Hospitality AI operating platform: market opportunity

Research draft v0.2 · 15 October 2026 · annual software revenue in USD

1. What the evidence supports

The opportunity is a focused, fragmented market of independently controlled multi-property operators. A platform that connects guest service, reservations, operational tasks, finance, commercial work and management information could support a larger recurring contract than a single guest chatbot. Public evidence does not yet establish how many qualifying groups would buy that platform or the price they would accept.

The covered-country scenario produces $53.3m / $261.6m / $787.3m in annual recurring software opportunity, with a base case of about 2.7 million rooms, 32,000 properties and 4,600 buyer equivalents. These are conditional allocation and pricing scenarios, not observed segment counts, confidence intervals or a sales forecast. They apply proposed exclusions to published accommodation stocks through explicit assumptions. The model has not enumerated and removed every major brand at property level.

The room-stock baseline covers 158 of 244 eligible reporting areas and 28.9 million reported rooms before target-segment filtering; 84% of covered stock uses a 2023-or-later reference or the current AHLA page. It combines different reference years and national definitions. 86 eligible reporting markets remain unquantified, including economically meaningful gaps. The table is a covered global subtotal, not a complete worldwide TAM or a statistically defensible lower bound. A full global figure equals this modeled subtotal plus an unquantified remainder; neither missing capacity nor uncertain eligibility should be treated as zero. Sources and raw observations are retained for revision. S01 S02

The base modeled recurring opportunity is concentrated in Europe, Rest of Asia, LATAM and Caribbean, United States. That does not establish where sales will be easiest. The recommended first validation sequence is UK/Ireland, Canada and Australia; a parallel UAE design-partner track could test the Gulf. Expand into selected continental European and Latin American markets after language, integrations, onboarding and prices have been tested. This sequence is an analytical recommendation with medium-to-low confidence, not a finding from customer interviews.

The three-year commercial model gives $0.05m / $0.51m / $2.41m in illustrative ending ARR. It is built from seller capacity, qualified opportunities, conversion, onboarding capacity and retention. It is not a percentage of TAM. The full platform opportunity and the narrower launch product have different capabilities and prices.

2. Define the customer before counting it

The proposed core is 2–100 operating accommodation properties and no more than 20,000 operating rooms at the parent/operator level, with at least two eligible non-major-brand properties and authority to purchase a shared workflow platform. The small band is 2–9 properties; the midsize band is 10–100. These thresholds are working commercial boundaries, not industry standards or a user-approved definition.

Two properties introduce cross-property knowledge, configuration and reporting needs. Beyond roughly 100 properties or 20,000 rooms, integration breadth, enterprise procurement and support expectations can become a materially different selling and delivery motion. The exact transition is not established by public data. Room and property limits are both necessary: an operator of eight very large resorts can be commercially larger than one with forty small hotels.

A property is one operating hotel, resort or centrally operated aparthotel/serviced-apartment building, with a distinct operating context and system configuration. An annex is not automatically a separate property. A room is a sellable bedroom or serviced unit, not a bed, occupied room-night, holiday-home listing or apartment scattered across an online marketplace. Pipeline projects are excluded from the current operating universe. Resort F&B can be a workflow inside a hotel account; standalone restaurant businesses are outside this sizing.

A chain/operator has shared operating control over multiple properties. An addressable purchasing account is the group budget and decision authority that could contract for the platform. Brand names, legal subsidiaries, hotel owners and management companies do not automatically create additional buyers. A group using three brands may still be one buyer. A property owner whose management company selects operational software is not an additional account.

Include independently controlled hotel and boutique-hotel groups, resort groups, and aparthotel/serviced-apartment operators with suitable hotel-like service workflows. Managed portfolios can qualify where the operator controls the relevant systems and workflows. Exclude single-property independents from core; they are an adjacent, separately unquantified market. Exclude cruises, standalone restaurants, campsites, ordinary residential landlords, marketplaces and unmanaged holiday-home inventories. Property-type uncertainty in broad supply sources is represented by explicit scope factors, not by relabeling every lodging room as a hotel room.

Major-group and affiliation rules

Marriott, Hilton, IHG, Accor and their controlled portfolios are excluded, including franchised and soft-branded properties. Comparable exclusion screening includes Wyndham, Choice, BWH, Hyatt, Jin Jiang/Louvre/Radisson, H World/Deutsche Hospitality, Ascott, Minor/NH, Meliá, Barceló and RIU, along with any other parent above the proposed size limits. This is a screening policy, not an exhaustive verified brand-parent database. Corporate disclosures illustrate the scale difference: Hilton reported more than 9,000 operating hotels in 2025, IHG 6,963 open hotels, and Accor 5,836 hotels in its reported perimeter. Their pipelines are not added. S23 S24 S25

Apply exclusions before defining a prospect's eligible footprint. A five-hotel franchise owner operating only global-brand properties is outside core. A mixed manager can contribute only independently branded properties for which it has purchasing rights, provided the parent/operator itself meets the working size boundary and at least two qualifying properties remain. Do not split a large enterprise into smaller brands to circumvent the boundary. Do not add both the manager and owner for the same deployment.

