Ahrefs’s AI search research points to an operating model change for B2B marketing, not a simple SEO adjustment. AI citation, search ranking, brand mention, third-party validation, and conversion pages now need separate owners, metrics, and budget decisions. Tim Soulo summarized the work as more than 1 billion data points across 14 studies in six months. That number should be read as a directional signal from one of the better data sets available to marketers right now, not as universal law. The pattern is consistent enough to matter: AI Overviews are reducing clicks to top-ranking informational content, schema markup did not create a meaningful citation lift in Ahrefs’ measured study, ChatGPT citations are not the same as Google rankings, YouTube and external mentions appear to matter, and “Best X” content can be influential for some queries while low-quality listicles remain a bad strategic bet. For B2B marketing leaders, this is a management story.
1. AI citation is not the same as SEO ranking
Traditional SEO trained teams to think in rankings, pages, impressions, clicks, and conversions. Those still matter, but AI search adds another layer: being retrieved, being cited, being mentioned, and being used as background context are different events. Soulo’s LinkedIn summary notes that ChatGPT cites about 50% of the URLs it retrieves. That means a page can influence the answer without receiving a visible citation. The same summary says 28.3% of ChatGPT’s most-cited pages had zero Google organic visibility. Ahrefs’s deeper citation analysis also found examples of pages with no traditional search visibility receiving significant ChatGPT citations. The business implication is that SEO ranking is no longer a complete proxy for AI visibility. A page can rank and lose clicks because the answer is resolved in the interface. A page can fail to rank in Google and still appear in ChatGPT’s cited set. A brand can be mentioned without a specific owned page getting the credit. Use rankings to understand search competitiveness, AI citation and mention tracking to understand answer influence, branded search to understand demand, and commercial-page conversion to understand whether buyers can still take action. This is why the opening piece in this series argued for query-intent reallocation. Informational content belongs in a citation-and-authority lane. Commercial and service content still belongs in a click-and-conversion lane.
2. Query intent determines the strategy
The Ahrefs AI Overview click data gets attention for a reason. Their 2026 update found that the presence of an AI Overview correlated with a 58% lower average click-through rate for the top-ranking page, up from 34.5% in the 2025 study. The more important executive detail is where AI Overviews appear. Ahrefs’s trigger research found that 99.9% of AI Overview keywords are informational. Tim Soulo’s summary also notes that transactional, navigational, and local searches remain far less exposed, with shopping triggering AI Overviews just 3.2% of the time in that synthesis. The strategy has to assign the right outcome by intent. Definitions, explainers, category POVs, and research claims should be built to be cited and remembered. Service pages, pricing pages, comparison pages, and case studies should still help buyers click, evaluate, and convert. Third-party surfaces should validate the brand where buyers and AI systems look for confirmation. The portfolio map for clicks vs citations covers that detail because this is where many teams waste money. They either keep funding old informational SEO as if clicks will behave the same, or they flatten every page into AI visibility work and weaken the pages that still need to convert.
3. Schema is hygiene, not the lever
The schema study removes a convenient fantasy. Ahrefs tracked 1,885 pages that added JSON-LD schema, compared them against 4,000 control pages, and found no major citation uplift across Google AI Overviews, AI Mode, or ChatGPT. AI Overviews declined 4.6% relative to controls, while AI Mode and ChatGPT moved 2.4% and 2.2%, changes Ahrefs described as indistinguishable from zero. Schema still helps machines understand a page and belongs in a clean technical foundation. The executive conclusion is that schema should not become the citation strategy. This distinction matters in budget meetings. Technical tasks are easier to approve because they look concrete. A team can show completed tickets, validation reports, and markup coverage. The harder work is improving the quality of the answer, strengthening the point of view, building external authority, and earning mentions from sources the model already trusts. The piece on schema markup and AI citations gives a reallocation map: keep schema hygiene, then move incremental effort toward single-question content, authorship, external mentions, YouTube, and cited-domain strategy.
4. External authority is part of the answer surface
Ahrefs’s ChatGPT citation analysis found that 67% of the top 1,000 cited pages came from sources marketers cannot directly influence. Wikipedia represented 29.7%, homepages 23.8%, and app stores 6.6%. Only 32.3% were categorized as influenceable content like educational pages, reviews, news, and blog posts. Marketers still have influence, but owned content is only one surface. Ahrefs’s “how to rank on ChatGPT” article points to third-party mentions and YouTube presence as meaningful signals. Tim Soulo’s summary cites YouTube mentions as having the highest correlation with AI brand visibility in their study, at 0.737. That number should be treated directionally, but it lines up with how buyers behave. They trust a company’s website more when other credible places describe the company in similar terms. This pushes AI visibility into work that often sits across teams: customer reviews, analyst-style content, podcasts, YouTube interviews, partner pages, credible roundups, community mentions, public case studies, and industry publications. A B2B company cannot ask the content team to “own AI visibility” if the most important sources are outside the blog.
