{"@context":"https://schema.org","@type":"BlogPosting","headline":"Apart-Hotels in Paris on ChatGPT (2026): ChatGPT Doesn't Have an Aparthotel Category — It Has a Duration Heuristic","description":"The standing weekly Paris hotel panel almost never surfaced aparthotel chains — Adagio, the city's largest operator, appeared zero times across 1,305 map entities over three months. A dedicated 68-prompt panel (a full dimension grid: intent × duration × budget tier × party × purpose × landmark, EN+FR) × 5 repeats × ChatGPT × FR proxy, 340 captures in a single run, found the chains were never invisible: tracked brands appear in 98.7% of mapped answers once asked directly (Adagio 72.9% of captures, Citadines 56.2%, Appart'City 39.1%) — not a controlled A/B, since intent, proxy country and prompt language all differ between the two panels. But even on the dedicated panel, 56.1% of surfaced venues are still typed plain 'Hôtel' vs 39.7% apart-typed, and the split runs along brand lines inherited from Google's own categorisation, not property type: Adagio is 'Hôtel' 97.8% of the time, Appart'City is 'Résidence hôtelière' 100% of the time. The category taxonomy is locale-locked to the Bright Data proxy's country, not the prompt language (99.8% French tokens even under English prompts). Length of stay is the strongest lever tested: three-month prompts return 70.5% apart-typed venues vs baseline; a gym-and-pool amenity request returns zero. Along the way, a pipeline gap: ChatGPT's own is_map flag is true on 92.4% of captures but the structured field we normally read only populates on 73.2% — the rest render the map as inline markdown text, recoverable with a dedicated parser. Repeat consistency across 5 identical prompts: mean Jaccard 0.270, with 65.3% of prompt-venue pairs appearing in exactly one of five repeats.","datePublished":"2026-08-25","dateModified":"2026-08-25","url":"https://nicolassitter.com/research/aparthotels-paris-ai-search-2026","category":"research","keywords":["aparthotel Paris ChatGPT","best aparthotel Paris AI","ChatGPT map widget categories","Adagio Citadines AI search","serviced apartments Paris AI","ChatGPT local search hotels","AI search extended stay","GEO aparthotel"],"articleSection":"Research","wordCount":3300,"readTime":"13 min","articleBody":"|\n\n|\n\nAugust 2026AI Search Studies\n\n# Apart-Hotels in Paris on ChatGPT (2026):It doesn’t have a category — it has a duration heuristic\n\n**TL;DR:** The standing weekly “best hotels in Paris” panel almost never surfaced aparthotel chains — Adagio, Paris’s largest operator, appeared **zero** times across 1,305 map entities over three months. We built a dedicated 68-prompt panel (EN+FR, 5 repeats each, FR proxy) to find out whether ChatGPT was ignoring the category or the panel had simply never asked. It was the panel: once asked directly, tracked chains show up in **98.7%** of mapped answers. But even here, **56.1%** of what surfaces is still typed plain “Hôtel” rather than an apartment category, and the split runs along **brand** lines inherited from Google (Adagio is “Hôtel” 97.8% of the time; Appart’City is “Résidence hôtelière” 100% of the time) — a labelling artifact of the underlying business listing. The one lever that reliably moves it is length of stay: ask for three months and **70.5%** of venues come back apart-typed; ask for a gym and a pool, **zero** do. And the category label is locked to the proxy country regardless of which language you prompt in.\n\nPublished August 25, 2026\n\n68\n\nPrompts (EN+FR)\n\n5×\n\nRepeats per prompt\n\n340\n\nCaptures\n\n1,978\n\nMap entities\n\n[Read the Report](#executive-summary)\n\n[Summary](#executive-summary)[1\\. The Panel Never Asked](#never-asked)[2\\. The Missing 21%](#extraction-gap)[3\\. Brand, Not Property](#taxonomy)[4\\. The Duration Lever](#duration)[5\\. Top Venues](#venues)[6\\. Citations](#citations)[7\\. Ask It Five Times](#consistency)[For Aparthotel Operators](#takeaways)[Caveats](#caveats)[Methodology](#methodology)[FAQ](#faq)\n\n## Executive Summary\n\nThe generic hotel panel’s blindness to aparthotel chains was never a ChatGPT behaviour — it was a question we hadn’t asked. Once asked, the interesting question moves to how ChatGPT categorises what it finds.