AEO is easy to pitch as a trend and much harder to defend in a budget meeting. The useful question is not how many times ChatGPT mentioned you, but whether better AI visibility is moving through buyer consideration into qualified demand.
What is a defensible ROI case for AEO?
AEO ROI should be measured as a chain: technical readiness → AI visibility → brand inclusion → buyer consideration → qualified demand → revenue. Direct AI referral traffic is useful, but it is still too small and too incomplete to carry the business case on its own.
The channel is in an unusual stage. Ahrefs measured AI assistants at only 0.43% of traffic across a fixed set of 115,254 websites in September 2026. At the same time, Similarweb estimates AI platforms generated an average 770.7 million referral visits per month globally between June 2025 and May 2026, up 117.4% year over year. These are different datasets, but together they describe the same market: direct AI traffic is still small for many individual sites while the channel is growing very quickly overall.
-
0.43%
AI assistants' share of traffic across 115,254 sites in Ahrefs Web Analytics, September 2026.
-
+117.4%
Year-over-year growth in global AI referral visits in Similarweb's June 2025–May 2026 dataset.
-
2.5×
Higher likelihood of visiting a brand's site within seven days after receiving an AI recommendation, in Similarweb's downstream-impact study.
Why does website traffic understate the value of AI visibility?
Because AI can influence the shortlist before the buyer ever visits your website.
Traditional digital measurement was built around the click. A user searched, saw a result, clicked through and entered the analytics funnel. AI increasingly breaks that sequence. A buyer can ask ChatGPT, Gemini, Claude or Perplexity to explain a category, compare vendors, identify strengths and weaknesses and build a shortlist before visiting a company website.
Google is showing a similar effect inside conventional search. Pew Research found that users who encountered an AI summary clicked a traditional search result in 8% of visits, compared with 15% when no AI summary appeared. Clicks on sources cited inside the summary were rarer still.
That does not mean the answer had no commercial influence. Similarweb tracked real user journeys across Finance, Travel and Beauty and found that people who received an AI recommendation were 2.5 times more likely to visit that brand's website during the following seven days, even when there was no direct AI referral to attribute. Similarweb also found that much of this downstream traffic arrived through search rather than directly from the AI platform.
The measurement chain
How should AEO activity connect to business value?
Measure the stages in order. Each stage answers a different management question, and skipping stages usually produces bad attribution.
-
Technical readiness
Can AI systems reliably access, render, interpret and retrieve the information that matters? Track crawler access, server-side content availability, site architecture, schema accuracy, entity consistency and other deterministic readiness checks.
-
AI visibility
Are you appearing for commercial questions your buyers actually ask? Track mention rate, citation rate, share of voice, recommendation or win rate, and accuracy of the claims AI systems make about you.
-
Buyer influence
Is visibility moving you into the consideration set? Look for recommendation frequency, branded-search movement, direct visits, post-conversion survey responses and assisted journeys rather than relying only on direct referrals.
-
Qualified demand
Are the people influenced by AI becoming useful leads? Track AI referral sessions, MQL or SQL rate, assisted conversions, enquiry quality, demo requests and other intent signals appropriate to the business.
-
Commercial value
Are those leads turning into opportunities, bookings, pipeline, revenue and gross profit? This is where the financial ROI calculation belongs — after the upstream chain is measurable.
Before the budget request
Which AEO metrics should marketing leaders put on the dashboard?
Separate leading indicators from influence and business outcomes. If they are mixed into one score, you lose the ability to diagnose what is actually moving.
| Layer | What to measure | What it tells leadership |
|---|---|---|
| Readiness | Technical readiness score, crawler access, renderability, schema accuracy, entity consistency | Whether structural problems are preventing retrieval before marketing work begins |
| Visibility | Mention rate, citation rate, share of voice, win rate, accuracy rate | Whether the brand is entering relevant AI answers and how it compares with competitors |
| Influence | AI recommendations, branded-search lift, direct visits, self-reported discovery, assisted journeys | Whether visibility is affecting consideration even when there is no AI click |
| Demand | AI referral sessions, MQLs, SQLs, demo requests, assisted conversions, lead quality | Whether AI-influenced discovery is producing commercially useful traffic and leads |
| Revenue | Opportunities, pipeline, bookings, closed revenue, gross profit | Whether the channel is producing financial value and deserves more budget |
Evidence
What do real AEO programmes show when measurement reaches the business layer?
The useful cases are not the ones that stop at citations. They show a path from visibility into qualified demand or revenue.
HubSpot: citations → qualified leads → conversion
HubSpot's published AEO case study reports a 433% increase in citations, an 1,850% increase in qualified leads from AI, and AEO leads converting at three times the rate of other sources. The programme did not begin with a revenue dashboard. HubSpot first defined buyer prompts across awareness, consideration, evaluation and decision stages, then tracked product-level visibility and citations before changing owned content, partner content and community presence.
