AI Overviews Are Eating My Traffic. Here's My Actual Response
This is for anyone watching their organic search traffic slide and wondering if AI Overviews killing organic traffic is permanent or fixable. Short answer: it's mostly permanent for informational queries, and you need to stop treating it as an SEO problem. It's a distribution problem now, and distribution problems have more than one channel.
Decide this first
Decide whether the traffic you're losing was ever going to convert. That's it. That's the whole first decision.
On the DTC brand I scaled from $100K to $3M, we pulled our analytics and found that 70% of our "lost" organic sessions were informational queries like "how to clean cast iron" or "best fabric for humid climates." Those visitors read a paragraph and left. Revenue per session on that traffic was $0.11. Revenue per session on branded and transactional queries was $4.80.
AI Overviews eat the informational stuff first because that's what they're built to answer. If your traffic loss is concentrated there, you didn't lose customers. You lost pageviews that were never going to become customers. Pull your GA4 or your analytics tool, segment by query intent, and look at revenue per session before you panic. This one variable determines everything you do next.
What to look for
Which pages actually lost money, not just traffic
Run a query-level report for the last 6 months. Tag pages as informational, commercial, or transactional based on the query they rank for. Then look at conversion rate and revenue by tag. I've seen teams cut content budgets in half after finding that 80% of their traffic drop was on pages with a 0.3% conversion rate. Cut your losses there. Don't cut your losses on the pages converting at 4%.
Whether you're still cited inside the Overview
Being cited inside an AI Overview isn't nothing. It's brand exposure without a click. Check Search Console for pages with rising impressions and falling clicks. That pattern usually means Google is showing your content inside the Overview box. It's worse than a click but better than invisibility. Track it separately from your regular rankings, because it changes what "success" looks like for that page.
Your dependency ratio on organic search as a channel
Calculate what percentage of your revenue comes from organic search alone. On the infrastructure project I ran PR and marketing budget for, no single channel ever exceeded 25% of total awareness reach, because government contracts get killed by single points of failure. Apply the same logic here. If organic is more than 40% of your revenue, you have a concentration risk that AI Overviews just exposed. That's the real problem, not the algorithm update.
What to ignore
Ignore the "answer engine optimization" services popping up promising to get you cited in AI Overviews for a monthly retainer. Nobody has cracked a reliable playbook for this because Google changes what triggers an Overview constantly. Anyone selling certainty here is selling you a story.
Ignore schema markup as a silver bullet. Structured data helps Google understand your page. It does not guarantee inclusion in an Overview, and I've seen sites with perfect schema get zero citations while a competitor with messy markup gets cited constantly because their content directly answers the query in the first sentence.
Ignore aggregate traffic dashboards that just show a downward line. A line going down tells you nothing about why. You need the segmented data from the section above, not a vanity chart your agency puts in a monthly report to look busy.
Common mistakes
The biggest mistake I see is teams doubling down on more blog content, publishing faster, hoping volume wins against an AI summary that can synthesize ten competitor posts into one paragraph. Volume was already a losing strategy before Overviews. Now it's actively wasting budget. If your content team's answer to declining traffic is "let's publish twice as often," you're feeding the same machine that's eating you.
The second mistake is treating this as a technical SEO fix. I watched a team spend three months restructuring their site architecture, rewriting title tags, and chasing Core Web Vitals scores, thinking that would win back Overview visibility. Traffic didn't move. The content itself wasn't the problem. The query intent was the problem. No technical fix solves for "Google decided it can answer this without sending anyone anywhere."
The third mistake, and the one that costs the most money, is panicking into paid search without checking unit economics first. A founder I know shifted 60% of a $40K monthly marketing budget from content into Google Ads the week their traffic dropped. Their cost per acquisition went from effectively $0 (organic, sunk cost content) to $38 overnight. They hadn't recalculated their margins to absorb that. Three months later they cut the budget back, having burned $70K figuring out what a spreadsheet could have told them in an afternoon.
My actual response
Here's what I did, in order, when a client's informational traffic dropped 40% in six weeks after an Overview rollout:
- Killed the bottom 30% of content by revenue per session. No rewrite, no "optimize and see." If it wasn't converting before Overviews, it wasn't going to convert after. That freed up $6,000 a month in content production spend.
- Redirected that budget into email and owned audience building. We added exit-intent capture and a real reason to subscribe (a genuinely useful resource, not "10% off"). Email list grew 22% over four months. Email now drives 18% of revenue, up from 11%.
- Rewrote the top 20% of surviving pages to answer the question in the first two sentences, then go deeper. Not for AI citation. For humans who still click through and bounce because the answer isn't obvious fast. Time on page for those pages went up, bounce rate dropped 9 points.
- Built out comparison and "vs" content targeting commercial intent queries. These are the queries Overviews handle worst, because the answer is genuinely "it depends on your situation," which an AI summary can't resolve as cleanly as a human evaluating options can.
- Started tracking branded search volume as a leading indicator. If people are searching your brand name directly, Overviews can't take that away. It went up 15% over two quarters, which told us the brand awareness work elsewhere (podcast sponsorships, partnerships) was working even while generic search traffic stayed flat.
None of this got our organic traffic number back to where it was. That number isn't coming back for informational content. What came back was revenue, and revenue is the only number that pays anyone's salary.
FAQ
Is AI Overviews killing organic traffic a temporary problem or permanent?
Permanent for informational and definitional queries. Google has no incentive to send that traffic anywhere when it can answer directly and keep the user on its own page. Transactional and commercial queries with genuine ambiguity ("best running shoes for flat feet under $100") are more resistant, because the answer depends on the person, not just the facts.
Should I stop investing in SEO entirely?
No. Stop investing in SEO for queries an AI can fully answer in three sentences. Keep investing in SEO for queries where a real decision needs to get made, where your brand's specific offer, pricing, or point of view matters. That's still search traffic, and it still converts.
How do I know if my content is likely to get cited inside an Overview versus just replaced by one?
Check if your content answers the question directly in the opening sentences with specific numbers or steps. Content that does this tends to get cited. Content that requires three paragraphs of throat-clearing before the answer tends to get replaced entirely, because the AI has already synthesized a better opening than yours.
What's the fastest way to check if this is actually hurting my revenue?
Pull your last 6 months of search console and analytics data. Segment landing pages by intent. Compare revenue per session, not traffic, month over month. If revenue per session held steady or went up while total sessions dropped, you're fine. If revenue per session dropped too, you have a real problem and need to act on the pages driving that drop specifically.
The practical takeaway: stop watching the traffic line and start watching the revenue line. AI Overviews took the free, low-value traffic first. That was never your business. Go find out what your business actually depends on, and put your budget there instead of arguing with an algorithm you don't control.