\n\nThis is a spin-off of our standing AI Hotel Landscape tracking. The weekly Paris panel runs 11 generic English prompts (“best hotels in Paris”-style) on a US proxy. Over three months and 1,305 map-widget entities, it surfaced exactly seven entities that belonged to a known aparthotel chain — three brands, and Adagio, the largest operator in the city, not once. That gap is the whole reason this study exists: is ChatGPT’s underlying knowledge of the category actually thin, or does a query for “hotels” simply never reach it?\n\nWe built a 68-prompt panel spanning duration, budget, party size, purpose and landmark, in English and French, fired five times each on a French proxy — 340 captures in one run, zero failures. Tracked chains appeared in 98.7% of mapped answers. The category was never invisible to ChatGPT; the standing panel had simply never asked a question shaped like this one.\n\nWhat the dedicated panel then exposes is more interesting than a coverage gap. Even here, a majority of surfaced venues are still typed “Hôtel,” and which label a brand gets is inherited from Google’s own business categorisation of that brand. The category label is also locked to the Bright Data proxy’s country, regardless of prompt language. And the dimension that reliably shifts the mix — length of stay — is not the one most people would guess.\n\n## 1\\. The chains were never invisible — the panel never asked\n\nThe standing weekly Paris hotel panel (public dashboard, US proxy, 11 English generic prompts, 2026-05-04 through 2026-08-24) produced 85 captures and 1,305 map entities over three months. Here is every aparthotel chain in it:\n\nAparthotel chains in the generic weekly Paris hotel panel, 3 months of data\n\nBrand\n\nEntities (of 1,305)\n\nCaptures\n\nMy Maison In Paris\n\n4\n\n1\n\nCitadines\n\n2\n\n1\n\nFraser Suites\n\n1\n\n1\n\n**Adagio, Appart’City, Odalys, Residhome, Staycity, Yuna, Edgar Suites, Highstay**\n\n**0**\n\n**0**\n\nSeven entities out of 1,305 — 0.5%, spread across three brands. Adagio, the largest aparthotel operator in the city, does not appear once. On the dedicated aparthotel panel, the same brands are the majority of what comes back:\n\nChain share: generic hotel panel vs dedicated aparthotel panel\n\nAparthotel chains on the dedicated 68-prompt panel, 340 captures\n\nBrand\n\nProse captures (of 340)\n\nMap appearances\n\nMap captures\n\nDistinct properties\n\nAdagio\n\n248 (72.9%)\n\n502\n\n227\n\n43\n\nCitadines\n\n191 (56.2%)\n\n252\n\n160\n\n14\n\nAppart'City\n\n133 (39.1%)\n\n179\n\n124\n\n9\n\nYuna\n\n87 (25.6%)\n\n51\n\n47\n\n11\n\nStaycity\n\n83 (24.4%)\n\n91\n\n79\n\n6\n\nResidhome (Réside Études)\n\n73 (21.5%)\n\n71\n\n58\n\n14\n\nMy Maison In Paris\n\n57 (16.8%)\n\n62\n\n50\n\n6\n\nFraser Suites\n\n56 (16.5%)\n\n41\n\n41\n\n3\n\nOdalys City\n\n49 (14.4%)\n\n48\n\n42\n\n3\n\n**This is not a controlled A/B.** The two panels differ on three axes at once: intent (generic “hotels” vs explicit “aparthotel”), proxy country (US vs FR), and prompt language (English-only vs EN+FR). The honest claim is that the generic hotel panel does not surface this category and a dedicated aparthotel panel does — not that any single one of those three changes is what did it. Isolating the axis would need a matched follow-up run.\n\nTwo chains from our target list are genuinely absent everywhere we looked: Nemea (2 prose mentions, 0 map appearances) and Cosy Home (0 anywhere). “Réside Études” never surfaces as a group name either — only its consumer brands (Residhome, Séjours & Affaires, Les Estudines) do.\n\n## 2\\. 