TL;DR
- AI Overviews and answer engines compress SERPs but do not end commercial search intent.
- Entities, structured answers and internal linking still move rankings; thin content and "generic informational" pages are the losers.
- Measurement has to switch from "position 3" to "incremental non-brand clicks" - and teams that cannot measure that will argue about vibes.
- The winning programmes are cross-functional: SEO, content, product and engineering working against one shared backlog.
What actually changed
Google's AI Overviews, Bing's Copilot surfaces and answer engines like Perplexity all compress the traditional SERP. For informational queries, the answer is increasingly above the ten blue links. This is a real structural change - but it is not new. Featured snippets, People Also Ask and knowledge panels have been eating top-of-SERP clicks for years. AI overviews accelerate the trend; they do not invent it.
The nuance matters. Commercial and navigational intent queries still route users to the blue links because they are deciding between options, not asking for a definition. The pain is concentrated in thin informational content that was never going to convert anyway.
Entity coverage beats keyword obsession
Answer engines and modern ranking systems both care about entities - the nouns your category is made of, and how they connect. A page that mentions the right twelve entities in the right context is more likely to be cited, retrieved and ranked than a page that repeats a target keyword seventeen times.
Practical move: take your top three commercial clusters, list the entities a domain expert would expect to see on any serious page about the topic, and grade your pages against that list. You will find missing entities more often than you find keyword gaps.
Structured answers are the new meta description
If you want to be quoted by AI surfaces, write pages that can be quoted. That means clean H2/H3 hierarchy, direct answers to direct questions, short declarative sentences near the top of each section, and tight paragraphs that survive being pulled into a bullet list. Schema (FAQ, HowTo, Article) still helps; more importantly, so does discipline in the first 80 words of every page.
Internal linking is the quiet winner
Two domains with similar content quality often separate on the strength of the internal link graph. Hub-and-spoke structures, descriptive anchor text, orphan-page cleanup and contextual in-body links are unglamorous - and they outperform another round of backlink buying. Npexpos's SEO engagements frequently allocate a whole phase to this workstream because the compounding is that good.
Measurement has to move
Position tracking in isolation is now misleading. "We rank position 3" can still mean a click-through collapse if AI overviews covered the query. The metrics that matter:
- Non-brand organic clicks on commercial clusters - segmented, not averaged.
- Impressions vs clicks by query type (informational, commercial, navigational) so you can see where compression is happening.
- Share of answer-engine citations on brand and category prompts - a new signal worth logging.
- Revenue per non-brand visit, to spot quality dilution early.
What we recommend for Q2
For most clients, the plan is boring and effective: audit entities on commercial clusters, restructure top pages for structured answering, invest in internal linking, and move reporting to incremental-click language before the next QBR. The loud work - new content velocity, more links - earns its spot after those four are in place.
Want this applied to your site?
Npexpos's SEO engagement runs the same audit, remediation and measurement rework in 8 phases, each with ticket-ready specs for your engineering team.