For two decades, search advertising worked on a simple bargain. Someone typed a question, the engine returned a page of links, and advertisers paid to appear near the top of that page. The click was the unit of value, and almost every tool, report and budget line in performance marketing was built around it.
That bargain is being renegotiated. Google now places AI Overviews above its traditional results for many queries, chat assistants such as ChatGPT and Perplexity answer questions directly, and a growing share of people get what they need without visiting a website at all. Google has also started showing ads inside AI Overviews, which tells you the platforms themselves see this as the new front page of search.
None of this makes search advertising obsolete. It does change where attention sits, what a click is worth, and what it takes to be the brand an answer points to.
What actually changes
The biggest change is that the answer, not the results page, becomes the place where decisions start. When an assistant summarises the options for "best running shoes for flat feet", it is doing part of the comparison work a shopper once did across several tabs. The brands it mentions gain a kind of visibility that no bid can fully buy.
Three practical effects follow:
| Fewer, more decisive clicks. | When people do click through from an AI answer, they have often already narrowed their choice. Traffic may fall while intent rises. |
|---|---|
| Longer, more conversational queries. | People ask assistants full questions, with context. Keyword lists built around two- and three-word phrases miss much of this. |
| New ad placements with less history. | Ads inside AI answers are new formats, with their own rules and limited benchmarks. Early buyers learn faster, but they also pay for that learning. |
Being part of the answer
The industry has started calling this Generative Engine Optimisation (making your brand easy for AI systems to find, understand and cite). It overlaps heavily with good search practice, but the emphasis shifts.
AI systems draw on content they can read and trust. Clear product information, consistent details across your website, marketplaces and listings, and genuinely useful explanations all make it more likely that an assistant describes your brand accurately. Thin pages written only to rank tend to add little.
Reputation matters too. Assistants lean on reviews, forums and third-party coverage when forming a view. A brand that is discussed well in the places people trust has an advantage that paid media alone cannot create.
Rethinking the search budget
For most advertisers, the right response is not to pull money out of search but to stop treating it as one channel with one metric.
Measure intent, not only volume.
If click volume falls while conversion rates hold or improve, that may be a healthy outcome, not a failing campaign. Review search performance on revenue and qualified leads, and expect cost-per-click comparisons with previous years to become less meaningful.
Test the new formats early, with a clear budget.
Ads in AI answers are still settling. Set aside a defined test budget, agree in advance what you want to learn, and review the results on their own terms rather than against mature campaigns.
Connect search to the rest of the plan.
People now meet a brand in a video, hear about it in a community thread, and then ask an assistant to compare it with alternatives. Search increasingly captures demand that other channels created. Planning search in isolation undercounts what the rest of your media is doing for it.
Start by asking the assistants
Search is moving from a list of options to a conversation. The brands that do well will be the ones that are easy to understand, widely and accurately described, and present in the new ad formats while the rules are still being written.
The practical starting point is modest. Audit how AI assistants currently describe your brand and your category. Fix the inconsistencies you find in your own content. Then give your team permission and a budget to test, so that you learn how these placements behave before your competitors do.