21% of the map data doesn’t look like map data\n\nChatGPT flags `is_map: true` on 314 of the 340 captures (92.4%) — but the structured field our pipeline normally reads, `business_locations`, is only populated on 249 (73.2%). The other 70 captures render the same widget in a second Bright Data format entirely:\n\nChatGPT's two mutually exclusive map render modes, this panel\n\nRender mode\n\nCaptures\n\nModel field\n\nWeb search triggered\n\nEntities\n\nPer capture\n\nStructured (business\\_locations)\n\n249\n\ngpt-5-6\n\ntrue\n\n1,120\n\n4.50\n\nMarkdown-serialised carousel\n\n67\n\nnull\n\nfalse\n\n858\n\n12.81\n\nNo map recovered\n\n24\n\nmixed\n\nmixed\n\n0\n\n—\n\nIn the second mode, ChatGPT writes the carousel straight into the answer’s markdown as plain text — `**Name**★ 4.8•Résidence hôtelière116 $US` — with the same underlying fields (name, rating, category, price) plus a static Mapbox pin URL carrying the coordinates. We wrote a parser to recover it. The two modes never overlap in the same capture: wherever the structured field is present, the markdown format is absent, and vice versa. Recovering it lifts total map coverage from 73.2% to **92.9%** (316 of 340 captures) and total entities from 1,120 to **1,978**. The recovered set agrees with ChatGPT’s own `is_map` flag on 312 of 314 flagged captures.\n\n**The two modes are not equal-depth.** The markdown carousel exposes roughly 12.8 venues per map; the structured field exposes 4.5. “How many venues does ChatGPT show for this query” is not comparable across the two modes unless you say which one you’re counting. We’ve filed the fix — folding the markdown parser into the extraction pipeline as a third fallback — as a follow-up ticket; every number below pools both modes.\n\nOne more trap sits in the raw response: a separate array, `raw_response.map[]`, carries 2,416 generic map pins across the same 249 structured captures. It looks like it should be the recommendation set. It isn’t — its mean name-overlap with the actual `business_locations` array is 0.01 venues per capture, and it repeats the same pin multiple times within a single capture. Anything treating it as ChatGPT’s picks is measuring the wrong array.\n\n## 3\\. The category label follows the brand and the proxy country\n\n2,123 category tokens across 1,978 entities resolve to 40 distinct labels. 2,118 of them (99.8%) are French — including under English-language prompts. Under the FR proxy, English prompts return French category labels: of 1,249 tokens attached to English-language captures, only 5 are English, and those five attach to just two structurally anomalous venue records (they carry extra fields absent everywhere else, suggesting a different upstream place source, not a language effect). Zero English tokens appear under the 874 tokens from French-language prompts.\n\nA sharper test: 113 venues appear under both English and French prompts in this panel. 93 of them carry an identical label set either way. Of the 20 that differ, 19 differ only in how many labels are listed (verbosity), and exactly one differs in language. The US-proxy version of the same Paris taxonomy, over the same period, returns English labels instead — Hotel, Extended stay hotel, Holiday apartment rental. “Résidence hôtelière” is the FR-locale name for the same “Extended stay hotel” class. The taxonomy is localised end-to-end by the proxy’s country, independent of the language you prompt in.\n\nTop category tokens across all 1,978 map entities (28 more, all ≤11, omitted)\n\nCategory token\n\nCount\n\nShare\n\nDistinct venues\n\nClass\n\nHôtel\n\n1,133\n\n53.4%\n\n256\n\nhotel\n\nRésidence hôtelière\n\n666\n\n31.4%\n\n77\n\napart\n\nLocation d'appartement de vacances\n\n40\n\n1.9%\n\n17\n\napart\n\nLogement indépendant avec services\n\n30\n\n1.4%\n\n12\n\napart\n\nAppartement avec services\n\n30\n\n1.4%\n\n13\n\napart\n\nAppartement de vacances\n\n24\n\n1.1%\n\n6\n\napart\n\nAppartement\n\n19\n\n0.9%\n\n16\n\napart\n\n**“Hôtel de ville” is not a hotel.