A marketing collection, membership consortium or soft-brand network is not itself an operating buyer unless procurement control is demonstrated. Membership in a non-controlling independent marketing consortium does not automatically make each hotel a separate chain, or the consortium one buyer. Global-brand soft affiliation is excluded under this brief. White-label managers and small franchisees may offer an adjacent route to market, but no numerical adjacent-market TAM is asserted here: the required property/brand/control join is missing.

Boundary sensitivity

The available sample illustrates different buyer sizes; it does not establish their distribution. For example, Boudl's site lists 47 stays across five brands, while Germain currently reports 20 hotels. A 25-property cutoff would exclude the former and retain the latter. Scandic's disclosed network of about 280 hotels, including development, falls outside even a 200-property screen. That is a credible regional enterprise opportunity that the narrow scope intentionally leaves out, rather than evidence that all regional chains qualify. E06 E01 S26

The broader 200-property / 40,000-room screen would admit commercially relevant regional groups that the proposed core omits. Motel One currently reports 111 operating hotels and about 30,000 rooms. Lemon Tree disclosed 130 operating hotels and 11,772 rooms in February 2026, separately from its pipeline. Both cross the 100-property limit; Motel One also crosses the 20,000-room limit. They illustrate the commercial cost of the boundary without reintroducing Marriott/Hilton-type portfolios. E13 E14

No public group-size census was obtained that can translate 25-, 100- and 200-property thresholds into supported global room allocations. The model therefore exposes the allocation inputs and tests them directly. It does not invent a threshold-to-TAM conversion. A larger parent cap could add buyers while removing no global-brand exclusions; the size of that addition requires an operator register.

3. From accommodation supply to a revenue scenario

The model begins with the latest numerical hotel-room observation in the UN Tourism December 2025 workbook, then replaces it with more comparable Eurostat 2024 observations or specified national/industry anchors. It retains the original value, year, unit, note and flag. Pre-2019 observations and missing rooms are unquantified in the revenue model. Recent publication does not make an old underlying observation current. No automatic room-stock growth projection is applied. UN Tourism warns that its new collection and processing system is transitional. S01

The complete country register includes every M49 reporting area plus separate Taiwan and Kosovo reporting rows, with each market assigned once. China, Hong Kong SAR, Macao SAR, Russia and North Korea are excluded. Including the SARs is an explicit conservative interpretation of the China exclusion. Taiwan is a separate reporting market; this is a commercial data treatment, not a political statement. Antarctica is not modeled as a commercial hotel market. S03

Europe includes Türkiye and Cyprus and excludes Russia. Armenia, Azerbaijan and Georgia remain in the rest of Asia. Middle East comprises Bahrain, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Oman, Palestine, Qatar, Saudi Arabia, Syria, UAE and Yemen. Africa includes Egypt. LATAM and Caribbean includes Mexico, Central and South America and Caribbean reporting areas. Bermuda, Greenland and Saint Pierre and Miquelon are shown separately as other Americas. Australia is separate from the rest of Oceania. Country detail is in the model and the searchable HTML appendix.

The calculation for each quantified country and case is:

1. Fit stock = reported rooms × property-scope factor. The factor is 100% where the chosen source already reports hotels/similar accommodation. It is varied for explicitly broad mixed supplies. This convention does not claim all remaining rooms are commercially ready.
2. Retained non-major stock = fit stock × retained non-major share. This scenario input represents removal of major-brand portfolios and parents above the working boundary. It is not an observed exclusion tally.
3. Eligible rooms = retained stock × qualifying multi-property share. This explicitly removes single-property stock and remaining nonqualifying purchasing structures. It prevents an independent-hotel total from becoming a small-chain total.
4. Eligible properties = eligible rooms ÷ assumed rooms per qualifying property. Assumed sizes vary by region, from 60 to 150 rooms. They describe the target property, not the national hotel average.
5. Buyer equivalents = eligible properties ÷ assumed seven properties per buying group. These are footprint-allocated buyer equivalents, not observed or deduplicated purchasing accounts.
6. Software ARR = buyer equivalents × group fee + eligible properties × property fee + eligible rooms × room fee. Usage, implementation and services are kept outside the recurring total.

The two allocation shares are the largest unresolved inputs. No public evidence directly measures them for this exact target. Country adjustments use chain research to avoid obvious scale contradictions, but are still scenarios. Low/base/high combine different allocations and different prices; they are not probabilities. Changing a case is not a forecast of a country becoming more or less independent.