The underlying experiments are useful as well. HubSpot reports that an FAQ glossary increased citation share for related prompts by 60%, while product-page changes produced a 56% increase in AI citations.
Hertz Iceland: share of voice → referrals → bookings
A Semrush Enterprise customer story for Hertz Iceland reports that, over three months, ChatGPT share of voice increased 67%, AI Mode share of voice increased 70%, combined AI citations increased 59%, AI referral traffic increased 81%, and bookings attributed to AI sources increased 50%.
For a travel company, bookings are the outcome that matters. The value of the case is not that every company should expect a 50% booking lift; it is that the reporting followed the chain from visibility to an actual commercial transaction.
Blend: visibility → MQLs → pipeline
B2B agency Blend used its own business as an AEO proof of concept. Across more than 500 high-value commercial prompts, it reports total AI presence increasing from 17.1% to 34.7% over 90 days, AI-sourced MQLs increasing 133%, and more than $500,000 in AI-sourced pipeline. It also reports AI-sourced pipeline growing nearly 3.5× from the first half to the second half of 2025.
For B2B marketing leaders, this is probably the more relevant ROI shape: visibility first, then qualified inbound, then opportunity and pipeline. The sales cycle may mean the revenue result appears months after the visibility movement.
Direct AI traffic is still small on many sites
Share of tracked traffic across 115,254 websites, September 2026. Ahrefs Web Analytics.
Bars are scaled to search traffic. Different analytics methodologies will produce different channel shares; this chart reproduces the Ahrefs dataset only.
A common budget mistake
Why is AI referral traffic the wrong first ROI KPI?
Because the most important AI influence can occur before the click, while direct referrals remain a small part of most sites' traffic mix.
Ahrefs measured AI assistants at 0.43% of traffic across its fixed site set in September 2026. That is not an argument against AEO. It is an argument against using direct referral volume as the only success criterion.
At the same time, Semrush's 17-month clickstream analysis found outbound referral traffic from ChatGPT grew 206% from January 2025 to January 2026. Similarweb's global data also shows rapid growth. The sensible reading is that the channel is early, uneven and accelerating.
Does AI-referred traffic convert better?
Sometimes. The more useful question is whether it converts better for your category and audience.
There is a reasonable theory behind higher conversion: someone who clicks after an AI-assisted research session may arrive further into the decision process. But the effect is not universal, and industry context matters.
Adobe's September 2026 U.S. retail data is a striking example. Across a dataset covering more than one trillion visits to U.S. retail sites, Adobe reported that AI-referred traffic converted 90% better than non-AI traffic and generated 83% higher revenue per visit. AI-referred visitors also spent 73% more time on site and viewed 25% more pages per visit.
HubSpot's 2026 CRM buyer research points in the same direction from a B2B angle. In a survey of more than 3,000 CRM purchase decision-makers, 42% said they used AI search during evaluation, and those buyers were 36% more likely to purchase than buyers who did not. HubSpot also reports that customers actively optimising for AI search generated more AI traffic, MQLs and deals than comparable customers who were not, though these are HubSpot proprietary data rather than independent market benchmarks.
Finance
How should you calculate AEO ROI without pretending attribution is perfect?
Keep the standard financial formula, but separate what is directly attributable from what is influenced and what is still a leading signal.
ROI = (incremental gross profit attributable to AEO − AEO investment) ÷ AEO investment
- Directly attributable valueLeads, opportunities, bookings and revenue where an identifiable AI referral exists in the recorded journey.
- AI-influenced valueJourneys where AI appears earlier in the path, buyers self-report AI discovery, or AI recommendations correlate with subsequent branded/direct behaviour.
- Leading-indicator movementReadiness, mention rate, citation rate, share of voice, recommendation rate and accuracy. These show whether the mechanism is improving before revenue attribution matures.
Do not collapse all three into one artificially precise number. Finance should get a hard ROI where attribution genuinely exists; marketing should separately report influence and leading indicators that explain where future value is developing.
The budget gate
What should be known before a CMO asks for AEO budget?
A budget request becomes much easier to defend when it starts from an evidence-backed gap rather than a generic claim that AI search is growing.
-
Establish the technical baseline
Know whether AI crawlers can access the site, whether important content survives rendering, whether schema matches visible content, and whether obvious retrieval blockers exist.
-
Measure the commercial prompt set
Build prompts around actual buyer problems, category discovery, comparisons, evaluation and brand questions. Measure mention rate, citation rate, share of voice and accuracy against named competitors.
-
Map where the engines get their answers
Separate citations coming from your own domain from publishers, review sites, communities, social platforms and other authority sources. The remediation plan depends on that source mix.
-
Confirm the downstream measurement you already have
Identify whether analytics, CRM, self-reported attribution, branded-search data and assisted-conversion paths can pick up any influence today. Fix measurement gaps before declaring an ROI target.