** It’s Google’s own label for a town hall, and it appears 17 times in this dataset. It only ever shows up alongside a real lodging label on the same venue, so it never changes a venue’s classification here — but a naive text match on `contains(\"Hôtel\")` would silently corrupt on it.\n\nRolled up into four classes, even an explicitly aparthotel-intent panel returns mostly plain hotels:\n\n56.1%\n\nPlain hotel\n\n1,110 entities · 247 venues\n\n39.7%\n\nApart-typed\n\n786 entities · 125 venues\n\n3.5%\n\nOther lodging\n\n69 entities · 42 venues\n\n0.7%\n\nNon-lodging\n\n13 entities · 12 venues\n\nAnd the hotel/apartment split runs along brand lines, not along what the property actually is. Adagio and Citadines run comparable serviced-apartment products in the same city and land on opposite sides of Google’s own category system:\n\nDominant category label by brand — inherited from Google's own business listing\n\nBrand\n\nDominant label\n\nShare of that brand’s tokens\n\nAdagio\n\nHôtel\n\n491 / 502 (97.8%)\n\nOdalys City\n\nHôtel\n\n48 / 48 (100%)\n\nMy Maison In Paris\n\nHôtel\n\n62 / 62 (100%)\n\nAppart'City\n\nRésidence hôtelière\n\n179 / 179 (100%)\n\nCitadines\n\nRésidence hôtelière\n\n246 / 296 (83.1%)\n\nResidhome\n\nRésidence hôtelière\n\n64 / 75 (85.3%)\n\n## 4\\. Length of stay moves the category more than anything else we tested\n\nAcross every prompt dimension we built into the panel — duration, purpose, budget tier, party, landmark, language — the apart-typed share of surfaced venues moves most with duration:\n\nApart-typed share of surfaced venues, by prompt duration\n\nA three-month relocation prompt returns 70.5% apart-typed venues against a 39.7% panel baseline — a factor of 1.8. A two-week stay returns less than baseline (27.8%, on a smaller 15-capture cell). ChatGPT doesn’t appear to treat “aparthotel” as a fixed category so much as infer it from how long you say you’re staying: ask for a month or a quarter and it reaches for a real serviced residence; ask for a weekend and it mostly hands back hotels that happen to be branded Adagio.\n\nApart-typed vs hotel-typed share, by purpose and by prompt language\n\nDimension\n\nValue\n\nCaptures\n\nEntities\n\nApart-typed\n\nHotel-typed\n\nPurpose\n\nRelocation\n\n20\n\n138\n\n61.6%\n\n37.0%\n\nPurpose\n\nMedical stay\n\n10\n\n61\n\n52.5%\n\n47.5%\n\nPurpose\n\nEvent\n\n5\n\n26\n\n42.3%\n\n57.7%\n\nPurpose\n\nLeisure\n\n255\n\n1,489\n\n38.4%\n\n58.4%\n\nPurpose\n\nBusiness\n\n30\n\n168\n\n36.9%\n\n58.3%\n\nPurpose\n\nRemote work / digital nomad\n\n10\n\n61\n\n29.5%\n\n60.7%\n\nPurpose\n\nStudent\n\n10\n\n35\n\n17.1%\n\n28.6%\n\nLanguage\n\nFrench prompt\n\n145\n\n823\n\n43.6%\n\n53.3%\n\nLanguage\n\nEnglish prompt\n\n195\n\n1,155\n\n37.0%\n\n58.1%\n\n**All captures**\n\n—\n\n**340**\n\n**1,978**\n\n**39.7%**\n\n**56.1%**\n\n**The digital-nomad framing underperforms.** The purpose most people would guess pulls hardest toward serviced apartments — “remote work,” “télétravail” — comes in at 29.5% apart-typed, below the leisure baseline. That phrasing pulls design-forward hotels with coworking space instead, not aparthotels.\n\nIndividual prompt intents show the sharpest swings, though each rests on just 5–20 captures (one or two prompts × 5 repeats) — illustrative, not a measured trend on their own:\n\nSharpest per-intent extremes (≥10 entities each)\n\nIntent\n\nApart-typed\n\nHotel-typed\n\nLong-stay relocation\n\n70.5%\n\n26.2%\n\n\"Citadines vs Adagio\"\n\n65.2%\n\n34.8%\n\nLong-stay, one month\n\n54.5%\n\n45.5%\n\n\"Aparthotel vs hotel\"\n\n23.9%\n\n73.9%\n\nBusiness convention (Porte de Versailles)\n\n5.9%\n\n88.2%\n\n**With a gym and a pool**\n\n**0.0%**\n\n**100%**\n\nAsking for a gym and a pool flips the category off entirely — 29 of 29 entities come back plain hotels. An amenity constraint overrides whatever category signal “aparthotel” was carrying in the rest of the prompt.