Google's November 2026 core update: how to measure it

Google's November 2026 core update started on Tuesday the 10th and may take up to two weeks to roll out, which means it ends right around the time retailers in Chile head into Black Friday week. I can already picture the December scene: someone opens Search Console, sees a line going down, pastes the screenshot in the team chat and writes "the update hit us".

Maybe. But at this time of year traffic drops on its own for most sites, and Google is reshuffling results at the same time. My take is that the before-and-after comparison Google recommends, done in this season, measures both things together. To separate them, you have to subtract the season using the same weeks from last year, take branded queries out of the math and read positions as clicks.

Here's what we know, why this update is harder to read than the previous ones and how I would measure it.

What did Google announce with the November 2026 core update?

The usual, in three lines. On November 10 at 10:06 a.m. (Pacific time), Google posted on its Search Status Dashboard: "Released the November 2026 core update, which applies globally and to all languages. The rollout may take up to 2 weeks to complete".

In other words, up to two weeks of movement in every country and every language, so Chile and Peru are in just like the United States. On LinkedIn, the Search Central account added that it is a regular update, designed to better surface relevant, satisfying content from all types of sites. There was no post on the official blog and no hints about what changed (as almost always).

What does stand out is the calendar. It is the third core update of the year and it arrives 173 days after the previous one, the longest gap of 2026, while spam updates kept coming closer together:

If the rollout uses the full window, it closes around the 24th. Black Friday in Chile falls on November 27. Three days of margin.

What is a core update and how is it different from a spam update?

A core update is a broad change to the systems Google uses to rank results. Its official documentation explains it with a pretty good analogy: it's like a list of the best restaurants that gets refreshed. New places show up, the assessment of the existing ones changes, and a restaurant that drops three spots didn't necessarily start cooking worse.

That's why it doesn't come with a penalty or a warning letter. It reorders: if you dropped, the most likely explanation is that other pages are now assessed as better answers for those searches.

A spam update works differently. It refreshes the systems that detect violations of the spam policies (scaled content without value, site reputation abuse, recycled domains), and if it hits you, there is usually a concrete practice to fix. I covered it in the analysis of the September update.

The difference matters because the fixes are different. After a spam update you clean up. After a core update you raise the relative quality of your content against what now ranks above you. For the wider context, there's the guide on how the Google algorithm works.

Why is this core update hard to measure?

Because it lands on top of the year-end season, and the season moves traffic more than many updates do.

Last year we measured what happens to year-end organic traffic across 178 company sites in the US, Latin America and other markets (the full study is here). At the median, November and December traffic was 16% lower than September and October, and only 22% of sites grew.

So the season pushes in two directions. Most sites (B2B, services, content) lose visits at year end, update or no update. And the ones that sell on Black Friday and Christmas get almost all of their seasonal demand right during the rollout weeks. In both cases, a naive comparison blames Google for something that is pure seasonality.

One piece of context: that sample is mostly US sites, and the season in Chile and Peru has its own events. Even so, the year-end drop also showed up in the 32 Latin American sites (-13% at the median).

How do you know if the core update affected you?

Google gives a reasonable method and I add three adjustments for this time of year. This is the sequence I would use:

1. Wait for it to finish. Check the Status Dashboard until Google marks the update as complete. During the rollout positions go up and down, and what you see on a Tuesday can undo itself by Thursday.

2. Build the two windows Google asks for. The documentation says to compare a week taken at least seven days after completion against the week before the start. If it closes on the 24th, that's December 1 to 7 against the week of November 3 to 9.

3. Apply the same windows to last year. This is the most important adjustment: compare those same dates for 2025 in Search Console. That window is clean, because the December 2025 core update only started on the 11th. Whatever changed last year between those two weeks is your season without an update. Also check that Black Friday falls in the same place within the windows (in 2025 it was on the 28th and this year it is on the 27th).

4. Separate branded queries. Filter the queries with your company name and look at the rest separately. In our branded traffic study those searches were 18% of the typical site's organic clicks (131 sites), with a 15.4% CTR against 0.5% for everything else.

5. Segment by pages, queries and search type. A drop concentrated in one group of pages (for example, the generic blog articles) tells a different story than an even drop across the whole site. Google also asks you to check web, image, video and news separately.

About step 4: searches with your name depend on how well known the company is, and at year end media campaigns move them. If you leave them in, they muddy the reading.

About step 5: Discover has its own report and doesn't go into this math. AI Overviews and AI Mode, on the other hand, are counted inside web search, so they are measured with the rest.