Modeled room, property and buyer counts in report tables are rounded to approximately two significant figures. The workbook and CSV retain the underlying calculations for audit. Rounded rows may not sum to displayed subtotals.

A cross-border group contributes rooms and property fees where its properties operate. In the modeled account measure, its group fee is allocated by footprint through the seven-property conversion. An actual account register would count the group once and allocate one group fee across its eligible properties. No country row should be read as a headquarters count, nor summed as if independently deduplicated named buyers. The global sum is an estimated number of account equivalents under an assumed mean portfolio, not evidence of that many identifiable buyers.

Calibration and comparability checks

Germany: the 2025 chain study reports 418,034 rooms in 3,054 chain hotels, including 168,228 domestic-chain rooms. Neither the domestic room count nor the full chain stock is automatically eligible: large national parents and global affiliations must also be removed. The base scenario retains 10% of Eurostat's broader national room stock as eligible. This is a chosen allocation, not a measured 10% segment share. The newer DACH 2026 publisher summary was found, but a full report download was blocked; its regional headline is not used to invent country-level counts. S16 S17

Italy: THRENDS' Q4 2025 public preview gives 2,607 chain hotels averaging 92 rooms. Its H1 2026 update describes around 21% chain room penetration. The base target allocation is 8% of national rooms, reflecting a smaller segment than all chains. The two previews have different vintages and should not be forced into a precise growth reconciliation. S14 S15

Spain: the publisher reports 424,807 chain rooms as of March 2026, with 330,263 domestic and 94,544 international rooms, in a 1–5-star hotel universe. Its 61% chain penetration uses a narrower denominator than Eurostat I551. The model does not subtract those chain rooms from an incompatible stock and call the remainder independent. Its base 12% target allocation is unvalidated. S18

Brazil: JLL estimates 565,987 rooms across 10,650 hotels, condo hotels and resorts as of July 2024. National brands account for 131,287 rooms and international brands 107,548. Importantly, national chains with fewer than 600 rooms are included in JLL's independent category. The base model allocates 14% of total stock, about 79,000 rooms, to the target. This is below all national-brand rooms, but not a census: some domestic branded rooms fail the boundary while small groups may be hidden in the independent bucket. The estimate is not added to the much broader older UN observation. S13

India: broad UN supply is about 2.84 million rooms, whereas Horwath's 2024 chain-affiliated supply only passed 200,000 rooms. The model therefore uses much smaller target shares of 0.5% / 1.2% / 2.5% of the broad stock. Even these require parent-level validation. This check demonstrates why applying one global small-chain percentage indiscriminately would mislead. S01 S27

Saudi Arabia: about 545,000 licensed keys in H1 2025 include 83% hotel keys and 17% apartments/other units. Scope factors of 83% / 92% / 100% explicitly vary inclusion of the latter. The base 92% assumes that some, not all, apartment/other keys fit hotel-like operating workflows. New licensing does not establish an equal quantity of newly built rooms. S08

The Eurostat EU27 aggregate is 6,173,288 rooms; available 2024 country observations sum to 6,165,498 because Luxembourg's country room observation is missing. The 7,790-room difference is disclosed and not inserted as a plug. Luxembourg uses its separate 2022 UN observation. This is a small stock reconciliation gap, not evidence of target buyers. S02

4. Geography: market size and quality of evidence

All revenue ranges below are modeled annual recurring software scenarios for the proposed core, excluding the specified countries and global-brand portfolios by assumption. The property and buyer measures are calculated from room allocations. A dash means unquantified, not no opportunity. Underlying country years are shown explicitly.

Europe

Europe offers a dense set of local and regional operating groups, but a different mix in each country. Eurostat supplies a consistent 2024 starting point for many countries; chain studies then provide checks on the branded subset. The UK anchor is Knight Frank's rounded 2025 supply of over 709,000 rooms. Neither hotel establishment counts nor the number of active hotel brands is a purchasing-account count. S02 S07

The UK and Ireland are practical first validation markets because an English-language, human-reviewed workflow can be tested without simultaneously validating another language. They offer an initial learning advantage, not proof of superior commercial returns. Germany, France, Spain, Italy, Portugal and the Netherlands merit subsequent focused work. They require language-specific knowledge, guest-message testing, local integrations and channel partners. Fragmented small hotel ownership can expand supply while increasing sales and onboarding effort. The Nordic market has technology-oriented regional enterprises but a meaningful portion is outside the proposed size boundary; the model does not treat those as small chains.

