” for the systems to learn and confirm that the whole site is producing helpful content in the long term. If nothing has happened after a few months, that could mean waiting for the next core update.

According to Google, it isn’t quite that rigid. The smaller, unannounced core updates can also make improvements visible. And Google makes clear: “there’s no guarantee that changes you make to your website will result in noticeable impact in search results.” According to Google, positions aren’t fixed, because user expectations and the web keep changing.

Gary Illyes showed concrete values on October 2, 2026 at Search Central Live Deep Dive in Barcelona. On his slide about core updates, the range starts with the rollout of two to four weeks. Recovery typically takes three to six months. In the slowest case it takes six months to a year, until the next core update. Below the slide it says: “Times expressed on the diagram are estimations based on internal analysis.” Search Engine Roundtable published photos of it. I did not find the slide on Google’s own pages as of October 8, 2026. Illyes later called the slides an exercise in a comment on LinkedIn, to see whether the audience can relate to the numbers pulled internally.

I checked the two to four weeks of rollout against the history above. Of the 13 core updates in the history, six fell within this range. Six were shorter, three of them by less than a day. One took 45 days. So the slide is a rough guide and not a schedule. For recovery, in my view, that means: plan in quarters, not in weeks.

The made-up core update of March 2026

In March 2026 an experiment showed how quickly an update that doesn’t exist can emerge. According to Search Engine Journal on 18 March 2026, an SEO had noticed an AI hallucination about a “March 2026 Core Update” in his newsletter and, by his own account, published it as an experiment. He wrote: “My LinkedIn article began ranking on the first page of Google for “Google March update 2026.”” According to his account, Google’s AI Overview also picked up the fabricated information. According to the dashboard, the real March update only started on 27 March, nine days after the report.

A spam update is not a core update

According to Google, a spam update is a notable improvement to the automated systems against spam, such as SpamBrain. Sites that see a change afterwards should, according to Google, review their compliance with the spam policies. According to the dashboard, the three spam updates from March to August 2026 each took less than three days, the August one 2 days and 16 hours. Google announced the fourth spam update of 24.09.2026 with up to two weeks instead. It ended on 08.10.2026 after 13 days and 16 hours.

Fluctuation is not the same as an update

Tools such as the Semrush Sensor show daily fluctuations in search results. A high value doesn’t prove a core update. A drop in impressions alone can also have causes other than rankings.

A core update is not a verdict on individual pages but a reassessment of many results. According to Google, most sites may not even realize a core update has happened. If you see drops, you first need the right periods and a look at individual queries. For my own site, that showed four large shifts across three updates and many small ones in both directions.