\n\n## 5\\. Which properties actually show up\n\n426 distinct venues appear across the panel; the top 10 account for 648 of 1,978 map appearances (32.8%).\n\nTop 10 venues by map appearances (of 426 distinct venues surfaced)\n\n#\n\nVenue\n\nMap appearances\n\nDistinct prompts\n\nClass\n\nRating\n\nReviews\n\nPrice\n\n1\n\nAparthotel Adagio Paris Montmartre\n\n100\n\n37\n\nhotel\n\n4.4\n\n1,115\n\n123 $US\n\n2\n\nAparthôtel Adagio Paris Centre Tour Eiffel\n\n95\n\n38\n\nhotel\n\n3.6\n\n4,181\n\n147 $US\n\n3\n\nCitadines Saint-Germain-des-Prés Paris\n\n83\n\n35\n\napart\n\n4.1\n\n1,110\n\n161 $US\n\n4\n\nAppart'City Collection Paris Gare de Lyon\n\n82\n\n35\n\napart\n\n4.8\n\n1,385\n\n122 $US\n\n5\n\nHôtel Staycity Appartement Gare de l'Est\n\n55\n\n27\n\napart\n\n4.4\n\n592\n\n114 $US\n\n6\n\nAparthotel Adagio Paris Buttes Chaumont\n\n50\n\n25\n\nhotel\n\n4.2\n\n942\n\n123 $US\n\n7\n\nCitadines Les Halles Paris\n\n50\n\n28\n\napart\n\n4.0\n\n1,298\n\n193 $US\n\n8\n\nCitadines Bastille Marais Paris\n\n49\n\n26\n\napart\n\n4.1\n\n764\n\n94 $US\n\n9\n\nMode Paris Aparthotel\n\n46\n\n25\n\napart\n\n4.7\n\n136\n\n290 $US\n\n10\n\nFraser Suites Le Claridge Champs-Élysées\n\n38\n\n17\n\napart\n\n4.4\n\n1,568\n\n619 $US\n\n**The one independent in the top 10 is the outlier worth noting.** Mode Paris Aparthotel places 9th on 46 appearances across 25 different prompts with only 136 reviews — an order of magnitude below every chain property around it (592–4,181 reviews). Its rating is 4.7. Whatever ChatGPT is weighting to surface it this often, review volume alone doesn’t explain it.\n\nProse and map mostly agree on which brands show up, but three run heavily prose-only: Zoku (21 prose captures vs 4 map appearances), Yuna (87 vs 47), Edgar Suites (13 vs 5). ChatGPT talks about these brands in the answer text more than it plots them. No brand shows the reverse pattern. Prose counts here are a folded substring match against the answer text, not named-entity recognition, so treat them as an upper bound.\n\n## 6\\. Brand sites lead the citations; Booking.com is still the single biggest domain\n\n1,050 citations across 128 domains, a mean of 3.09 per capture. 323 of 340 captures carry at least one citation.\n\nCitations by source-taxonomy bucket\n\nTop cited domains\n\nDomain\n\nCitations\n\nCaptures\n\nBucket\n\nbooking.com\n\n149\n\n96 (28.2%)\n\nOTA\n\nall.accor.com\n\n89\n\n66\n\nBrand site (Adagio's parent)\n\ndiscoverasr.com\n\n80\n\n53\n\nBrand site (Ascott / Citadines)\n\nadagio-city.com\n\n67\n\n53\n\nBrand site\n\ntripadvisor.fr\n\n59\n\n50\n\nReview aggregator\n\nparisjetaime.com\n\n43\n\n36\n\nCity tourism office\n\n45.4% of citations go to an operator’s own domain — a strong first-party web presence for this category. Booking.com is nonetheless the single most-cited individual domain, present in 28.2% of all captures. The split flips for comparison prompts: asked “aparthotel vs hotel, which should I pick,” ChatGPT’s citations shift to city tourism offices (39.0% of that slice’s citations, on a thinner 77-citation base) instead of brand sites.\n\n**No reliable retrieved-vs-cited split here.** The two render modes from Section 2 emit different citation schemas — 892 citations from the structured mode carry rank/title/snippet fields with no `cited` flag; 158 from the markdown mode carry a `cited` boolean instead. Domain counts above pool both, but a citation-throughput metric can’t be computed cleanly on this dataset.\n\n## 7\\. Ask the same prompt five times, get a mostly different map\n\nEach of the 68 prompts fired five times, back to back, in the same run — this is exactly what the repeats were for.