5. The KPI cannot be one screenshot
Soulo’s summary notes that AI Mode and AI Overviews can reach the same conclusion for a query 86% of the time while citing different sources, with only 13.7% citation overlap. It also says AI Overviews change every 2.15 days on average, with 70% of content differing between consecutive observations while semantic similarity remains high at 0.95. The business interpretation is important: sources rotate faster than meaning. If a marketing team treats a single citation screenshot as the win, it will chase volatility. If the underlying answer keeps saying the same thing but the cited sources change, the durable goal is to influence the semantic territory: the category language, the buyer’s mental model, the repeated brand associations, and the external sources that reinforce those associations. That is why the AI Visibility Review proposes a monthly meeting: a meeting where leaders inspect priority queries, identify missing sources, assign owners, and make budget decisions.
Executive implications
- Organic traffic decline needs diagnosis by query intent, not a blanket panic response.
- Informational content should be funded for citation, mention, authority, and category framing.
- Commercial pages still need investment because clicks and conversions still matter near evaluation.
- Schema should stay in the technical SEO backlog, but not absorb the AI visibility budget.
- Third-party authority needs an owner because many AI-visible sources sit outside owned media.
- YouTube, podcasts, reviews, and credible roundups should be reviewed as visibility assets, not just awareness channels.
- AI citation screenshots should be treated as observations, not executive KPIs.
- The operating cadence should produce decisions about pages, sources, owners, and spend.
Operating model changes
| Operating area | Old habit | New requirement |
|---|---|---|
| Content planning | Build around keywords and volume | Build around query intent and desired outcome |
| SEO reporting | Rank, impressions, clicks, traffic | Add AI citations, mentions, answer accuracy, and source patterns |
| Technical SEO | Treat schema as a growth lever | Maintain schema as hygiene and fund substance separately |
| Content quality | Long guides and broad coverage | Clear answer assets, POV pages, research claims, and proof |
| Distribution | Publish on owned site and promote | Build credible external mentions and cited-domain presence |
| Executive review | Monthly dashboard readout | Monthly decision meeting with owners and next actions |
| Budgeting | Separate SEO, content, PR, and video | Allocate against visibility jobs across owned and external surfaces |
The practical work starts with rebuilding the portfolio around intent, keeping technical hygiene in place, earning mentions where AI systems and buyers look for validation, and reviewing the answer surface every month. AI search is making weak marketing operating models more visible. It forces the content program to stop treating traffic as influence, and it forces leadership to decide what kind of visibility is actually worth paying for. The funnel-signal review and the reporting workflow as Service-as-Software are the two adjacent operating moves that make this shift legible inside an existing marketing function.
Frequently asked questions
What did Ahrefs find about AI Overviews and B2B traffic?
Ahrefs found that AI Overviews correlated with a 58% lower average click-through rate for the top-ranking page in their 2026 update, up from 34.5% in their 2025 study. Their trigger research also found that 99.9% of AI Overview keywords are informational, leaving commercial and transactional searches far less exposed.
Does schema markup help with AI citations?
Ahrefs measured no meaningful citation lift from adding JSON-LD schema across Google AI Overviews, AI Mode, and ChatGPT in their study of 1,885 pages. Schema still belongs in technical SEO hygiene but should not be treated as the main AI visibility strategy.
What sources do AI systems cite most often?
Ahrefs’s ChatGPT citation analysis found that 67% of the top 1,000 cited pages came from sources marketers cannot directly influence: Wikipedia, homepages, and app stores. Only 32.3% were influenceable content such as educational pages, reviews, news, and blog posts.
How should B2B marketing leaders respond to the Ahrefs data?
Reallocate content investment by query intent (citation lane vs click-and-conversion lane), keep schema as hygiene but not as the strategy, build a presence on third-party sources that AI systems trust, and run a monthly AI Visibility Review that produces decisions about pages, sources, owners, and spend.
Next read: What to Review When AI Search Makes Funnel Signals Less Complete — the demand-side companion to this synthesis. To turn the operating-model shift into a structured engagement, see our methodology.
— Fernando González Aguirre, Founder, Structured Rebellion