Middle East

The Gulf combines hotel/residence operations, central teams and cross-property service coordination. UAE supply is about 216,900 rooms in the January–November 2025 government release; Qatar's 42,469 keys include 10,564 hotel-apartment keys. Oman reports 35,948 rooms and 1,031 hotels for 2024. These numbers support a hotel/aparthotel operating-unit opportunity; they do not establish how much inventory is controlled by qualifying local groups. S09 S10 S11

Use the UAE for a parallel design-partner experiment rather than assuming the largest Gulf stock automatically provides the easiest entry. A multilingual guest base, channel variation, group/property authority and enterprise support demands need explicit testing. Saudi Arabia has the largest measured stock in this regional model, but accommodation mix, procurement ownership, implementation capacity and local-language quality deserve separate discovery. Bahrain, Oman, Qatar and Kuwait may be reachable through regional relationships; they are much smaller operating markets. Iran, Levant markets and other eligible countries remain part of the opportunity scope without implying launch readiness or a legal assessment. Iraq and Yemen lack recent usable room anchors here.

Latin America and the Caribbean

Mexico and Brazil are the first commercial research priorities within this region. Both have visible regional operating groups and enough stock to justify buyer-level enumeration. Brazil's stock anchor is an industry estimate with unusually useful chain segmentation. Mexico's 899,389-room figure is Horwath's 2024 estimate using historical official inventory. It is not a verified count of eligible chains. S12 S13

The Chile government's 2026 investor guide provides a useful four-country comparison, but its accommodation years differ: Chile is 2021, Colombia and Peru are 2024, and Mexico is 2019. The model uses that comparison for Chile, Colombia and Peru with broad-lodging scope factors; it retains the newer Mexico anchor. Those values are not represented as one contemporaneous census. A newer Peru document was discovered but returned 403; its headline is not used. Dominican Republic remains unquantified because an adequately verified numerical supply table was not retrieved, despite secondary figures being discoverable. S28 S20

Spanish and Brazilian Portuguese need separate product and sales validation. An English-only launch is not automatically serviceable here. Multi-property service routing and centralized finance may create useful entry points, but local software budgets, cloud/PMS access, channel economics and support expectations remain untested. Caribbean resort groups can have large room counts with few properties and major-brand or resort-parent affiliations; use both size thresholds and portfolio exclusions.

United States, Canada and Australia

The US has the largest single-country accommodation anchor: AHLA reports 5.7 million guest rooms and 64,000-plus lodging properties. Its small-business property count must not be interpreted as the independent-chain population. A family-owned hotel under a major global flag is excluded. The national source does not state a precise stock reference year, so the model labels the source as current-page evidence rather than a 2026 census. Labor intensity is credible context: BLS records about 264,600 hotel/motel/resort desk clerks in 2025. That is not evidence that those roles can be eliminated, nor a software spending pool. S04 S22

Canada's source is a 2025 industry report updated in May 2026: 477,378 rooms and 8,391 establishments. Its smaller country footprint could make account mapping and reference development more manageable than a broad US launch. English/French workflows and dispersed regional operations still require local validation. Australia has 340,662 rooms in establishments with at least ten rooms at December 2025. Its geography and time zones increase implementation and support considerations. Neither source identifies the precise target-chain share. S05 S06

Remaining eligible regions

The rest of Asia is commercially substantial but has some of the widest definition and integration uncertainty. Japan, Indonesia, Thailand, Vietnam, Malaysia and Singapore carry meaningful source capacity. Much of it is single-property, large domestic chain, global brand or a different accommodation type; the allocation scenarios explicitly assume removal. Pakistan, Bangladesh, Nepal and several Central Asian markets remain unquantified with current evidence. The product is not assumed ready for all local languages, channels or PMS families.

Africa includes Egypt, Morocco, Tunisia, South Africa, Kenya and other eligible countries. South Africa's 73,700 hotel stay units are separately identified from camping, guest houses and other accommodation in the national survey. Country room stocks elsewhere are often older and do not reveal local group ownership. Nigeria, Algeria, Tanzania and other material markets are gaps. These should not disappear in a global headline. S01 S19

New Zealand and Pacific markets are retained in the rest of Oceania, with limited smaller-island coverage. The model does not extrapolate a tourism-arrivals ratio into hotel rooms. Other Americas is a small separate reporting bucket to avoid double-counting Bermuda and similar territories. The complete coverage register lists every unquantified area and its missing/stale input.

5. Commercial model and sensitivity

The model uses one hypothetical additive package, not the sum of competing alternative price models. The group fee funds shared control, knowledge, configuration and reporting. The property fee reflects per-site configuration/connectors. The room fee represents operational scale. This packaging is a proposal to test; some buyers may prefer a flat property subscription or outcome/usage pricing instead.