\n\n0.270\n\nMean Jaccard\n\nacross 5 repeats, n=68 prompts\n\n65.3%\n\nAppear in exactly 1 of 5\n\n783 of 1,199 prompt-venue pairs\n\n4.1%\n\nAppear in all 5\n\n49 of 1,199 prompt-venue pairs\n\n52 / 68\n\nMap in all 5 repeats\n\nrest flicker on/off\n\nTwo-thirds of the venues ChatGPT shows for a given aparthotel query appear once in five identical asks. Stability tracks how constrained the prompt is: tight geographic anchors and named-brand comparisons hold together; open-ended ranking requests do not.\n\nRepeat consistency, selected prompts (full 68 in the underlying dataset)\n\nPrompt\n\nRepeats with a map\n\nUnion of venues\n\nMean Jaccard\n\nbest aparthotel in Montmartre Paris\n\n5\n\n5\n\n0.840\n\nbest aparthotel near the Eiffel Tower Paris\n\n5\n\n3\n\n0.683\n\nCitadines vs Adagio aparthotels in Paris\n\n5\n\n10\n\n0.481\n\n_(panel mean)_\n\n—\n\n—\n\n_0.270_\n\nappart hotel paris famille avec cuisine\n\n4\n\n18\n\n0.043\n\nbest aparthotel in the 1st arrondissement Paris\n\n4\n\n7\n\n0.039\n\nNo language effect on stability: English prompts averaged 0.271 (n=39), French 0.268 (n=29). And this is a lower bound on real-world drift — all 340 captures were fired within a single hour on a single day. Week-to-week variation would plausibly be larger than what five identical prompts in one hour already show.\n\n## For aparthotel operators\n\n-   Your category label on ChatGPT’s map widget is Google’s business category, not a description of your product. If your Google Business Profile is filed as “Hôtel,” that’s what surfaces on aparthotel-intent queries too — correcting the underlying listing category is upstream of anything ChatGPT does.\n-   Long-stay framing works in your favour if your listing is already categorised as serviced accommodation: relocation and month-plus queries pull apart-typed venues at nearly double the rate of short-trip queries. Content and metadata aimed at “monthly stay,” “relocation,” or “corporate housing” searches sits on the strongest lever we measured.\n-   Don’t assume digital-nomad-flavoured content is the right hook. In this dataset, “remote work” framing pulled ChatGPT toward design hotels with coworking space, not aparthotels — it underperformed the leisure baseline.\n-   Your own site still matters: 45.4% of citations in this category went to a brand/operator domain. Booking.com remains the single biggest individual domain (28.2% of captures) — owning your Google category and your own site does not displace the OTA layer, it sits alongside it.\n\n## Caveats\n\n-   **The old-panel comparison is confounded three ways** (intent, proxy country, prompt language). Read it as “the generic panel doesn’t surface this category and a dedicated one does,” not as a clean isolated effect.\n-   **Small cells in the dimension analysis.** Most purpose, budget-tier and landmark values rest on 5–20 captures (one to four prompts × 5 repeats). Duration, purpose and language are the only cuts wide enough to publish as trends; individual intents are illustrative.\n-   **Two render modes, different depth** (4.5 vs 12.8 entities per map). Pooled totals are fine; per-capture “how many venues” comparisons are not, unless the mode is stated.\n-   **Brand matching is a folded-text substring match, not NER.** “La Clef” can collide with the ordinary French word “la clef” (the key); its 19 prose captures are an upper bound, like every prose count here.\n-   **33 map entities (1.7%) have a null name** in the structured mode; they still carry address, rating, category and website, and are named from their domain for classification purposes but excluded from the venue leaderboard.\n-   **Single run, single hour, single day (2026-08-14), single platform.** All 340 captures are one hour of ChatGPT. Repeat consistency is measured within that hour and is a lower bound on real week-to-week drift.\n-   **No ground-truth entity seed was built for this study.** Nothing here measures hallucination or verifies that a named property is real — the underlying venues are well-known chains, but that was not independently checked property by property.