What is The AEO Index?

The AEO Index is a monthly report from Results AI on how AI engines — ChatGPT, Perplexity, and Gemini — recommend businesses. Each issue draws from the Results AI audit network, which runs the same buying-intent prompts every month and records which sources the AI cites, which industries are gaining or losing visibility, and how platform behavior is changing.

Where does the data come from?

Every figure in The AEO Index comes from audits run on the Results AI platform across paying clients and internal benchmark audits during the month of the issue. We run a fixed set of buying-intent prompts across ChatGPT, Perplexity, and Gemini, record which businesses are named and which sources are cited, and aggregate the results. Client identities are never disclosed without permission.

How can a business get into The AEO Index data set?

Any business that runs a Results AI audit becomes part of the anonymized data set for that month. A free AI Snapshot delivers results in about 5 minutes and contributes one anonymized data point to the next issue. A full monthly audit provides the deeper data used in trend sections.

How often does The AEO Index publish?

Once per month, mid-month, covering the prior reporting window. This cadence matches the natural pace at which AI engines update their underlying models and search indexes — fast enough to catch real shifts, slow enough to avoid daily noise.

The AEO Index — June 2026: What AI Engines Cited This Month

Welcome to The AEO Index

AI engines are now a real channel for how customers find local businesses, and yet almost no one publishes month-over-month data on what is actually happening inside that channel. The AEO Index closes that gap. Every issue draws from the Results AI audit network — the same buying-intent prompts run across ChatGPT, Perplexity, and Gemini every month — and surfaces the patterns that matter for businesses trying to get recommended.

This is Issue 1. Five sections, all grounded in audit data from the June 2026 reporting window: the sources AI engines cited most often this month, the industries that moved up and down, the platform behavior shifts we noticed, one anonymized client story, and one emerging prompt pattern worth watching.

Every figure below comes from audits run on the Results AI platform during the June 2026 reporting window. Client identities are anonymized unless permission has been granted. The data set will grow each month as more businesses run audits — the patterns are early, the direction is real.

1. Top Citation Sources This Month

When the three major AI engines name a local business or describe its services, they cite the underlying sources they are pulling from. Across all audits run in the June 2026 window, the most-cited source domains were consistent enough across industries to form a clear hierarchy.

The pattern: AI engines lean heavily on a small number of high-trust sources, and the gap between the top tier and everything else is significant.

June 2026 · Most-cited source types

Where AI engines pulled their answers from

Google Business Profile — cited in 71% of local-intent prompt responses across all three engines

The business’s own website — cited in 58% of responses where the business was named

Reddit threads and discussion forums — cited in 34% of responses, with Perplexity weighting these much more heavily than ChatGPT or Gemini

Industry directories and review aggregators — cited in 29% of responses (Yelp, TripAdvisor, BBB, niche industry directories)

News articles and local press — cited in 14% of responses, almost always for businesses with a recent newsworthy event

The practical read: a complete, accurate, regularly updated Google Business Profile is still the highest-leverage AEO investment any local business can make. It is the one source the AI engines reach for first, and the only one your business directly controls without depending on a third-party platform.

The second read: your own website still matters. In every audit this month where a business was named by an AI engine, the business’s own domain was cited as a supporting source more than half the time. The pages that got cited tended to be specific service pages and FAQ-style content, not the homepage.

2. Industries That Moved This Month

The Results AI audit network covers roughly two dozen industry verticals. Some moved up significantly in average Mention Rate this month, some moved down, and most stayed flat.

Up

Home services (HVAC, plumbing, electrical). Average Mention Rate across audited businesses rose from 28% in May to 36% in June. The driver appeared to be a wave of seasonal content — cooling, summer maintenance, emergency repair — that AI engines surfaced more aggressively as search interest spiked.
Accounting and bookkeeping firms. Up from 21% to 27%. We suspect this is a post-tax-season effect: less time-sensitive informational content, more advisory-tone content that AI engines tend to recommend.
Hotels and short-stay hospitality. Up from 24% to 30%. Summer travel intent driving higher AI engagement on hospitality prompts, especially on Gemini, which appears to weight Google Travel data heavily.

Down

Tax preparers (separate vertical from accounting firms). Down from 34% in May to 19% in June. Expected and seasonal — the tax preparer category is volatile and AI engines mirror search intent.
Real estate agents. Down from 26% to 22%. Possibly a market effect, possibly a content stagnation effect as agents who relied on listing-only content lose ground to agents publishing market commentary.

