How search has changed with AI, and what still works
Search now happens across Google, AI answers, Reddit, YouTube, and other surfaces. Here’s what actually changed, what did not, and where SEO work needs to change with it.
4 min read
Search is no longer one results page.
People still search Google, but they also get answers directly from AI Overviews, ask ChatGPT and other AI tools, research products on Reddit, and use YouTube when video is a better answer.
That changes where discovery happens and what you can measure. It does not mean the fundamentals of search disappeared.
The useful question is not whether SEO is dead. It is which parts of the work need to change.
What has changed about search?
The biggest change is that discovery now happens across more surfaces, and not all of them produce a click.
Search engines answer more questions directly. AI Overviews and other search features can answer a query before someone visits a website.
Some searches now start in AI tools. When someone asks ChatGPT or another model a question, there may be no traditional results page or ranking position to measure.
Third-party sources can shape brand discovery. Reddit, review sites, publishers, forums, and other sources can influence what people find in Google and what gets referenced in AI answers.
Video is part of search discovery. For queries where demonstration or explanation matters, YouTube and video results can compete directly with traditional pages.
Your own analytics show less of the journey. A person can discover your brand, compare it with competitors, and form an opinion before ever visiting your site.
What still works in SEO and AI search?
The places people search have changed more than the basic job.
You still need to understand what people are trying to find, create something that answers that need better than the alternatives, make it accessible to the systems that retrieve it, and establish enough authority for it to be trusted.
That means much of the work still matters:
Understanding search intent
Technical crawlability and indexability
Clear site and page structure
Useful internal linking
Content that answers the actual question
Original expertise, evidence, and data
Authority beyond your own website
Testing whether changes improve performance
The distribution systems changed. The need to be the best available answer did not.
Is SEO dead because of AI search?
No. But some SEO strategies are becoming less valuable.
Ranking for an informational query matters less when the search result answers the question without generating a click. Producing large amounts of interchangeable informational content is harder to justify when an AI answer can summarize the same information immediately.
At the same time, people still need to discover products, compare options, evaluate companies, solve problems, and make decisions.
The job is still discovery. The surfaces where discovery happens are expanding.
That means SEO needs to account for Google results, AI answers, third-party sources, video, and the other places your audience actually researches.
What does this change about SEO strategy?
Stop treating organic discovery as a single-channel problem.
For an important topic, ask:
What does the traditional Google result look like?
Is there an AI Overview?
Which brands are mentioned in AI answers?
Which sources are cited?
Does Reddit rank?
Does YouTube appear?
Which third-party sites influence the decision?
Where does your brand appear across that journey?
Where are you absent?
You do not need a separate strategy for every surface.
Start with the questions and decisions that matter to your audience, then determine where you need to be present to influence them.
What SEO work should you stop doing?
Stop treating rankings as the entire outcome. A ranking matters only if it contributes to visibility, traffic, influence, or a business result.
Stop using word count as a measure of quality. Write enough to answer the question completely. More words do not make a page more useful.
Stop assuming your analytics capture all demand. Someone can discover or evaluate your brand in an AI answer without ever visiting your site.
Stop producing content because a keyword exists. Decide whether the query matters, what the searcher needs, and whether you have something worth adding.
Stop measuring every surface the same way. Rankings, mentions, citations, traffic, and conversions tell you different things.
What should SEO teams start doing differently?
Check whether your brand appears in AI answers. Test the questions that matter to your customers and record who gets mentioned and cited.
Look beyond your own website. Find the Reddit threads, publishers, review sites, videos, and other sources appearing for important searches.
Put important answers where people can find them. Do not bury the answer under several paragraphs of setup.
Create content around decisions, not just keywords. Comparisons, costs, tradeoffs, alternatives, use cases, and specific questions often matter more than another broad informational article.
Connect informational content to the next decision. If someone learns something useful from you, make the relevant product, service, comparison, or next step easy to find.
Test changes. Use control groups when possible so you can distinguish the impact of your work from normal search volatility.
Want to check yours?
Run the AI answer checker →
How should you prepare for the next change in search?
Do not build your strategy around predicting which search surface will win.
Build assets that remain useful as distribution changes:
Original research and data
First-hand expertise
Strong product and service information
Clear answers to important questions
Useful comparisons and decision support
Recognizable brands and experts
Content that can work across text, video, and other formats
Then watch where your audience is discovering information and adjust distribution accordingly.
The surface will change again. Your ability to answer real questions better than the alternatives is more durable.
What should you prioritize if you have limited SEO resources?
Start with the searches and questions closest to the decisions that matter to your business.
For those topics:
Check what currently ranks in Google.
Check whether AI answers mention you and which sources they cite.
Look for Reddit, YouTube, publishers, and other third-party sources influencing the results.
Compare what people need with the pages you already have.
Fix the most important gaps first.
Do not try to optimize for every possible query or platform.
Start where better discovery could change a customer decision.
What should SEO teams invest in most?
Invest in knowing whether your work actually changed the outcome.
Search now has more surfaces, more metrics, and more confident advice about what supposedly works. That makes measurement more important, not less.
Define what you expect a change to do before you make it. Use control groups when you can. Separate rankings, visibility, mentions, citations, traffic, and business outcomes instead of collapsing them into one measure of “SEO performance.”
The teams that can distinguish a real improvement from normal volatility can learn faster than the teams following every new tactic.
That advantage matters regardless of what search looks like next.
From this guide
Related tools
Keep reading
