How Often Do Google's Top 10 Results Actually Change? We Re-Ran 1,460 Searches Up to Six Months Later

Google's top 10 changes all the time, but most of that change isn't caused by time passing. In our experiment, the same search repeated a few minutes later already showed about 3 different URLs in the top 10. Results from up to four months earlier differed by less than one more URL than that. The biggest jump came from results collected around Google's May 2026 core update.

If you track rankings, write content briefs from search results or compare SERPs over time, here is what that means for you.

What we did

We took 1,460 Google results pages that were collected between April and early October 2026, for keywords people researched in Seonio. On 9 October 2026 we searched each keyword again in the same country and language and compared the two top 10 lists of organic results.

• What counts as a change: a URL that was in the earlier top 10 and is no longer in the top 10.
• The baseline: we searched 50 keywords twice, about 8 minutes apart, to measure how much the top 10 moves without any time passing.
• The keywords: 91% were searched in the United States and 9% in Germany. They are keywords our users looked up, not a random sample of all searches.

About 3 of the top 10 change within minutes

Two searches for the same keyword, a few minutes apart, differed by 2.8 URLs on average. Counting positions instead of URLs, 4 to 5 of the 10 places showed a different result. That's normal behavior, not a ranking change. In the main comparison, the newest results, mostly one or two weeks old, matched this baseline, so about 3 URLs is a solid floor.

So if a page drops from position 4 to position 6 in a single check, you can't tell yet whether anything happened.

Age adds less than one URL over four months

The chart compares the earlier results with today's, grouped by how old the earlier results were. The dashed line is the minutes-apart baseline. Click a difficulty band in the legend to add its line.

Average number of top 10 URLs that changed, by age of the earlier results: 2.8, 3.1, 3.6, 3.5, 4.8 and 3.9 for 0–30 up to 151–180 days, against 2.8 for the same search minutes apart.

Up to four months, the earlier results differed by 0 to 0.8 URLs more than the baseline. Ranking changes do pile up over time, but slowly: a results page you analyzed last month is still mostly the page your readers see today.

The 121–150 day group is the exception, with about 2 extra URLs. Most of them were collected during a Google core update or in the two weeks after it.

Results collected during a core update changed most

Here the same comparison is grouped by the month the earlier results were collected.

Average number of top 10 URLs that changed by October, by month collected: April 3.8, May 4.4, June 4.7, July 3.7, August 3.0, September 3.2, October 2.3.

Google's May 2026 core update rolled out from 21 May to 2 June, according to the Google Search Status Dashboard. Results collected from late May to mid-June, during the rollout and the two weeks after it, changed most: about 4.8 URLs, 2 more than the baseline. Results from April, before the update, changed 3.8, about as much as July. From the second half of June on, change shrank with age: 3.3 to 3.9 in late June and July, about 3 in August and September, and the baseline in early October.

If age were the main cause, the oldest results from April would show the most change, but May and June changed more. The keywords researched each month also differ, so read the exact monthly numbers as approximate.

Harder keywords hold their top 10 longer

At every age, keywords with a difficulty of 60 or more changed less than keywords with a difficulty of 20–39. The middle band sat in between, except in the oldest group, where the two harder bands were level. In the newest group, hard keywords changed 2.2 URLs and keywords with a difficulty of 20–39 changed 3.1. Pages that rank for competitive keywords tend to stay there, while easier results pages leave more room for new pages.

What this means for your SEO work

• Don't react to a single rank check. A move of a couple of positions can happen within minutes. Compare several checks or a weekly average before you change a page.
• Reuse a recent SERP analysis. If you analyzed a results page in the last few months, most of it still applies. Check again before you rely on it for a big decision.
• Check again after a Google update. In our data, an update changed more than four months of normal movement, so re-check your important keywords once Google confirms one has finished, and treat rankings you see during a rollout as temporary.
• Expect more movement on easier keywords. They change more often, which also gives a new page a better chance to get in.

Limits of this experiment

Each keyword was compared once, at one point in time. Because every keyword was searched again on the same day, the age of a result and the month it was collected are the same thing in this data, so we can't fully separate them. Most searches were in the United States, so other markets may behave differently. Because the keywords came from Seonio research, very easy keywords (difficulty under 20) were too few to report separately.

To see today's top 10 for any keyword, run it through the SERP API or ask your AI assistant through Seonio MCP.