Halfway through the year is a good time to check whether your media plan still matches how people are actually finding you, and for most of the clients I talk to right now, it doesn't. The gap is showing up first and worst in organic search.
Seer Interactive tracked click-through rates on Google queries where an AI Overview appears and found organic CTR has collapsed by 61 percent on those queries. AI Overviews now show up on roughly a quarter of all Google searches. Read those two numbers together and the conclusion is not subtle: a meaningful share of the traffic your SEO program used to deliver is simply not arriving anymore, not because your rankings dropped, but because the person searching got their answer without clicking through to anyone.
Do the composite math and the planning implication gets concrete. If a quarter of your query landscape now carries an AI Overview, and those queries lose 61 percent of their clicks, the blended hit to organic traffic runs around fifteen percent, before accounting for the queries that migrated out of Google entirely into ChatGPT and Perplexity. Fifteen percent of a channel that most companies treat as free is a real budget number. For a site doing a million organic sessions a year, it is 150,000 visits that the H2 plan is silently assuming will arrive, and they will not, and no line item anywhere accounts for their absence.
I bring this up because almost every budget conversation I'm having this quarter still treats organic search as a stable, self-sustaining channel that doesn't need active investment beyond the usual content calendar. That assumption was reasonable in 2022. It is not reasonable now, and treating it as though it still holds is the single biggest blind spot I see in H2 planning this year.
The traffic didn't die. It moved upstream.
Here's the uncomfortable part. The traffic isn't gone, it's just harder to earn and different to measure. Someone still asked the question. They just got the answer synthesized inside the search results page, or inside ChatGPT, or Perplexity, rather than by clicking a blue link. Whether your brand shows up inside that synthesized answer is now a separate battle from whether you rank on page one, and I've written before about why GEO is a genuinely different discipline from SEO, not a rebrand of it. This is the practical consequence of that difference showing up in a client's traffic report.
And the decline is not evenly distributed, which is where the planning opportunity hides. Informational queries, the what-is and how-does and best-way-to questions, are the ones AI Overviews answer completely, and they are bleeding clicks fastest. Transactional and navigational queries, the ones typed by someone ready to act, still click through at close to historical rates, because an AI summary cannot complete a purchase or open your login page. Which means your organic channel is not shrinking uniformly. It is shedding its top-of-funnel volume while keeping most of its bottom-of-funnel value, and a traffic report that averages the two tells you neither story. Segment your own query data along that line before you touch the budget; the split will tell you how much of your loss is recoverable clicks versus permanently absorbed answers.
What to actually do with the H2 budget
What I'm telling clients: stop measuring organic search success purely on click volume and rankings, and start tracking share of voice inside AI-generated answers as its own line item, with its own budget. That likely means shifting some of the content budget that used to go toward chasing keyword rankings into building the kind of citable, specific, well-sourced content that gets pulled into AI answers in the first place. The pages that win citations are not the pages that won rankings; they are the ones with original numbers, named sources, and claims specific enough to quote. It also means accepting that some historically reliable organic traffic is not coming back, and paid channels need to pick up more of that demand in the meantime while your GEO presence builds. Budgeting for that honestly beats discovering it in the Q3 numbers.
One more measurement note, because it changes how the loss reads: the visitors who do still click through from AI-assisted search arrive better informed than any traffic your site has ever received. They read the synthesis, compared the options, and clicked anyway. Several clients are seeing conversion rates on that smaller stream run well above the old organic baseline. Fewer sessions, better sessions. A dashboard still graded on raw session volume will report catastrophe while revenue holds, which is its own kind of reporting failure, and a fixable one.
None of this is a reason to panic. It's a reason to update the plan. The teams quietly losing ground right now are the ones still reporting organic search health with the same dashboard they built three years ago, measuring a channel that has fundamentally changed how it delivers value. If your H2 planning hasn't touched that assumption yet, it's worth doing before you finalize the budget, not after Q3 numbers come in soft and nobody can explain why.