All prices are annual USD scenario inputs, defined directly in USD. The same fee cases are used across countries to keep footprint comparisons transparent; there is no validated regional pricing or affordability adjustment. No exchange-rate conversion is performed and no local-currency vendor price is extrapolated. There is consequently no FX date embedded in the model. Implementation, bespoke connectors, managed services and usage pass-through require separate commercial decisions. A zero modeled usage charge means excluded from TAM, not free underlying inference, telephony or messaging. One-time implementation revenue and customer operating savings are not added to software ARR.

A seven-property, 630-room group would imply full-platform ARR of $29,520 / $59,040 / $88,560 under the three price cases. These are arithmetic illustrations, not validated contract sizes. The base launch package is assumed to cost half the full-platform price, or $29,520 for that same group, because its initial workflows are narrower.

These one-driver changes hold the other base inputs constant. Low/base/high in the headline instead change both allocation and price, producing a much wider combined envelope. The seven-property mean is particularly uncertain: the public operator sample is biased toward visible larger groups and does not measure the long tail of two-to-nine-property operators. Property size also changes the property-fee and group-fee components even if eligible rooms stay constant. A measured portfolio distribution and real commercial offers would improve this model more than additional decimals.

6. TAM, launch SAM and obtainable revenue

TAM: the covered-country, steady-state scenario for the full proposed platform. It assumes every modeled eligible unit purchases one full package, representing theoretical full penetration rather than expected adoption. It includes broader workflows that are not established as built or operator-accepted. It is an opportunity envelope for a future product, not current product revenue potential.

Currently verified saleable SAM: not established. The benchmark evidence contains developer-tested/demo capabilities and unresolved business acceptance. It does not establish a generally deployable, commercially accepted multi-tenant platform. This is an evidence gap, not a claim that demand is zero.

Conditional launch SAM: UK, Ireland, Canada, US, Australia and UAE, restricted to a supported PMS/API path, a suitable English-language/human-review workflow and a customer ready to provide approved knowledge, access controls and staff ownership. Other eligible regions have zero allocation in this launch plan because localization and commercial coverage are not yet planned there; this is not a country ban or a zero TAM assumption.

For the low/base/high launch scenario, the sequential conditional gates are 5% / 15% / 30% compatible PMS, then 60% / 80% / 90% language/workflow fit among the compatible subset, then 40% / 60% / 75% operationally ready among the preceding subset. These are not empirically measured shares. The multiplication is a transparent conditional assumption, not a claim of statistical independence. The 1 October 2026 Mews release reports 15,000 customers across 85 countries. It does not establish whether a customer corresponds to one property, a legal entity or a purchasing group; other Mews pages also use the term properties. No one-to-one conversion is established. These vendor figures are not used as property inventory, target-buyer counts or evidence for the compatible-PMS fraction. S21

Applying those gates and half the corresponding full-platform prices gives $75k / $2.0m / $17.2m of conditional launch ARR opportunity. This should be replaced by a named compatible-account register before an investment or sales-target decision. Supported connector documentation, production security/access approval, customer UAT, reliable handover and an accountable support model are prerequisites for converting it into a credible SAM.

Illustrative SOM, first three commercial years after launch:

New accounts each year equal the lower of rounded-down sales capacity and onboarding capacity. Active accounts equal retained prior-year accounts plus new accounts. Annual retention assumptions are 90% / 93% / 95%. Ending ARR is a run rate, not revenue recognized during the year; no partial-year revenue is implied. The model assumes seven properties and ninety rooms per property per customer, the narrower launch price, and no expansion to the full suite. Fractional retained accounts are expected values, not fractions of actual contracts.

No claim is made that qualified opportunities or conversion rates have been observed. The high case requires multiple sellers and implementation specialists; their salaries, support costs, channel fees, gross margin and cash needs are outside this market-sizing assignment. The SOM cases fit below their corresponding conditional launch SAM. That arithmetic does not validate an acquisition channel or a forecast.

7. Where to validate first

The shortlist weighs the fit with a controllable initial deployment as well as market size. UK/Ireland provides an English-language learning route and proximity to the demonstrated workflow context. Canada offers a manageable national account-mapping exercise and recognizable independent groups; bilingual support remains relevant. Australia offers a similar initial language fit with its own delivery and support constraints. UAE offers concentrated regional operating relationships and hotel/residence use cases, balanced against multilingual and support requirements. None has yet demonstrated willingness to pay for this platform.

The US deserves a targeted regional or channel-led test rather than assuming all 5.7 million rooms are accessible. Germany/France/Spain/Italy/Portugal are attractive expansion research tracks after localization and vendor coverage are known. Mexico and Brazil should receive separate Spanish and Portuguese design-partner research. Saudi Arabia merits dedicated discovery after Gulf procurement, language and apartment/hotel scope are clearer. Remaining regions retain upside but are lower-confidence initial commercialization choices in this version.

