Here’s who is privileged and who is rejected in the up-to-date era

AI assistants prescribe the rules when it comes to online visibility. These websites are selected on the basis of various criteria than conventional search rankings.

Instead of just looking at which pages are mentioned, I wanted to understand whether these references actually correspond to the popularity of the topics discussed.

I compared every domain Memory Share (as often appears) Participation (How often can you expect that it will appear on the basis of the volume of searching for these topics).

This comparison helps to discover prejudices and show whether a given system is based on some sources more or less than expected depending on the popularity of topics that the site includes.

I looked at the 50 best sites cited at the Brand Ahrefs radar for Google AI, Chatgpt and embarrassment reviews.

This is an AI ~ 76.7 m review, horses of ChatgPT by 957 thousand. And 953.5k hints of embarrassment for the month of 2025.

• Google AI review Resist tough on UGC pages such as Reddit, Quora and YouTube. Much more than the Google search engine. Wikipedia and health places, such as Mayo Clinic and Cleveland Clinic, are insufficiently represented. YouTube is the property of Google, so they can prefer it. Google has a license agreement with Reddit and has already given them great visibility in Google search engine.
• Chatgpt Underpressentents wikipedia and reuters.
• Embarrassment It is generally quite neutral.
• Wikipedia He was somewhat insufficiently represented in embarrassment, but slightly slightly represented in Google and Chatgpt. Despite the popularity of Wikipedia in AI assistants, it seems that these systems are quite biased.

Here is Nerdy.

Google seems to be about sites with the content generated by users (UGC) than trustworthy sites that are considered more EEAT.

Here are some definitions to keep in mind:

• Memory Share This is how often the domain was mentioned in all AI answers.

Mention Share = (Number of responses that mention the domain ÷ Total number of AI responses analyzed) × 100

• Participation The weight of memories according to Google search volume to estimate how much potential visibility the site gets on the topics of a high and low demand.

Impression Share = (Sum of search volume for queries where the domain is mentioned ÷ Total search volume of all AI-analyzed queries) × 100

The difference between the mention of the action and impressions informs whether the site is cited more in queries of high visibility or low visibility. It reveals system prejudices in how AI assistants show different websites.

Excessive religion:

• Reddit (7.4% vs 4.0%): mentioned 3.4% more than expected visibility
• Quer (3.6% vs 1.4%): Mentioned by 2.2% more than the expected visibility
• YouTube (9.8% vs 7.8%): Mentioned by 2.0% more than the expected visibility

Insufficient:

• Wikipedia (8.4% vs 11.6%): mentioned 3.2% less than the expected visibility
• Mayo Clinic (2.9% vs 4.5%): 1.6% less than expected visibility was mentioned
• Cleveland Clinic (2.9% vs 4.5%): 1.6% less than expected visibility was mentioned

You can apply Ahrefs radar to see when these prejudices could be introduced.

For example, here are the 5 best sites in AI reviews in time, but burdened on the market. It seems that Google is moving away from Wikipedia, strongly heading towards Reddit in December and March, towards Google’s translation in March, towards Quora in December and March, but away from the quora in April, and maybe he left YouTube in October, but he restored it in April.

In this way you see the algorithm and trends updates in the up-to-date era.

Chatgpt was a system that in my opinion would have the most radical differences. In general, they seem not to represent Wikipedia and information services.

One more definition:

• Potential participation in the range (%) estimates how much exposure to the domain could Complete the answers of AI assistants. There is no reliable volume of searching for AI assistants, so the weight in Google search engines and acts as proxy based on the historical popularity of the network.

Potential Reach Share = (Sum of search volume for prompts where the domain is mentioned ÷ Total search volume for all prompts analyzed) × 100

Excessive religion:

• Google.com (2.3% vs 1.7%): 0.6% above
• Apple.com (4.0% vs 3.8%): 0.2% above

Insufficient:

• Wikipedia (16.3% vs 19.3%): 3.0% below
• Reuters (4.3% vs 6.4%): 2.1% below

I expected more embarrassment towards Wikipedia. The general director commented on Wikipedia’s bias, and even offered the support of anyone who wanted to build an alternative. I couldn’t be wrong.

Trouble seems to be the most balanced, showing references, which roughly comply with the popularity of the subject in the conventional network.

Excessive religion:

• YouTube (16.1% vs 15.7%): 0.4% above
• Apple.com (3.5% vs 2.6%): 0.9% above

Insufficient:

• Wikipedia (12.5% ​​vs 13.4%): 0.9% below
• Tasaude.com (1.6% vs 2.5%): 0.9% below

Final thoughts

The search is fragmented. We are so used to optimization for Google. Now you may need to look at optimization of various systems with different prejudices.

If you have questions, keep asking me LinkedIn Or X.