“Do you write texts tailor-made to AI Overviews?” It’s the query purchasers most frequently ask our content material studio, WordFactory.
“Sure, we do. Since 2016!” That is our utterly sincere reply. The purpose is that the principles of the sport have modified, however the demand for helpful, extremely detailed content material from trusted authors has solely grown stronger.
Let’s check out AIO, which may complicate enterprise promotion and, on the identical time, open up new alternatives for it. Kateryna Zolotaryova, founder and CEO of the Site24 advertising and marketing company will information us.
AI Overviews are opinions that Google generates independently primarily based on the evaluation of dozens of sources. Now, customers on the SERP get a ready-made, concise reply to their question on the prime of the checklist of hyperlinks. It’s not mandatory to go to web sites — that is how the “zero-click” phenomenon emerged. On the Google I/O 2025 convention, CEO Sundar Pichai confirmed that AI Overviews is not going to be an addition however a full-fledged a part of Google’s search ecosystem. Which means the modifications are irreversible, and you might want to put together for them not “later” however proper now.
This can be a actual revolution in info habits, which web site homeowners, entrepreneurs, and search engine marketing specialists are already experiencing. At WordFactory, we’re more and more being requested by purchasers whether or not to disregard AI Overviews or, conversely, adapt to them. One factor is definite: visitors is not a assure however a reward for accuracy, well-thought-out construction, and useful content material. To maintain up with the brand new actuality, you might want to perceive how this expertise works and the way it modifications the strategy to promotion.
Nonetheless, there may be some disagreeable information. Kateryna explains:
“The largest problem with optimizing for AI Overviews and different LLM programs continues to be the dearth of ordinary efficiency metrics. Conventional instruments like Google Analytics can solely present referral visitors from AI instruments, comparable to ChatGPT — and even then, it requires superior GA4 configurations to isolate that information.
We nonetheless cannot immediately monitor how typically an internet site is referenced inside AI Overviews or how incessantly it is talked about by massive language fashions. Whereas some instruments promise this, the outcomes are sometimes imprecise or unreliable.
So whereas the panorama continues to be forming, it is not utterly darkish — we now have instruments that supply a minimum of partial visibility into how AI programs understand and reference your model.”
Happily, we’re seeing new options emerge. For instance, Serpstat has already launched:
- AIO monitoring in Rank Tracker
- Key phrase analysis tailor-made for AI Overview
- A devoted LLM Model Monitor, which detects how your model is represented in AI-generated solutions.
AIO vs. search engine marketing: Analytics No Longer Works as It Used to
In basic search engine marketing, all the things was comparatively clear: positions, clicks, conversions — a large number of metrics that could possibly be measured, analyzed, and used as the premise for a technique. We knew which queries had been driving our development, the place the visitors was coming from, and which pages had been performing nicely. However with the appearance of AI Overviews, all the things has modified, and worst of all, the efficiency monitoring system has actually disappeared.
This can be a critical sign for companies. The interplay funnel breaks down proper at first: the person will get a solution with out visiting the location. Marketing campaign planning has develop into tougher: whether it is not possible to precisely measure which efforts have been efficient, it’s troublesome to optimize them. Nonetheless, as Kateryna assumes, Google and OpenAI will quickly launch their analytical instruments for monitoring outcomes.
Why Do Companies Have to Perceive AI Search?
Like all revolutionary expertise, AI Overviews is each a problem and a brand new development level. Sure, the necessity for clicks is falling: in keeping with numerous estimates, greater than 50% of queries at this time will not be accompanied by a transition to the location. This has had the best impression on info visitors. Kateryna shared the most recent statistics: “70% of informational visitors has now shifted to queries to ChatGPT and different LLMs.”
She can be assured that Google will actively develop AI Overview as a separate mode within the close to future — AI Mode alongside Google Maps, Photos, and so forth. It will likely be an interactive chat-based dialogue mannequin. And this may additional change how we work together with info on the Web.
We’re already observing such modifications:
- Being on the prime of search outcomes not ensures visibility. Even websites on the primary web page of SERP lose visitors if they do not make it into the AI Overview.
- The winners are matters with a excessive diploma of personalization. Tourism, HoReCa, tradition, historical past, schooling — something the place feelings, visible context, and engagement are vital.
- In case your website is included within the pool of sources for AI, it is a new peak of visibility. Our inside calculations present {that a} point out within the AI Overview can improve model belief by 1.5–2 occasions even with out clicking the hyperlink.
What Ought to You Do?
You shouldn’t simply “make it to the highest” however develop into a supply that AI makes use of to construct its solutions. To do that, you first want to grasp how neural networks work. There are two approaches: theoretical and sensible. Our professional, Kateryna Zolotaryova, is able to share her insights in each instructions.
Theoretical strategy
The place does AI overview get its info? Theoretical evaluation requires delving into the related Google patents, as that is at present the one publicly out there official supply of data. For instance, for those who learn this patent fastidiously, you’ll discover an fascinating element: in some circumstances, Google AI generates a response primarily based on an current data base, whereas in others, it performs further evaluation of paperwork within the index for particular queries and generates a brand new response. This response can then be added to the everlasting data base — that is how the system learns.