The qualitative assessment considers operator density, centralized decision rights, multi-property complexity, labor capacity, integration readiness, budget, language, sales effort and routes to market. Density and labor-economics rankings have not been quantified consistently across countries. Do not present the recommendation as a measured league table or a regulatory risk score. No legal or data-residency conclusion is used to exclude an otherwise eligible country.

8. Operator examples and purchasing-unit tests

These are publicly sourced illustrations of the intended operating pattern, not an enumerated prospect universe, verified procurement map, customer relationship or outreach list. Listed property counts are observations of current public pages where available. Complete room totals, ownership, affiliation and purchasing rights still need checks before final qualification.

The examples show why country stock, brands and buyers diverge. Germain's three brands belong to one national operating group; Evok and Fasano span countries; an annex may share one property context; an own-brand franchise network may not control franchisees' operational software. A collection website alone proves neither central procurement nor independent ownership. The full core estimate cannot be inferred from the average size of these selected examples.

9. Decisions and evidence needed for judge review

The core opportunity appears large enough to justify a bounded commercial validation programme. It is not yet established as a billion-dollar verified target segment. The present base scenario is useful for comparing assumptions and research priorities; its allocation and pricing uncertainties are too large for an investment forecast.

Before promoting a TAM headline into a stakeholder investment claim:

1. Confirm the working parent/property/room boundary and whether to maintain separate regional-enterprise and franchise/white-label adjacent segments. A 25-property and a 200-property screen should be tested against named groups before recalculating aggregate shares.
2. Build an account/property/brand/manager register for at least UK/Ireland, Canada, Australia, UAE, Germany/Spain and Mexico/Brazil. Capture the controlling buyer, operated versus pipeline hotels, rooms, affiliations, supported systems and geographic footprint. Deduplicate at parent and property level. A purposive sample supports an observed floor; it does not by itself estimate an unseen population.
3. Replace eligibility priors with a documented enumeration or a defined, sampled allocation method. Public hotel counts alone cannot supply this bridge. Obtain fuller chain research only through authorized, unpaid access; paid purchases require separate instruction.
4. Test actual annual offers with small and midsize operators and identify who owns the budget, the existing tools displaced, implementation burden and renewal criteria. Record accepted/rejected prices rather than interpreting competitor presence as willingness to pay.
5. Validate a repeatable PMS/channel configuration, production controls, operator UAT and onboarding effort. Count serviceable accounts from actual compatibility, language and operating readiness rather than vendor-wide property totals.
6. Resolve material missing stock and vintages, especially Dominican Republic, Nigeria, Algeria, Tanzania, Pakistan and Bangladesh. Refresh older Asia/Africa/Caribbean anchors, and check whether broad lodging includes serviced apartments, hostels or other non-target units.
7. Rebuild the obtainable-market case from observed qualified pipeline, win rates, deployment capacity, support load and churn. Keep professional services and software economics distinct.

The deliverables are a research draft for judge review. Readiness for combination with the product and competitor reports remains pending that review. The source register records retrieval dates, reference periods, definitions and limitations. The private benchmark evidence is kept in a separate internal note and is not embedded in this report, HTML or model.

Appendix: model files and public sources

The editable spreadsheet contains linked pricing and allocation inputs, a country build, regional outputs and the three-year capacity model. The accompanying country CSV provides all reporting areas, including excluded, missing and stale observations. Source-register CSV/JSON preserve original URLs and reference periods. The HTML's country explorer changes between the same three modeled cases; it does not run a new market estimate or collect data. Derived room, buyer and revenue displays use approximately two significant figures. Published stock facts retain their source precision, while full arithmetic remains in the workbook, CSV and JSON.

S01 — UN Tourism. Accommodation in hotels and similar establishments, December 2025 data workbook. Reference: 2015–2024; country-specific latest years may be older. Retrieved 2026-10-15. Country rooms and establishments, original indicator/unit/flags/notes retained. ISIC 5510 includes country-specific scope. New collection/process is transitional; some observations may not reflect intended standards. No ownership, chain size or purchasing account data.

S02 — Eurostat. tour_cap_nat: establishments, bedrooms and bedplaces in hotels and similar accommodation. Reference: 2024. Retrieved 2026-10-15. NACE I551, unit NR; rooms = BEDRM, establishments = ESTBL. Country values replace UN observations where available. National thresholds/scope differ. EU aggregate is a reconciliation control, never added to countries.

S03 — UN Statistics Division. M49 geographical groupings. Reference: Retrieved 2026-10-15. Retrieved 2026-10-15. Geography register only. Custom commercial regions move Türkiye and Cyprus to Europe; Iran to Middle East; Egypt remains Africa; Mexico and Caribbean in LATAM/Caribbean. No political inference.

