Flight Bookings: AI Market Discovery Index

Benchmark Summary

AirHelp and Flightright share the top position in valid recommendation coverage for September 2026, each at 2.3%. Both figures represent a gain from the July 2026 baseline, when AirHelp led the category at 1.9% and Flightright registered 0.0%. That 2.3-point gain makes Flightright the largest riser across the series. The remaining five tracked brands — AirAdvisor, Compensair, Refund.me, ClaimFlights, and Skycop — recorded 0.0% valid recommendation coverage in July, August, and September 2026, and no tracked brand declined in valid recommendation coverage since the July 2026 baseline.

The move from August 2026 to September 2026 was a quiet month: every tracked brand is classified as stable on valid recommendation coverage, with no change large enough to register as significant. Within that stable month, Flightright's coverage has moved upward in each of the two months since the July 2026 baseline, extending its gain across the series, while AirHelp's cumulative gain since the baseline also remains positive.

The benchmark began with 261 prompt-surface observations in September 2026, producing 44 qualified observations after qualification.

Current Benchmark at a Glance

Qualified surface breadth measures how many of the six canonical AI/search surface families (ChatGPT, Copilot, Gemini, Perplexity, AI Overviews, and AI Mode) produced at least one qualified observation. All six families were represented in each month of the series. August 2026, the intermediate month, recorded 55 qualified observations.

For the strategic interpretation of this benchmark, read CiteWorks Studio's analysis of How AI Search Is Recommending Flight Bookings

Research Scope and Qualification

The public benchmark is narrower than the raw collection universe by design.

Brand-level percentages use the 107 qualified observations from July 2026 and the 44 from September 2026 as the public denominator, not the raw collection. AI Market Discovery Methodology

Current Brand Standings

This table is the primary current-month benchmark view for September 2026.

How to Read the Standings

• Presence rate: The share of qualified observations where the brand is mentioned, regardless of whether the mention is a recommendation.
• Valid recommendation coverage: The share of qualified observations where the brand receives a valid recommendation, the primary metric of the benchmark.
• Top-three rate: The share of qualified observations where the brand appears in the first three recommended positions.
• Rank-one rate: The share of qualified observations where the brand is the first recommendation.
• Net sentiment: The balance of positive over negative mentions (positive share minus negative share), ranging from -1.0 to 1.0.

For formulas and denominator rules, see AI Market Discovery Metric Definitions.

Recommendation Coverage Movement

This was a quiet month for recommendation coverage. No brand's movement was beyond normal month-to-month variation, and no brand was classified as a significant riser or decliner. Flightright has risen in each of the two months since July 2026, from 0.0% in July to 1.8% in August to 2.3% in September, on one valid recommendation in the current month (1 of 44 qualified observations). AirHelp returned to recommendation coverage in September 2026 after a month at 0.0% in August 2026, and its raw presence rate rose from 40.2% in July 2026 to 79.5% in September 2026. AirAdvisor's presence rate moved in the opposite direction, down from 57.9% in July 2026 to 29.5% in September 2026, while its recommendation coverage remained at 0.0% throughout.

Category Leader: AirHelp

AirHelp leads the category with 2.3% valid recommendation coverage in September 2026, up from 1.9% in July 2026, after dipping to 0.0% in August 2026. This is based on one valid recommendation within 44 qualified observations. AirHelp's presence rate rose from 40.2% in July 2026 to 79.5% in September 2026, and its net sentiment settled at 0.0 in September 2026, down slightly from 0.1 in July 2026.

Flightright matches AirHelp's 2.3% valid recommendation coverage but on a far smaller presence base of 4.5% in September 2026 versus AirHelp's 79.5%, and both brands recorded zero top-three and zero rank-one placements.

Skycop: Current Position

Skycop recorded 0.0% valid recommendation coverage and 0.0% presence rate in September 2026, unchanged from both July and August 2026. Skycop is the only tracked brand with no recorded presence at all in the current measurement month, alongside a 0.0 net sentiment reading that reflects an absence of measured mentions rather than negative sentiment.

AirHelp records 2.3% on valid recommendation coverage, while Skycop stands at 0.0%. Flightright achieves the same 2.3% coverage figure, compared with 0.0% for Skycop. The more important issue is what that gap means for Skycop's recommendation position: two competitors are now being recommended in roughly one out of every forty-four qualified AI responses in this category, and Skycop is being recommended in none of them, while also not yet registering measured presence in the qualified observation set. Leaving this unresolved for another measurement cycle means Skycop enters the next reporting period from the same zero baseline while AirHelp and Flightright continue to accumulate recommendation history.

Highest-priority diagnostic: Why does Skycop register 0.0% presence and 0.0% valid recommendation coverage across all six qualified AI/search surfaces in September 2026, while AirHelp and Flightright each reach 2.3% valid recommendation coverage in the same measurement period?

Recommendation Placement Snapshot

Coverage alone does not show how prominently a brand is recommended. Placement rates reveal where a brand appears in the recommendation order, not just whether it appears at all.

Neither AirHelp, Flightright, nor AirAdvisor recorded a top-three or rank-one placement in September 2026, despite AirHelp and Flightright both carrying recommendation coverage. Similar coverage levels can still hide different first-position rates, and in this month, no tracked brand achieved either a top-three or a rank-one placement.

Buyer-Intent Distribution

All qualified observations in September 2026 fell into the Brand Recommendation class, which covers discovery and consideration of a named compensation service.

The current public series measures brand-recommendation discovery only. It does not yet contain qualified observations in the Pricing & Value or Multi-Brand Comparison classes.

Historical Measurement Record

This is an evergreen benchmark URL. New measurements are added to the same report.

Evidence and Source Layer

The benchmark is built from prompt-level observations that retain the query, AI/search surface, answer, brand outcome, recommendation placement, sentiment, and, where exposed, citations or attributable evidence sources. Each observation carries the full context needed to understand how a brand surfaced in an AI response. Source presence is evidence about the information environment. It is not automatically proof that the source caused the recommendation.

Scope Boundaries

This public benchmark does not measure:

• Market share or sales attribution
• Every possible AI response a brand might appear in
• Organic-search ranking performance
• Social media mention volume
• Private or sponsored channels
• Causality from a metric movement alone

About This Benchmark

The LLM Authority Index AI Market Discovery Index tracks how brands appear in AI-generated answers across the major AI and search surfaces. This neutral, industry-level benchmark measures presence, recommendation coverage, placement, and sentiment to show how the AI information environment treats each tracked brand.

• AI Market Discovery Methodology
• AI Market Discovery Metric Definitions
• AI Market Discovery Research Standards
• Modeled AI Authority Value

Get a Company-Level Authority Index

The public industry benchmark shows category-level standings. A company-level Authority Index can go deeper.

Get my free AI Company Index