This article is a companion to our search engine optimization SEO content. The Google core update of October 2025 has been released and there was a great deal of ranking change with it. In the October 2025 update Google carries on developing a system that moves more cleverly inside the logic of search ranking.

Since 1999 we have watched Google's updates, the ones it named like Panda and Penguin and the core updates it left unnamed, and they have generally produced better results. We now see that Google's search results behave rather like an A/B test, running tests continuously inside themselves in order to present the most useful site to the user. Beyond the A/B test between sites, we feel that A/B testing in the logic of search ranking too.

There are words and phrases that certain companies have become identified with, or that they own. Those are generally untouched by any update. But on searches such as the name of a sector or a product category, meaning the searches where there is competition, we can say that positions are no longer very stable and nobody's position is guaranteed. The importance of sustainable work in SEO has come out once again.

Core Updates, A/B Tests and Changes on the Site

These three ideas are not really separate processes, they are connected parts of a single system.

  1. The core update (the change in the main rules): this is Google recalibrating its definition of “the best result”. The algorithm updates its rules on which site will satisfy the user more. It might decide, for instance, to give more weight to the “E-E-A-T” signals of experience and expertise.
  2. The A/B test (applying the rule and confirming it): this is the live testing mechanism Google uses to find out whether its new rules really work. What you observe as Google “giving a lot of sites a chance” is exactly what happens here. For a query it takes the site in 3rd place up to 1st place and watches how users react, through time on site, bounce rate and the like. If the site that was 3rd delivers better user satisfaction in the A/B test than the site that was 1st, it can win that position for good. Rankings “not being stable” and “nobody's place being guaranteed” are the direct result of that constant testing.
  3. The effect of changes on the site (entering the test again): when you carry out “a broad edit” on your site, you have given Google the signal “test me again against the new rules”. Google then puts your site back into that A/B test.
    • If your changes fit the new definition of quality that the core update is looking for, and users are satisfied with the result, your site passes the A/B test and your ranking rises.
    • If your changes do not satisfy the user, it loses the test and you live through a “loss of ranking”.

AI Content in October 2025

Reaching information, or reaching conditional and personalized information, is very easy now. The AI models themselves now hold 40 years of internet data, produced across the whole internet in every language it was written in. Writing content out of general knowledge does not mean what it used to. You can produce content with AI support, but there are things you have to add, and places in the skeleton where you have to leave your own mark. The most important point, as in the updates before this one, is E-E-A-T.

On every piece of content you should ask these questions:

  • Who is writing this article, and what is their expertise in what they wrote?
  • What does this article add to the database of humanity that was not there before?
  • Which parts of this content are current?
  • Which data can someone reading this content reach that they could not have obtained from an AI?

E-E-A-T and the October 2025 Updates

In this update we can say that the whole approach to E-E-A-T is now more targeted and more clever. According to our own analysis as Türk SEM, we can say that some changes and improvements were made in the logic of E-E-A-T scoring. You may well have taken a qualification in your field, run superb training courses, or your company may be a firm with a great deal of experience. What changed here is the distribution between the E-E-A-T scores. We think the weight of real world cases increased in the October 2025 update. This information, of course, holds our own personal experience as Türk SEM. So the distribution may not be quite like that. But one thing is a firm piece of data: you have to be sure you have explained yourself correctly and clearly on E-E-A-T.

Spam and Manipulative Tactics in the Crosshairs

20 years ago, when I was an editor inside the ODP (the Open Directory Project), we would spend a great deal of time fighting spam. With the human eye we analyzed whether the sites applying had got mixed up in spam, and we tried to work out whether the sites we had added earlier had got mixed up in it since. The logic of spam back then was primitive and rather funny. White text written on a white background. Poor quality automatically generated content, and at its simplest that was what spam looked like.

Now content can be produced with AI, and that content can be “QUALITY” content. It has become quite hard for a person to tell. But Google is in a position to tell better than a person can. Whether a piece of content was written for Googlebot or for humanity is now something it reads more strongly, with the October 2025 update. Articles are being categorized more clearly now, and their effects separated out with labels.

Local SEO Changes in October

Another thing that changed in October 2025 was the ranking logic in local searches. With this update, reviews are now analyzed by Googlebot and their effect on value has become sharper. Businesses whose name, address and phone number are consistent are assessed as more successful. Changing your location or your phone number often can have a negative effect on your Google Business Profile.