S04 — American Hotel & Lodging Association. Our Industry. Reference: Current page; stock reference year not stated. Retrieved 2026-10-15. 64,000+ lodging properties and 5,700,000 guest rooms. Not an operator census. Small-business properties are not synonymous with unbranded multi-property groups.

S05 — Hotels Canada / CBRE. Industry overview. Reference: 2025 Business Intelligence Report, updated May 2026; economic impact uses 2024 operations. Retrieved 2026-10-15. 8,391 establishments and 477,378 rooms. Counts not purchasing accounts.

S06 — Tourism Research Australia / CoStar. Annual benchmark report: Australia's visitor economy in 2025. Reference: December 2025. Retrieved 2026-10-15. 340,662 rooms in accommodation establishments with at least 10 rooms. Smaller establishments omitted. No published ownership/control segmentation.

S07 — Knight Frank. UK Hotel Trading Performance Review 2026. Reference: 2025 actual supply. Retrieved 2026-10-15. Over 709,000 rooms; model conservatively uses 709,000 as rounded anchor. Commercial stock definition differs from UN 2016 observation. Not an additive series or growth bridge.

S08 — Saudi Ministry of Tourism. Hospitality Sector Performance, first half 2025. Reference: H1 2025. Retrieved 2026-10-15. About 545,000 keys and 5,326 licensed facilities; 83% keys hotels, 17% apartments/others. Model applies scope factors to the mixed stock. Licensing expansion is not necessarily all newly constructed capacity.

S09 — UAE Ministry of Economy and Tourism. Hospitality Advisory Council reviews initiatives for 2026. Reference: January–November 2025. Retrieved 2026-10-15. Around 216,900 rooms and 1,260 hotels. Rounded stock retained; do not imply exact 2026 census.

S10 — Qatar Tourism / STR. 2025 Annual Performance Report. Reference: December 2025. Retrieved 2026-10-15. 42,469 keys including 10,564 hotel apartments, with hotel categories separately identified. Operating hotel/aparthotel inventory.

S11 — Oman NCSI. Tourism Statistics Bulletin 2024. Reference: 2024 data; August 2025 issue. Retrieved 2026-10-15. 35,948 rooms and 1,031 hotels, printed p13 / PDF page13 and table p58. Scope/chain membership not identified.

S12 — Horwath HTL. Mexico Hotel Market 2024. Reference: 2024 estimate, March 2025 publication. Retrieved 2026-10-15. 899,389 rooms and 26,911 hotels estimated using historical SECTUR/DATATUR inventory. Estimates, not an official 2024 complete census.

S13 — JLL / FOHB. Lodging Industry in Numbers, Brazil 2024. Reference: July 2024 stock. Retrieved 2026-10-15. 10,650 hotels/condo hotels/resorts, 565,987 rooms; national brands 131,287 rooms, international brands 107,548; independent 327,152. Domestic chains under 600 rooms are classified independent. Stock estimate uses proprietary database; not directly comparable to UN's broader 2016 stock.

S14 — THRENDS. Chains Monitor Italy Q4 2025. Reference: December 2025. Retrieved 2026-10-15. 2,607 chain hotels, average 92 rooms; 31% international share in preview. Neither all domestic chains nor all independent hotels equal the target segment.

S15 — THRENDS. Chains Monitor Italy H1 2026. Reference: H1 2026, published 12 October 2026. Retrieved 2026-10-15. Approximately 21% room penetration and 8% hotel penetration; 197 entries and 27 exits. Definitions/vintages differ from Q4 2025 preview. Full free report form not submitted.

S16 — Horwath HTL. DACH Hotel & Chains Report 2025. Reference: 2025. Retrieved 2026-10-15. Germany 418,034 chain rooms across 3,054 hotels; domestic chain rooms 168,228. Global owner/brand mix still needs exclusion. Web-readable indexed report; direct host download 403.

S17 — Horwath HTL. DACH Hotel & Chains Report 2026. Reference: 30 June 2026. Retrieved 2026-10-15. Summary reports 4,287 DACH chain hotels / 552,132 rooms / 39% room share. Full source 403; summary not used to replace country stock or infer target accounts.

S18 — Horwath HTL. Spain Hotels & Chains 2026 publisher summary. Reference: March 2026. Retrieved 2026-10-15. 2,889 chain hotels / 424,807 rooms; domestic 330,263 and international 94,544. 61% room penetration covers narrower 1–5-star universe than Eurostat broad I551. Calibration only; no subtraction across incompatible scopes.

S19 — Statistics South Africa. Tourist accommodation, January 2026. Reference: January 2026. Retrieved 2026-10-15. 73,700 hotel stay units, separately from guest houses, camping and other accommodation; estimated survey universe. No chain/control split.

