The rest resolve inside the result page itself — an AI Overview, a featured snippet, a knowledge panel, a People Also Ask stack — and the user moves on without ever visiting the source. That's the environment every brand is now marketing into.
The instinct is to fight it: rewrite pages to claw back clicks, or pull out of SEO because the funnel looks broken. Both reactions skip the real work. The search result has become the destination, and how a brand chooses to show up there — or whether to try — decides who gets remembered and who gets summarized away. Four decisions matter more than the rest.
Decide Whether a Click Is Still the Right Success Metric
The oldest habit in search marketing is counting sessions. It stops working the moment the answer box eats the query. If a page ranks first for a definition-style question and Google paraphrases the answer above the fold, traffic to that page can fall by half even as visibility rises. Report on clicks alone and the campaign looks like it's failing while it's doing exactly what you hired it to do.
Sorting out what to measure instead is the core of building a strategy for zero-click search — deciding which signals reflect performance once the click is no longer the default outcome. The trade-off is real. Impressions and citations are softer numbers, and finance teams don't fund soft numbers forever. A workable middle ground:
- Keep click tracking. You still need it for commercial and comparison queries, where the click is the intent.
- Add a visibility layer. Track share of voice inside AI Overviews, featured snippets, and answer engines — how often your brand gets named when the query is answered without you.
- Watch branded search. If people see your name in an answer and later search for you directly, that's the conversion. Count it as one.
Decide Which Pages Should Fight to Be Quoted
Not every page belongs in the answer box. Being cited inside an AI summary is powerful for awareness and terrible for direct response, because the reader gets what they came for and leaves. A Pew Research analysis of U.S. adults' browsing found that users click a traditional search result just 8% of the time when an AI summary appears, compared with 15% when one doesn't. That's the tax on being the source.
So pick sides, page by page. Educational content, glossary entries, and top-of-funnel explainers should be engineered to be extracted: clean definitions, structured lists, a direct answer in the first 40 words. Product pages, pricing, case studies, and lead-gen assets should do the opposite — stay rich enough that a summary can't replace the visit, and lean on brand and intent signals a paraphrase can't carry.
Trying to make one page do both usually loses on both. Sort your URLs up front by which ones are there to be quoted and which are there to be visited, and the rest of the plan gets easier.
Decide How Much to Engineer for the SERP Itself
Getting pulled into a SERP feature isn't luck. Google publishes the mechanics: schema types that make pages eligible for rich results, snippet controls, image and video markup, FAQ and HowTo structures. A brand that ignores structured data is asking to be summarized without attribution. A brand that uses it well shows up in the panel with its logo, its rating, and its name.
There's a ceiling on how far the technical work carries you. Schema qualifies a page for a feature; it doesn't win it. What wins the feature is the same thing that wins ranking generally — the page has to be the best available answer, from a source the engine trusts.
That trust comes from outside the page: editorial coverage, brand mentions, links from sites the crawler already respects. On-page structure and off-page authority have to be run as one project, not two.
Decide Where to Show Up When the SERP Won't Have You
Some queries are gone. If Google can answer "what time does the Super Bowl start" or "how many ounces in a cup" inside its own result, no amount of optimization brings that traffic back. Fighting for it is a sunk cost. The smarter move is to accept the loss and reallocate.
Two directions tend to pay:
- Deeper-intent queries. Comparisons, buying decisions, troubleshooting inside a specific stack — questions a two-sentence summary genuinely can't close. These still send clicks, and they convert better than the traffic being lost at the top.
- Off-search surfaces. Newsletters, podcasts, communities, YouTube, and the AI answer engines themselves. Being cited inside ChatGPT or Perplexity is the new version of ranking on page one, and it's driven largely by whether credible third-party sources talk about the brand at all.
The brands adapting fastest have stopped asking how to get the click back. They're asking where the attention went, and building for the surfaces that still reward being useful. The answer box didn't kill demand. It moved where you have to earn it.