I want to be honest about something a lot of SEO people are still dancing around: for informational queries, the click is dying, and it is not coming back. Between AI Overviews, AI Mode, and the assistants people use instead of Google entirely, I have watched top-three rankings on informational terms lose a large share of their historical click-through on properties I manage.
The wrong response is panic. The other wrong response is pretending nothing changed. This post is about the third option: reallocating effort based on what still pays.
The Fallout, Query Type by Query Type
Not all queries got hit equally. When I segmented my own portfolio and client accounts, a clear pattern emerged.
Informational queries: heavily absorbed
"What is," "how does," "why do" queries are increasingly answered in place. If your content strategy was built on definitional traffic, you are the one funding Google's answer box with nothing in return. This is the natural endpoint of the trend I wrote about in zero-click searches, just accelerated.
Commercial investigation: partially absorbed, still valuable
Comparison and "best X for Y" queries now trigger AI summaries, but users still click through to verify before spending money. The clicks that survive are higher intent than before, because casual browsers get satisfied by the summary. My conversion rates on surviving commercial traffic went up even as volume went down.
Navigational and brand queries: nearly untouched
People searching your name or your brand still click. This is why I keep saying that brand demand is the most defensible asset in search. If someone types "Pierre Subeh" or your company name, no AI summary satisfies them.
Local and transactional: resilient
Bookings, purchases, directions, appointments. The AI can describe a restaurant but it cannot seat you.
The Reallocation: Where My Hours Go Now
Here is roughly how I shifted content and SEO effort across my projects over the past eighteen months.
Less effort on: encyclopedic explainers, glossary content, broad beginner guides with no unique perspective. These are AI food now. I still maintain the ones that support my topical authority structure, but I stopped producing new ones for volume.
More effort on:
1. Experience-based content. Case narratives, teardowns, opinions with receipts. An AI Overview can summarize a definition. It cannot replicate "here is what happened when I actually did this." This is E-E-A-T as a survival strategy, not a compliance checkbox.
2. Brand demand generation. Podcasts, speaking, social, PR. The goal is people searching for me by name, because that traffic is AI-proof.
3. Owned audience. Email above all. Every AI-vulnerable page on my sites now works harder to convert a visitor into a subscriber before they leave.
4. Being the cited source. If the summary is the new page one, being named in the summary is the new ranking. That work looks different from classic on-page SEO, and I broke down my process in how I get cited by ChatGPT and Perplexity.
What I Tell Clients Who Want the Old Numbers Back
At X Network, the hardest conversations in the last two years were with clients whose dashboards showed declining organic clicks despite stable rankings. My position, delivered plainly:
- Traffic was always a proxy metric. Revenue, leads, and brand demand are the real ones. If leads held while clicks fell, the business is fine and the dashboard needs new metrics, not the strategy.
- Impressions in AI surfaces have value we cannot fully measure yet. A prospect who reads an AI summary that mentions your brand three times may convert weeks later through a branded search. Attribution was already broken. It is more broken now.
- The competitors losing their minds over this are cutting content budgets exactly when the bar for citable, experience-rich content went up. That is an opening, not a threat.
Tactical Adjustments That Actually Moved Numbers
Beyond strategy, a few concrete changes earned their keep:
Restructure for partial extraction
I write key sections so that if Google lifts them, the lift includes my name or brand. First-person framing does this naturally. "In my agency's experience running retention campaigns" travels with the quote.
Push comparison content up the funnel
Since commercial investigation clicks survived, I moved effort there. Honest comparison pages, pricing explainers, "who this is not for" sections. These pages convert the high-intent clicks that still happen.
Double down on formats AI cannot swallow
Tools, calculators, templates, original data from my own operations. An AI Overview can describe a checklist. It cannot be the downloadable, interactive thing. My technical SEO checklist style content keeps earning links and visits precisely because it is a usable artifact, not just prose.
Watch the query mix, not the total
I now report organic performance segmented by query intent. Total organic clicks as a single number hides everything important. Branded up, informational down, commercial flat is a healthy chart in 2026 even when the topline is negative.
The Uncomfortable Conclusion
The open web traded content for traffic with Google for twenty years. Google changed the terms unilaterally, and no amount of industry complaining will reverse it. The businesses that thrive now are the ones that treat search as one channel among several, own their audience directly, and build a brand strong enough that machines and humans both name it unprompted.
I did not enjoy watching those click curves bend. But the discipline it forced, less commodity content, more brand, more owned audience, has made every business I touch more durable. That is the fallout worth keeping.