{"@context":"https://schema.org","@type":"BlogPosting","headline":"AI Hotel Volatility 2026: The Top 10 Rewrites Itself Weekly","description":"An 11-week time series (2026-05-11 to 2026-07-20) of AI hotel recommendations: the same 616-prompt library fired weekly at ChatGPT, Gemini, Perplexity, Copilot and Google AI Mode, hotels entity-resolved to canonical IDs and ranked per engine per week by mention count (top-500 pool). Week-over-week, only 29.2% of a top-10 and 49.7% of a top-100 survives, and the median surviving hotel still moves 55–80 ranks. Cohort survival separates three churn personalities: Gemini is noise around a stable core (60% of its May 11 top-100 alive ten weeks later; 42 of 60 survivors left and returned), ChatGPT is compounding drift (24%), Perplexity near-lottery (19%, zero all-season hotels, 10.5% newcomer rate). Pool membership is far stickier than position: top-10 hotels remain in the 500-deep list 94–99% of the time on four of five engines. Only 29 distinct hotels held a top-100 seat all 11 weeks (41 hotel–engine seats, 28 chain-branded); four did it on three engines at once. Chains outlast independents everywhere except Perplexity. Grok excluded (absent from the data since 2026-05-25, scraper suspended) — disclosed, not imputed.","datePublished":"2026-07-22","dateModified":"2026-07-22","url":"https://nicolassitter.com/research/ai-hotel-volatility-2026","category":"research","keywords":["AI hotel volatility","AI hotel rankings stability","ChatGPT hotel recommendations change","AI search volatility","Perplexity hotel recommendations","AI visibility tracking","AI rank tracking hotels","week over week AI rankings"],"articleSection":"Research","wordCount":2200,"readTime":"9 min","articleBody":"July 202611-week time series\n\n# AI Hotel Volatility 2026:the top 10 rewrites itself every week\n\nSince May, the [AI Hotel Landscape](/projects/ai-hotel-landscape) pipeline has asked five AI engines the same 616 hotel questions every Monday. That gives us something a snapshot study can’t: the ability to hold the questions still and watch the **answers move**. Eleven consecutive weeks, one prompt library, five engines — how much of an AI’s hotel ranking is still there seven days later?\n\n**TL;DR.** Averaged across the five engines, only **29.2%** of a weekly top-10 survives into the next week — roughly **7 of 10 top picks are replaced every Monday**. Depth of churn is an engine personality: Gemini’s week-one cohort is still 60% intact ten weeks later, ChatGPT’s erodes to 24%, and Perplexity kept **zero** hotels in its top 100 for all eleven weeks. Meanwhile the pool itself is calm: a top-10 hotel almost never falls off the 500-deep list — it just loses its slot.\n\n### Summarize with AI\n\n70.8%\n\nof the top 10 replaced weekly\n\navg across 5 engines\n\n11\n\nconsecutive weeks\n\nMay 11 – Jul 20, 2026\n\n0\n\nPerplexity anchors\n\nhotels in its top 100 all 11 weeks\n\n29\n\nanchor hotels\n\nheld a top-100 seat all season, any engine\n\nBack in February, the [rankings-consistency study](/research/ai-hotel-rankings-consistency-study-2026) re-ran identical queries minutes apart and found 50.5% position-1 stability — ask twice, get a different winner half the time. The open question was whether that shakiness averages out once you aggregate hundreds of prompts into a weekly leaderboard. It does not. Even at the aggregate level — 616 prompts, thousands of captures per engine per week — the composite top-10 keeps just 29.2% of its members from one Monday to the next, and the top-100 keeps 49.7%.\n\nThe interesting part is _how_ each engine churns. Some jitter around a stable core, one drifts steadily away from where it started, and one behaves close to a weekly lottery. The label on the churn matters more than its size: jitter is noise you can average away; drift means the model’s picture of the market is actually moving under you.