With that you can do a simple calculation. An illustrative example: if your non-branded clicks dropped 20% between this year's two windows and the same weeks in 2025 dropped 15%, the part you could attribute to the update is about 6% (0.80 divided by 0.85). It is still correlation and not causation, but it's a number you can discuss in a meeting without making anything up.

Two warnings about data sources. First: a rank tracker can show dramatic drops during the rollout that Search Console doesn't confirm, because it measures a sample of searches from one location on one specific day. I would use the rank tracker to see movement and Google's tool to make decisions.

Second: on November 12 Google published a guide to identifying Google Search traffic in your analytics. It explains that the referer header carries only the search engine's domain and that Discover uses the same values. It also clarifies that Google Analytics counts visits once the page loads, while Search Console counts clicks. If the two numbers don't match in December, that's normal and proves nothing.

Is a drop from position 2 to 4 a small one?

For Google, yes. Its documentation uses that example as a small drop that doesn't call for drastic changes, and a drop from position 4 to 29 as a large one that deserves a deeper review.

For your business, it depends on what that search is worth. In our CTR by position study (153 sites, non-branded searches, October 2026), the median CTR was 3.0% in second position and 1.3% in fourth. Dropping from 2 to 4 on an important query means losing more than half of its clicks.

Google is right that you shouldn't rewrite the page in a panic. But if that query is the one that brings in quote requests, it's worth looking at closely as soon as the update ends. That's why I would sort the analysis by lost clicks. A list of the 20 non-branded queries that lost the most clicks (with last year's control) says more than the site average.

What should you do if the core update dropped you?

First, confirm that the drop is sustained and survives the seasonal control. If it does, the work is about quality, and it's a lot less mysterious than it sounds.

Search your lost queries and look at who moved up. What those pages have and yours lacks (an original data point, a more complete explanation, a table that answers the question at once, an author who knows the subject) is the most honest diagnosis you will find.

Then run your most affected pages through Google's self-assessment questions, updated in October. The ones I use most:

• Does the content provide original information, research or analysis?
• Was it written or reviewed by someone with first-hand experience of the topic, the first E in the famous E-E-A-T?
• Does it have easily verified factual errors?
• Why does this page exist, beyond ranking?

With that you decide page by page: improve, merge with another one or delete. The documentation is explicit that deleting is the last resort, for what can't be salvaged. If you need to do it at scale, the process I use is in the guides on content audits and content pruning.

What not to do during the rollout?

The most tempting moves and the ones that do the most damage:

• Making big changes while the update is in progress (rewriting sections, changing URLs, deleting categories). If you change half the site midway, you'll never know what caused what.
• Hiring someone who promises to "recover you from the update" in 30 days. There is no fine to lift.
• Buying links or publishing content at scale to compensate. That puts you in line for the next spam update.
• Chasing whatever you think the algorithm rewarded this time. Nobody published what changed, and guessing is expensive.

And one that is more about reporting than SEO: don't send leadership a conclusion based on first-week data. Write down the dates, note which campaigns ran in parallel and wait for the clean window.

How long does recovery from a core update take?

Longer than anyone would like. According to Google, some changes show up within a few days, but confirming that the site as a whole improved can take several months, and sometimes the effect only arrives with the next core update.

There's a nuance worth knowing: there are also smaller core updates that aren't announced, and those can move positions in favor of sites that improved. So recovery doesn't have to wait for the next announcement, although there is no guarantee it will come either.

The practical move is to use that time to raise the quality of the pages that really matter, instead of sitting and waiting.

Frequently asked questions

The three I get most when there's an update: whether it's a penalty, where to see it on the Status Dashboard and how long it lasts.

Is a core update a penalty?

No. It reorders results and doesn't target specific sites. A manual action comes with a notice in Search Console, and a spam update targets specific practices.

Where can I see whether the core update has finished?

On the Search Status Dashboard (status.search.google.com), where Google announces the start and marks the end of every update. When it closes, it also updates its ranking release history.

How long does the November 2026 core update last?

Google gave a window of up to 14 days from the 10th. The two previous core updates of 2026 took about 12 days each.

What I would do this week

Before the rollout ends, open Search Console and export the four weeks you're going to need right now: the two from 2025 and this year's week before the start (the fourth gets added in December). Set up the branded filter and save the list of non-branded queries with the most clicks. It takes an hour, costs nothing, and in December you'll be able to answer with a number instead of a screenshot.

And if your site doesn't sell on Black Friday, tell whoever reviews your reports today that traffic will probably drop in December, update or no update. It's a lot easier to explain it before than to justify it after (the same goes for the clicks that don't arrive even when there are impressions).