The conclusion is that the system capabilities not solely as a search engine AI but additionally as a studying system. That’s:
- For frequent and normal queries, a ready-made, confirmed data base is used.
- For brand spanking new, uncommon, or situational queries, Google AI analyzes contemporary content material, creates a response, and presumably saves it as a brand new “truth.”
Sensible strategy
Let’s transfer on to follow. Right here, our professional recommends listening to three major areas of labor.
1. Content material
The AI overview supplies recommendations on content material optimization. We enter our search question, for instance, “bus Kyiv — Warsaw,” and have a look at the particular content material within the AI overview. We complement our web page with detailed and correct info on all out there factors. It’s important to grasp that this output could differ for every person at totally different occasions, so it’s best to conduct this evaluation repeatedly, at various occasions and with totally different question formulations.
The AI overview is created and can develop as a conversational mannequin, so to get into these search outcomes for person queries, you might want to place high-quality content material on the location to have the ability to get into the Google database as a supply of useful thematic info.
Due to this fact, we return to the 2 common rules of search engine marketing content material: writing helpful info for readers and producing content material in an FAQ type. The latter doesn’t essentially should be in a “question-answer” format; it could merely be a high-quality article that solutions all potential questions on the subject for customers. Gemini can be utilized to generate such potential questions, as a result of Google AI overview works on its foundation.
Let’s take a look at an instance. The question is “Paris to London bus tickets.” The search outcomes present the next info:
Accordingly, it’s clear what we have to add to the content material on our web page:
- Particulars of on-line reserving (together with phrases and situations of buy, ticket refunds, and so forth.);
- Value and what it consists of;
- Reductions and situations for acquiring them;
- Journey time, schedule, and different route particulars.
One other technique is to make a request in conversational kind within the AI dialog field. For instance:
After analyzing the solutions, add all the required info to the web page.
On this case, Kateryna recommends utilizing Gemini as a result of Google makes use of this AI mannequin to generate responses. However you can too discover this difficulty utilizing ChatGPT or different programs and enhance the knowledge on this web page primarily based on them.
When inserting content material on the location, our professional adheres to the precept of distributing clusters throughout pages from basic search engine marketing. The design technique balances the “one intent — one web page” precept with an evaluation of Google’s prime outcomes to group associated question clusters onto shared pages. She additionally advises remembering clear H1–H3 headings, quick paragraphs, lists, and markers — all the things that each customers and AI can simply scan and perceive.
To seek out actual person questions and key phrase intent clusters, instruments like Serpstat’s Key phrase Analysis plus Clustering may help establish what searchers wish to know — even earlier than AI Overviews decide it up. This information can information your FAQ-style content material and guarantee it aligns with each person queries and search engine AI context.
2. Technical optimization
Based on Kateryna, correct technical search engine marketing is about sustaining an internet site’s “hygiene.” It have to be optimized and freed from errors, however the next parameters affect whether or not an internet site seems in AI search outcomes:
- No web site indexing points, no 3**, 4**, and 5** server responses, right sitemap, robots.txt, and meta robots tags.
- Website pace, Core Internet Vitals indicators, and file compression strategies.
- Structuring information utilizing Schema.org microformats and JSON-LD. All potential web site parts for which there’s a corresponding markup have to be marked up. And take into account that format libraries are consistently increasing.
- Cellular optimization and responsive design. Google AI Overviews are constructed into cell SERP, so the location have to be mobile-friendly and show shortly on any system.
3. Exterior hyperlinks
Right here, our professional factors out that it’s essential to separate hyperlinks and website status — each components affect the outcome. Hyperlinks are vital, however not directly. They impression the general authority of the area and, consequently, its inclusion within the AI Overview. The aforementioned Google patent states that the mannequin generates a response from sources chosen for his or her authority and thematic relevance. If we analyze the websites that seem within the AI Overview snippet, they’ve a excessive DR.
It is usually assumed that the location’s presence within the “Greatest X”-articles positively influences its inclusion to the AI search. Nonetheless, such rankings have to be revealed on authoritative websites and have natural visitors. In different phrases, the checklist have to be compiled and revealed in a manner that enables individuals to learn it and belief it. Faux rankings on weak web sites is not going to be useful.
Implications
We agree with Kateryna that AI programs don’t basically change the strategy to search engine marketing. They modify the best way info is obtained and the way customers work together. Nonetheless, the strategies stay largely the identical, with solely a transition to the subsequent degree of high quality. That is evolution, not a funeral.
Optimization for AI Overview doesn’t exclude the promotion to the highest of Google, because the AIO outcomes typically align fairly carefully with the highest 10 outcomes. It’s not possible to definitively decide if the connection is causal or because of shared components influencing each basic prime rankings and AI responses. Nonetheless, it’s affordable to imagine each mechanisms are most definitely concerned.