\n\n## 1\\. The setup\n\nEvery Monday the landscape pipeline fires the same 616-prompt library (56 destinations × 11 templates) at ChatGPT, Gemini, Perplexity, Copilot and Google AI Mode, then resolves every recommended hotel to a canonical hotel ID. Ranking hotels by weekly mention count per engine yields a 500-deep leaderboard per engine per week. This study compares those leaderboards across the 11 consecutive weeks from May 11 to Jul 20, 2026 — the full run of the current prompt library, which went live on May 11 (earlier weeks used a smaller library and are excluded, because a prompt-set change reshuffles the hotel universe all by itself).\n\n-   **Identity across weeks** is the canonical hotel ID, so a renamed listing or alternate spelling doesn’t count as churn.\n-   **Same questions, same cadence.** With the prompt side held fixed, any movement comes from the engines’ answers themselves.\n-   **Grok is absent:** its scraper broke in late May and the platform has been suspended from the panel since; it only has two weeks inside the window and is excluded rather than imputed.\n-   **Depth guardrail:** at rank 100 a hotel has 3–5 mentions in a week, so tie-order noise is real deep in the list. Headline numbers therefore use the top 10 and top 100, and section 4 shows how churn scales with depth.\n\n## 2\\. The weekly rewrite\n\nTake an engine’s top-10 on any Monday and check the following Monday: on average only 29.2% of it is still there. ChatGPT is the steadiest at the very top (34%), Perplexity the wildest (17% — in a typical week it keeps fewer than two of its ten). Widen to the top 100 and the spread between engines opens up: Google AI Mode retains 57.9% week over week, Perplexity just 28.4%.\n\nai-hotel-volatility-weekly-retention-2026\n\nWeek-over-week stability by engine, averaged over the 10 week-pairs. 'Median rank move' = median |Δrank| for hotels present in consecutive top-500s. 'Newcomers' = share of each week's top-100 never seen in any earlier week's top-500.\n\nEngine\n\nTop-10 kept\n\nTop-100 kept\n\nMedian rank move\n\nAnchors (11/11 wks)\n\nWk-0 cohort alive @ wk 10\n\nNever-seen newcomers\n\nChatGPT\n\n34%\n\n51.8%\n\n80\n\n5\n\n24%\n\n2.8%\n\nGemini\n\n29%\n\n56.7%\n\n55\n\n18\n\n60%\n\n2.5%\n\nPerplexity\n\n17%\n\n28.4%\n\n78\n\n0\n\n19%\n\n10.5%\n\nCopilot\n\n33%\n\n53.5%\n\n58\n\n6\n\n37%\n\n2.1%\n\nGoogle AI Mode\n\n33%\n\n57.9%\n\n60\n\n12\n\n44%\n\n2.7%\n\nA single week’s “AI rank” is close to unquotable. The median hotel that stays on a 500-deep list still moves **55–80 places** between Mondays depending on the engine. Any tool selling you a weekly AI-visibility position is largely reselling this noise.\n\n## 3\\. Three volatility personalities\n\nSame-size churn can hide opposite behaviours, so we tracked each engine’s May 11 top-100 as a cohort: week by week, how many of those exact hotels are still in the top 100? The shape of that curve — flat, sloping, or cratered — separates noise from drift from chaos.\n\nai-hotel-volatility-cohort-survival-2026\n\n### Gemini: noise around a core\n\nThe curve is flat: 53% after one week, **60% after ten**. Hotels leave and come back — 42 of its 60 surviving cohort members dropped out of the top 100 at some point and returned. 18 never left, the most of any engine.\n\n### ChatGPT: slow drift\n\nDecent weekly retention (51.8% at top-100) yet the cohort erodes almost monotonically: 59% → 40% → … → **24%**. Hotels that leave tend to stay gone. Its list isn’t jittering around a fixed set; the set itself is migrating.\n\n### Perplexity: near-lottery\n\nThe cohort collapses to **31% in a single week** and ends at 19%. Not one hotel held its top 100 for all eleven weeks, and 10.5% of each week’s top-100 are hotels never seen before in any prior week’s top-500 — the other engines sit at 2.1–2.8%.\n\nCopilot (37% at week 10) and Google AI Mode (44%) sit between the flat and the drifting curves. Worth noting against the engine-personality picture from the [landscape study](/research/ai-hotel-landscape-2026): Copilot may be the entity engine in _what_ it cites, but its hotel picks churn like everyone else’s.