When customers ask an AI which hotel brand to choose, they do not all get the same Marriott. In the aided awareness question, the models split. One model ranks Marriott at the top of the field. Another places it lower, behind competitors. The gap is not cosmetic — it changes which brand gets recommended.
Claude is the friendliest model to Marriott. GPT-5.6 is the harshest. The distance between them is the story of this report.
The ranking table below is the clearest view of the split. In aided awareness, Claude puts Marriott first. Gemini and GPT-5.6 both place Hyatt ahead of it. So a customer asking Claude hears Marriott as the leader; a customer asking the other two hears Hyatt first.
Note what the models agree on: Hyatt leads or ties for the lead in two of three models, IHG and Accor sit at the bottom across all three. The disagreement is concentrated at the top — exactly where a recommendation is made.
Swipe horizontally to view full table →
| Brand | claude-opus-4.8 | gemini-3.6-flash | gpt-5.6-terra | AVG |
|---|---|---|---|---|
| Hyatt | 2.3 | 1 | 1 | 1.4 |
| Marriott | 1 | 2.5 | 3 | 2.2 |
| Hilton | 2.8 | 2.5 | 2 | 2.4 |
| IHG | 4 | 4 | 4 | 4 |
| Accor | 5 | 5 | 5 | 5 |
The disagreement is sharpest in aided awareness, and it is a top-of-list disagreement. Claude reads Marriott's scale as the deciding strength. GPT-5.6 and Gemini treat Hyatt's consistency and loyalty program as more recommendable, pushing Marriott down.
The models' own reasoning points to why. Every model cites the same tension: Marriott's unmatched footprint and Bonvoy program versus property-to-property variability from its franchise-heavy model. The friendly read weights the footprint. The harsh read weights the variability.
“Marriott offers the broadest global footprint and diverse portfolio, but its scale and franchise-heavy model produce more property-to-property variability and less uniform elite recognition than smaller competitors.”
“Hyatt earns top guest satisfaction scores and an industry-leading loyalty program, with a smaller curated portfolio that translates into higher consistency.”
On attribute scores the models are tighter. Loyalty program and location are Marriott's strongest attributes across all three models; value is the weakest for every model. There is little room for disagreement here — the split lives in the ranking, not the ratings.
Personality and head-to-head data are pooled across all models, so they cannot show which model differs. Read them as consensus only. Pooled, the models cast Marriott as the Ruler archetype — ambitious, polished, authoritative — and Marriott wins most head-to-head dimensions against every competitor except on loyalty value, where Hyatt takes it.
The two models least favourable to Marriott are not confused about the brand. They rank it lower because they weight the same fact Marriott's marketing downplays: uneven quality across a franchise-heavy portfolio. Both cite property-to-property variability, and both hand the loyalty-value edge to Hyatt.
That is the actionable point. Marriott cannot out-scale its way past these two models — they already grant the footprint. What moves them is consistency and loyalty economics. Until property variability narrows in the sources these models cite, GPT-5.6 and Gemini users will keep hearing Hyatt first.