The honest caveat: a one-month change is not a trend. We expect movement on most of these to be partly noise. The reason to publish them anyway is that AEO movements are happening in the dark for most businesses, and even noisy monthly numbers are better than guessing.

3. Platform Behavior Shifts

The three engines we audit do not behave the same. June surfaced two notable platform-level shifts worth flagging.

Perplexity is leaning harder on Reddit

Across the June audit set, Perplexity cited Reddit threads in 47% of local-recommendation responses — up from roughly 38% in May. ChatGPT and Gemini cited Reddit far less often (under 20% combined). For businesses operating in categories where Reddit has active threads, a single popular thread can now meaningfully shift Perplexity Mention Rate. For businesses in categories where Reddit barely exists, this widens the gap between Perplexity performance and the other two engines.

Gemini is tightening on Google Business Profile completeness

Gemini’s recommendations this month were noticeably more conservative when Google Business Profile data was incomplete — missing hours, missing service area, no recent posts. Businesses with thin profiles dropped out of Gemini recommendations even when they appeared on ChatGPT and Perplexity. The signal: if you only fix one thing this month, fix your Google Business Profile.

4. One Anonymized Client Story

One client in the home services vertical — a multi-location HVAC company — ran their second monthly audit in June after starting AEO work in late May. The results illustrate how fast some signals move and how slowly others do.

The fast movers: their Google Business Profile work landed within the audit window. They added detailed service descriptions, posted updates twice a week, and asked recent customers for reviews. Their Mention Rate on Gemini went from 18% to 41% in a single cycle, almost entirely driven by Gemini’s sensitivity to GBP completeness flagged above.

The slow movers: their content work — four new service pages targeting specific buying-intent prompts — had not yet meaningfully moved ChatGPT or Perplexity Mention Rate. Content takes longer to be discovered, indexed, and integrated into recommendations. We expect to see the impact in the July or August audit cycle, not this one.

The takeaway is not the specific numbers — every business will move at its own pace. The takeaway is the asymmetry. GBP work pays off in one cycle. Content work pays off in two or three. A business doing both in parallel is making the right bet; a business doing only one is leaving the other gain on the table.

5. One Emerging Prompt Pattern

The most interesting prompt pattern we noticed in June is what we are calling the “trustworthy” qualifier. Across all three engines, prompts that include words like trustworthy, honest, reputable, or reliable produced meaningfully different recommendation lists than the same prompt without the qualifier.

Example: a prompt like “Best plumbers in Saskatoon” and the same prompt phrased as “Most trustworthy plumbers in Saskatoon” returned overlapping but distinct sets of businesses on every engine we tested. The qualifier shifted which sources the AI weighted — review depth, BBB rating, response to negative reviews, and length of business history all became more prominent.

The practical implication: there are now two related but separate AEO games. Performing in generic recommendation prompts is one. Performing in trust-qualified prompts is another, and the latter rewards a different content mix — long-tenure signals, response-to-review content, transparent pricing pages, and explicit credibility cues like accreditation or guarantees.

We will keep tracking this in subsequent issues to see whether the gap between generic and trust-qualified results widens or narrows.

How to Use The AEO Index

The Index is not a tactic guide — see our complete guide to AEO for the playbook, or what is AEO and why it matters for the foundational explanation. The Index is the dashboard above the playbook: month-over-month signal on what AI engines are actually doing right now, so you can prioritize the AEO work that is most likely to move the needle this cycle.

Three suggested uses:

Pick one platform shift per month to act on. This month, that is Gemini’s sensitivity to Google Business Profile completeness. Next month it will be something else.
Sanity-check your audit against industry context. If your industry moved down 8 points this month, your Mention Rate might be flat even though your work is paying off. Knowing the baseline matters.
Watch the emerging prompt patterns. Trust-qualified prompts are showing up more often. If your content does not signal trust explicitly, you are leaving recommendation surface area on the table.

Next Issue

The AEO Index publishes monthly. Issue 2 lands in mid-July with the same five sections, refreshed data, and trend lines that will start to form as the data set grows. The fastest way to be part of the next issue is to run an audit — the free AI Snapshot delivers in about 5 minutes and contributes one anonymized data point to the July report.

If there is a specific industry, platform, or prompt pattern you want covered in a future issue, send a note to [email protected]. The Index is most useful when it answers the questions our readers are actually asking.