\n\n## 4\\. Membership is calm, position is not\n\nHere’s the reassuring counterweight to all that churn. Ask a different question — not “did the hotel keep its slot?” but “is it still anywhere on the 500-deep list?” — and the picture inverts. A hotel ranked in the top 10 this week stays somewhere in the top 500 next week **94–99%** of the time on four of the five engines (Perplexity, again the outlier, 86%). The deeper the current rank, the leakier it gets, down to 30–45% for ranks 251–500 — the zone where hotels sit on 1–2 weekly mentions and a single conversation can flip membership.\n\nThis week’s rank\n\nChatGPT\n\nGemini\n\nPerplexity\n\nCopilot\n\nGoogle AI Mode\n\n#1-10\n\n99%\n\n96%\n\n86%\n\n97%\n\n94%\n\n#11-50\n\n91.8%\n\n93.2%\n\n68%\n\n93%\n\n94%\n\n#51-100\n\n85.4%\n\n81.2%\n\n53.4%\n\n86.2%\n\n88.8%\n\n#101-250\n\n71.5%\n\n64.1%\n\n42.7%\n\n67.9%\n\n72%\n\n#251-500\n\n45.2%\n\n41.3%\n\n30.1%\n\n45%\n\n43.2%\n\nReading: of the hotels an engine ranks #1–10 this week, the share still anywhere in its top 500 next week. Averaged over all 10 week-pairs.\n\nBeing _known to the engines_ is the durable part. AI hotel visibility behaves like a two-layer asset — a stable pool of hotels the engines keep reaching for, and a weekly shuffle deciding which of them get surfaced on top.\n\n## 5\\. The anchor hotels\n\nAcross five engines and eleven weeks there were 41 engine-seats that never turned over — held by just 29 distinct hotels. 8 of those hotels held a permanent top-100 seat on two or more engines, and 4 managed it on three: Hotel Monteleone (New Orleans), the Cairo Marriott, Marina Bay Sands, and The Langham Gold Coast. The Cairo Marriott’s median position on ChatGPT over the whole run: **#1**.\n\nGemini hosts 18 anchors, AI Mode 12, Copilot 6, ChatGPT 5 — and Perplexity zero, so no hotel can anchor more than four engines even in principle (none reached four). These are disproportionately branded flagship properties: 28 of the 41 seats belong to hotels carrying a chain brand.\n\nEvery hotel that stayed in an engine's top 100 for all 11 weeks (May 11 – Jul 20, 2026), with its best/worst/median weekly rank on that engine.\n\nEngine\n\nHotel\n\nCountry\n\nBrand\n\nBest\n\nWorst\n\nMedian\n\nGoogle AI Mode\n\nThe Taj Mahal Palace, Mumbai\n\nIN\n\nTaj\n\n#4\n\n#66\n\n#5\n\nGoogle AI Mode\n\nHotel Monteleone\n\nUS\n\nindependent\n\n#1\n\n#25\n\n#8\n\nGoogle AI Mode\n\nCairo Marriott Hotel\n\nEG\n\nMarriott\n\n#1\n\n#45\n\n#10\n\nGoogle AI Mode\n\nMarina Bay Sands Singapore\n\nSG\n\nindependent\n\n#1\n\n#57\n\n#11\n\nGoogle AI Mode\n\nHotel Adlon Kempinski Berlin\n\nDE\n\nKempinski\n\n#6\n\n#57\n\n#13\n\nGoogle AI Mode\n\nJW Marriott Gold Coast Resort & Spa\n\nAU\n\nMarriott\n\n#3\n\n#27\n\n#15\n\nGoogle AI Mode\n\nHotel Brunelleschi\n\nIT\n\nindependent\n\n#1\n\n#63\n\n#20\n\nGoogle AI Mode\n\nThe Langham Gold Coast\n\nAU\n\nLangham\n\n#9\n\n#53\n\n#22\n\nGoogle AI Mode\n\nFour Seasons Hotel Seoul\n\nKR\n\nFour Seasons\n\n#8\n\n#45\n\n#24\n\nGoogle AI Mode\n\nFour Seasons Hotel Hong Kong\n\nHK\n\nFour Seasons\n\n#16\n\n#97\n\n#39\n\n## 6\\. Chains vs independents\n\nThe anchor list hints at it and the retention split confirms it: on four of five engines, chain-branded hotels in the top 100 survive to the next week more often than independents. Google AI Mode shows the widest gap (61.6% vs 53.8%). Perplexity is the lone inversion — chains actually churn slightly _more_ there (27.3% vs 29.7%), consistent with an engine that re-rolls most of its list regardless of who you are.\n\nEngine\n\nChain-branded\n\nIndependent\n\nGap\n\nChatGPT\n\n53.6%\n\n49.3%\n\n+4.3 pts\n\nGemini\n\n57.8%\n\n55.3%\n\n+2.5 pts\n\nPerplexity\n\n27.3%\n\n29.7%\n\n\\-2.4 pts\n\nCopilot\n\n56.2%\n\n50%\n\n+6.2 pts\n\nGoogle AI Mode\n\n61.6%\n\n53.8%\n\n+7.8 pts\n\nWeek-over-week top-100 retention for chain-branded vs independent hotels. Positive gap = chains stickier.\n\n## 7\\. What it means\n\n-   **Judge AI visibility on a rolling window, never a single week.