-
Fund a measured test, then scale
Run the programme against a fixed baseline and prompt set. Scale investment only when readiness and visibility move first and downstream demand begins to follow.
A stronger budget statement sounds like this: “We are absent from most commercial AI conversations we care about, two competitors dominate the shortlist, our own domain is rarely cited, and we have identified specific technical and content gaps. Here is the 90-day plan, the baseline, and the business signals that will determine whether we scale.”
Is the market mature enough to justify AEO investment?
Mature enough to measure and test; not mature enough to abandon SEO or pretend the attribution problem is solved.
There is enough evidence now to treat AI-assisted discovery as a real part of the buyer journey. Similarweb's global referral dataset more than doubled year over year. Semrush measured ChatGPT referral growth of 206% from January 2025 to January 2026. HubSpot's State of AEO research says 58% of more than 4,000 marketers surveyed are already optimising content for answer engines.
There is also enough evidence to avoid overstatement. Direct AI referral traffic remains small for many sites. Attribution is incomplete. Engine behaviour changes quickly. A brand can be mentioned without being cited, cited without receiving a click, and influence a purchase that analytics later credits to another channel.
That argues for disciplined investment rather than either extreme. Ignoring the channel because AI traffic is currently small risks missing where buyer research is moving. Throwing a large budget at it purely because AI is fashionable is equally difficult to defend.
The practical test: are we becoming more visible during buyer research, are we entering more consideration sets, and is that increased consideration beginning to show up in qualified pipeline and revenue?
Common questions
AEO ROI — answered
What is the best leading indicator for AEO ROI?
There is no single best metric. Start with technical readiness, then measure mention rate, citation rate and share of voice across a stable set of commercial prompts. These tell you whether the mechanism is moving before enough downstream traffic or revenue exists for reliable financial attribution.
Should AI referral traffic be the primary AEO KPI?
No. Direct AI traffic is measurable and useful, but it misses the large class of journeys where an AI recommendation influences the buyer and the eventual visit arrives through branded search, direct traffic or another channel. Treat referrals as one performance metric, not the whole business case.
Can AEO revenue be attributed accurately?
Direct AI-referred revenue can usually be attributed with conventional analytics. AI-influenced revenue is harder because the discovery may happen without a click or on a different device. Use assisted journeys, CRM source data, post-conversion surveys and branded-search movement, and keep influenced value separate from directly attributable value.
How long should an AEO programme run before ROI is judged?
There is no universal period because site condition, engine retrieval behaviour, sales cycle and category competition differ. A practical first phase is a fixed-baseline test long enough to implement technical and content changes and re-measure visibility consistently. Revenue should be judged on the normal sales-cycle lag, not on the date a page was edited.
Does AEO replace SEO budget?
No. Search visibility, crawlability, site architecture, content quality and authority remain important inputs into AI retrieval as well as conventional search. AEO adds measurement and optimisation around AI mentions, citations, source mix, representation and buyer prompts; it does not remove the need for SEO.
What should a CMO ask an AEO vendor to report?
Ask for a fixed prompt set, repeated observations, competitor share of voice, mention and citation rates, source mix, technical remediation evidence, accuracy monitoring and a clear separation between direct AI referrals, assisted influence and business outcomes. Avoid reports built mainly from screenshots or one-off prompts.
Sources
Every figure on this page, with its origin
| Claim | Source | Date |
|---|---|---|
| AI assistants 0.43% of tracked traffic across 115,254 sites | Ahrefs — AI vs Search Traffic Analysis | Sep/Oct 2026 |
| 770.7M average monthly AI referrals; +117.4% YoY | Similarweb — AI Referral Traffic by Industry | 3 Sep 2026 |
| Search-result clicks 8% with AI summary vs 15% without | Pew Research Center | 22 Jul 2025 |
| AI-recommended users 2.5× more likely to visit brand within seven days | Similarweb — Downstream Impact of AI Visibility | 23 Jun 2026 |
| +433% citations; +1,850% qualified AI leads; 3× conversion; glossary +60%; product pages +56% | HubSpot AEO case study — first-party | 24 Aug 2026 |
| 42% of CRM buyers used AI search; users 36% more likely to purchase | HubSpot CRM buyer research — proprietary survey | 15 Sep 2026 |
| Hertz Iceland: +59% citations, +81% AI referrals, +50% AI bookings | Semrush Enterprise / Hertz Iceland — vendor case study | 2026 |
| Blend: presence 17.1% → 34.7%; +133% MQLs; $500K+ pipeline | Blend AEO proof of concept — first-party case study | 2026 |
| U.S. retail: AI traffic converted 90% better; revenue per visit +83% | Adobe Digital Insights | 29 Oct 2026 |
| ChatGPT outbound referrals +206% Jan 2025 to Jan 2026 | Semrush clickstream analysis | 7 Apr 2026 |
| 58% of 4K+ marketers say they optimise content for answer engines | HubSpot State of AEO 2026 — survey | 2026 |