\n-   **Disclosure:** no affiliation with any operator named in this article.\n\n## Methodology\n\n68 prompts were built as a full dimension grid — intent × duration × budget tier × party size × purpose × landmark — in English and French, covering general recommendations, price bands, family and group stays, business trips, long-stay and relocation framing, digital-nomad framing, amenity constraints, landmark- and arrondissement-anchored queries, and direct brand/format comparisons (“aparthotel vs hotel,” “Citadines vs Adagio,” “aparthotel vs Airbnb”).\n\nEach prompt was fired 5 times, ChatGPT only, on a single Bright Data proxy country (France), in a single run on 2026-08-14. 340 captures, 0 failures, 0 empty answers. We did not run a fresh Apify seed for this study: an earlier check of three months of the existing weekly Paris hotel panel showed that the map widget’s own `categories` field already carries the taxonomy needed (Extended stay hotel / Holiday apartment rental / plain Hotel, in the US-proxy English equivalent) — the gap wasn’t a missing taxonomy, it was that the standing panel never asked an aparthotel-shaped question.\n\nMap entities were extracted from `raw_response.business_locations` where present, and recovered from a second, markdown-serialised render mode with a dedicated parser where it was not (Section 2). Category labels were read from each entity’s `categories` array (the singular `category` field is confirmed dead in the current ChatGPT response format — null on every structured row in this dataset). Brand identity was resolved by case- and accent-insensitive substring matching against a curated list of known Paris aparthotel operators — not named-entity recognition — so every prose mention count in this article is an upper bound. Citations were parsed from each capture’s citation array and bucketed into a source taxonomy (brand/operator site, OTA, review aggregator, editorial, city tourism office, other) by domain.\n\nThe generic-panel comparison in Section 1 draws on the separate, standing weekly AI Hotel Landscape pipeline: 11 English prompts, US proxy, 2026-05-04 through 2026-08-24, 85 captures, 1,305 map entities — queried directly for this article, not re-run.\n\n## Frequently Asked Questions\n\nYes, heavily — but only once asked directly. On the standing weekly “best hotels in Paris” panel, Adagio never appeared once across 1,305 map entities over three months; only 7 entities (0.5%) belonged to any aparthotel chain. On a dedicated 68-prompt aparthotel panel, Adagio appeared in 72.9% of captures and 98.7% of mapped answers contained at least one tracked chain. ChatGPT wasn’t ignoring the category — the generic panel had never asked about it.\n\n![ChatGPT map widget for 'meilleurs apart hotel a paris', showing venues tagged both 'Hotel' and 'Extended stay hotel'](/_next/image?url=%2Fresearch%2Faparthotels-paris-ai-search-2026%2Fchatgpt-answer-example.png&w=3840&q=75&dpl=dpl_ERPXjuLofJRGTM2F5AmtjGVJ8AwH)\n\nThe screenshot that started this study: ChatGPT’s map widget tagging “MY MAISON IN PARIS - Louvre” as a Hotel and “Citadines Saint-Germain-des-Prés” as an Extended stay hotel, for the same query.\n\n### Summarize with AI","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/aparthotels-paris-ai-search-2026","mainEntityOfPage":{"@type":"WebPage","@id":"https://nicolassitter.com/research/aparthotels-paris-ai-search-2026"},"tags":["AI Search","Local Search","Hotels","Paris","Map Widget","GEO"],"sameAs":["https://hotelrank.ai/research/aparthotels-paris-ai-search-2026"],"alternateFormat":{"html":"https://nicolassitter.com/research/aparthotels-paris-ai-search-2026","json":"https://nicolassitter.com/api/post/aparthotels-paris-ai-search-2026","rss":"https://nicolassitter.com/rss.xml"},"datasets":[{"name":"summary","contentUrl":"https://nicolassitter.com/data/aparthotels-paris-ai-search-2026/summary.csv","encodingFormat":"text/csv"}]}