** With the median surviving hotel moving 55–80 ranks between Mondays, a 4–8 week average is the shortest honest read. One-week “we entered the AI top 10!” claims — including celebratory vendor screenshots — will usually un-happen by themselves.\n-   **Pick the metric that matches the layer.** Pool membership (are you in the top 500 at all?) is stable enough to track weekly and to alarm on. Position is only meaningful smoothed.\n-   **Know which engine you’re reading.** A Gemini drop is likely jitter that mean-reverts; a sustained ChatGPT slide is more likely real drift worth investigating; a Perplexity swing, in this dataset, is close to weather.\n-   **The anchor tier is small and brand-heavy.** 29 hotels worldwide held a permanent top-100 seat anywhere this season, mostly chain flagships. For everyone else, volatility is the normal condition of AI visibility in 2026 — plan measurement (and expectations) around it.\n\n## Methodology\n\n**Data:** the weekly AI Hotel Landscape pipeline — 616 prompts (56 destinations × 11 templates, EN, mixed personas/budgets) fired each Monday at ChatGPT, Gemini, Perplexity, Copilot and Google AI Mode; answers entity-resolved to canonical hotel IDs; hotels ranked per engine per week by mention count. Analysis window: the 11 consecutive weeks 2026-05-11 to 2026-07-20 — every week of the current 616-prompt library. The two prior weeks (library v1, a smaller prompt set) are excluded because churn across a prompt-library change mostly reflects the new seeding. Grok is excluded: absent from the data since 2026-05-25 and suspended after persistent scraper failures (2 usable weeks in the window); its absence is disclosed, not imputed.\n\n**Pool & metrics:** all comparisons use each engine’s global top-500 per week (the uniform retention cap in the published table; per-country floor rows beyond rank 500 are dropped so every week is compared under the same policy). Week-over-week retention = share of this week’s top-N present in next week’s top-N, averaged over the 10 consecutive week-pairs. Cohort survival tracks the May 11 top-100 membership forward. Rank displacement = median |Δrank| for hotels present in consecutive top-500s. Newcomer share = fraction of a week’s top-100 absent from every earlier week’s top-500 in the window. Anchors = hotels in an engine’s top-100 in all 11 weeks.\n\n**Caveats:** mention counts around rank 100 are 3–5 per week, so tie-ordering contributes noise deep in the list — the depth-gradient section quantifies exactly this, and headline claims stay in the top-10/top-100 where counts are higher. Results describe English prompts at weekly cadence on this prompt library; different phrasings or cadences could churn differently. Scrape-side variance (which conversations a platform returns) is part of what “volatility” means here — it is what a hotel tracking its own AI visibility would experience too. Analysis code and per-metric CSVs: [summary.csv](/data/ai-hotel-volatility-2026/summary.csv) plus the full retention/survival/gradient/anchor tables in the same folder.\n\n## FAQ\n\n[← All research](/research)","author":{"@type":"Person","name":"Nicolas Sitter","url":"https://nicolassitter.com/about","sameAs":["https://www.linkedin.com/in/nicolassitternolleau/","https://github.com/Nicositter88","https://hotelrank.ai"]},"publisher":{"@type":"Person","name":"Nicolas Sitter","url":"https://nicolassitter.com"},"image":"https://nicolassitter.com/api/og/ai-hotel-volatility-2026","mainEntityOfPage":{"@type":"WebPage","@id":"https://nicolassitter.com/research/ai-hotel-volatility-2026"},"tags":["AI Search","Hotels","Volatility","AI Visibility","Time Series"],"sameAs":["https://hotelrank.ai/research/ai-hotel-volatility-2026"],"alternateFormat":{"html":"https://nicolassitter.com/research/ai-hotel-volatility-2026","json":"https://nicolassitter.com/api/post/ai-hotel-volatility-2026","rss":"https://nicolassitter.com/rss.xml"},"datasets":[{"name":"summary","contentUrl":"https://nicolassitter.com/data/ai-hotel-volatility-2026/summary.csv","encodingFormat":"text/csv"}]}