AX is the new AEO

00 / the shift, in public data

Agents rely far less on training data

We put our buyer questions to earlier OpenAI generations and measured how much of each answer came from memory rather than retrieval. The last non-reasoning flagship, gpt-4.1, still built about half its answer from memory. Then reasoning and tool use became the default, and it falls off a cliff: down to 7–10% in the models agents run today.

Even the sources it cites are a moving target. Reddit's share of ChatGPT citations fell 86% in a week once the model began running site: searches. Any off-site channel can vanish in one update, so none is worth optimizing for. The page it keeps returning to is your own.

Note: models don't search on everything. Stable facts still come from memory. But pricing, setup, and comparison questions need current information, and in agent harnesses fetching is the default.

the experiment

How we designed the research

We ran the study across 1,000+ live business sites. Real buyer questions went to four agent harnesses spanning Claude and GPT models: claude-agent-sdk, claude-code, openclaw, and eve. That comes to ~38,000 agent journeys and 140,000+ page fetches from those sites, each journey run in its own isolated environment. Beforehand we scored, ranked, and tagged every site for agent-readiness and controlled for confounds like brand fame and industry. Full experimental design is in the preprint.

01 / trace replay · one ask, two journeys

Where the agent actually goes

Two real traces for the same buyer question, "what can I try for free?", on the same harness. twilio.com answers from its own docs. hashicorp.com blocks the agent twice and the journey ends on two competitors' blogs.

two real traces, claude-agent-sdk · claude-sonnet-4-6 · click a panel to expand

02 / recommendation

Agents recommend you twice as often when they can read you

We measure how strongly an agent recommends each business. An agent-ready site earns a clear, confident recommendation nearly twice as often as a blocked one.

The lift is steepest exactly where buying decisions are researched: 2.5× in IT infrastructure, 2.3× in sales & marketing, and up to 2.6× depending on which stack the buyer's agent runs on. The flip side holds too: when a site is not agent-ready, answers that both judges rate too weak to recommend at all are 2.5× more common.

03 / grounding

When agents can't read you, the answer is built without you

Across the finished answers, training knowledge stays at 7–10% either way, so agents are not falling back on training data. Whatever the site does not provide, web search fills in: on sites that are not agent-ready, almost half the answer (42%) comes from outside them.

04 / cost & performance

Blocked sites cost the agent's operator more per answer

On sites that are not agent-ready, runs take 23% more turns (5.5 → 6.8), get blocked by the site 2.1× as often, and take 15% longer to reach an answer (48s → 55s). The extra turns and retries are billed to whoever runs the agent.

Cost per run understates it. A 403 is cheap; the expensive part is that only 56% of runs on not-agent-ready sites end grounded in the business, against 78% on agent-ready sites. Spread the cost of the failed runs over the answers that did use your site and the price is +64% per grounded answer, averaged across all harnesses using claude and gpt.

05 / accuracy

The common failure is a missing fact

Every answer is graded fact by fact against ground truth captured from each business's own site: 30,633 asked facts across 2,426 answers and a sampled 131 of the 1,056 businesses. Wrong facts are rare in both conditions. The failure that grows when the agent can't read the site is an answer that contains none of the facts the user asked for. It is fluent and confident, so hallucination metrics and AEO dashboards do not flag it.

06 / what to do about it

The fix is a site agents can read.

AEO optimized what models knew about you from training data. AX means making your site readable to the agent that is already on it. Serve the answer in the initial HTML, don't block the agents your buyers send, and put pricing, trials, and setup on canonical pages an agent can fetch. Everything this study measured moves together once agents can read you: recommendation, grounding, cost, and accuracy.

We think the arbitrage is closing. Tactics that worked on the model's training knowledge (seeded threads, listicles on other sites, citation placement) lose leverage each time retrieval moves closer to the source, and the Reddit data in section 00 is what that looks like in practice. What is left is the basics: SEO so agents can find you, AX so agents can read you.

07 / conclusions

What we conclude

Five findings hold across the four Claude and ChatGPT harnesses and every business category we tested. Together they describe a market where training data no longer decides who gets recommended.

1. 01
   Training knowledge is no longer where the answer comes from.

   Agents now ground answers in the live web: frontier models search about nine times out of ten when a fact is fresh, and in our traces training knowledge supplies just 7 to 10% of the finished answer, readable site or not. Optimizing what the model knows about you optimizes a shrinking slice.
2. 02
   Readability is the largest single lever we measured.

   Agent-ready businesses are recommended 1.9× more often on average and up to 2.6× depending on the harness. When the agent can read the site, 78% of the finished answer is built from the business's own pages, against 58% when it cannot.
3. 03
   When the agent can't read you, it builds the answer without you.

   Blocked agents do not give up. They go to web search, whose share of the answer doubles from 12% to 25%. The answer built off-site is 3.7× more likely to contain none of the facts the buyer asked for, yet it reads fluent and confident, so nothing flags it. And accuracy follows the source: an answer grounded in your own pages is 1.4× as accurate.
4. 04
   Blocking agents is paid for by the agent's operator, then by you.

   On not-agent-ready sites, runs take 23% more turns, hit a block 2.1× as often, and each grounded answer costs the operator up to 93% more. Every harness is under pressure to route around sites that behave this way.
5. 05
   The AEO arbitrage is closing.

   Tactics aimed at the model's training knowledge (seeded threads, listicles on other sites, citation placement) lose leverage each time retrieval moves closer to the source. In August 2026, Reddit's share of ChatGPT citations fell 86% in a week. What remains is the basics: SEO so agents can find you, AX so agents can read you.

The full report publishes the traces behind every number.

08 / sources