S20 — Peru MINCETUR. Oferta y demanda de establecimientos de hospedaje 2025. Reference: 2025. Retrieved 2026-10-15. Direct document retrieval returned 403. Discovery-only: no stock value, reference-year input or property-scope adjustment is drawn from S20. The current Peru input and related scope assumptions are attributed to S28, not this inaccessible source.

S21 — Mews. Guest Messaging and Automations launch: About Mews section. Reference: 1 October 2026 release; rechecked 15 October 2026. Retrieved 2026-10-15. The 1 October 2026 release reports 15,000 customers across 85 countries. It does not define the customer unit as a property, parent group, legal entity or deduplicated chain buyer. Other vendor pages, including the press index (https://www.mews.com/en/press), also use properties; a one-to-one conversion is unverified. No vendor-wide count is used as accommodation stock, target-buyer evidence or a calibrated PMS-compatibility share.

S22 — US Bureau of Labor Statistics. Information clerks: Occupational Outlook Handbook. Reference: 2025 employment; retrieved October 2026. Retrieved 2026-10-15. 264,600 hotel/motel/resort desk clerks, with shifts including evenings/weekends. Context for operations; not savings, software budget or TAM.

S23 — Hilton. 2025 Annual Report: year in review. Reference: 2025. Retrieved 2026-10-15. Surpassed 9,000 operating hotels. Supports exclusion including capital-light/franchise brands; not used as a global supply subtraction.

S24 — IHG. 2025 full-year results. Reference: 31 December 2025. Retrieved 2026-10-15. 6,963 open hotels, more than one million rooms / 20 brands. Group excluded; future pipeline omitted.

S25 — Accor. 2025 Integrated Report. Reference: 2025. Retrieved 2026-10-15. 5,836 hotels / 881,427 rooms in reported perimeter. Exclusion includes controlled brands; joint-venture perimeters require account-level screening.

S26 — Scandic. Capital Markets Day 2025 release. Reference: 2025. Retrieved 2026-10-15. About 280 hotels / 58,000 rooms includes operating and development; exceeds narrow boundary. No operating-only count inferred.

E01 — Germain Hotels. Our hotels. Reference: Current 2026. Retrieved 2026-10-15. 20 hotels across Canada, three brands. Strong candidate for one group buying unit; room cap and procurement authority still need validation.

E02 — Staypineapple. Media / 2025 property awards. Reference: 2025–2026. Retrieved 2026-10-15. 10 hotels nationally in 2025 awards. Independent boutique-group illustration; no relationship or willingness-to-pay claim.

E03 — Classic Lodges. Our hotels. Reference: Current 2026. Retrieved 2026-10-15. Multiple named UK/Isle of Man hotels. Use listed examples, not an asserted audited portfolio total; associate structure needs buyer verification.

E04 — Evok Collection. Hotels. Reference: Current 2026. Retrieved 2026-10-15. Six non-coming-soon hotel listings in Paris, Madrid, Venice, Saint-Tropez. Two coming-soon listings excluded from observed example count. Cross-border account illustration.

E05 — Levahotels Mallorca. Our hotels. Reference: Current 2026. Retrieved 2026-10-15. Four listed hotels; La Luna is an annex and may not be separate commercial/operational property. Model example uses 3–4 operating units, one group, not four buyers.

E06 — Boudl. Hotel brand inventory. Reference: Current 2026. Retrieved 2026-10-15. Lists 27 Boudl, 8 Aber, 9 Braira, 2 Narcissus, 1 Narcissus The Royal stays: 47 listed units. Size candidate, room cap/operating status/procurement not independently audited.

E07 — TIME Hotels. Brands. Reference: Current 2026. Retrieved 2026-10-15. Regional multi-brand hotels and residences across Middle East and Asia. Open versus pipeline inventory not counted; regional group candidate, buyer/room cap unverified.

E08 — Fasano. Hotels. Reference: Current 2026. Retrieved 2026-10-15. Multiple operating hotels Brazil, Uruguay, US; coming-soon/restaurant/beach-club entries excluded. Cross-border group, membership collections are not assumed purchasers.

E09 — Hoteles Misión. Hotels. Reference: Current 2026. Retrieved 2026-10-15. Regional Mexico hotel portfolio / own-brand loyalty. Current complete count not inferred; franchise central procurement rights require validation.

E10 — Lancemore Group. Hotels. Reference: Current 2026. Retrieved 2026-10-15. Eight non-coming-soon hotel listings across Lancemore / LMG; 2027 developments excluded. Group purchasing authority is an assumption.

E11 — Crystalbrook Collection. Hotels and resorts. Reference: Current 2026. Retrieved 2026-10-15. Multi-property Australian collection; listed Adelaide property timing not independently verified, excluded from example count. Candidate rather than account census.

E12 — Sarova Hotels. Book hotels in Kenya. Reference: Current 2026. Retrieved 2026-10-15. Kenyan hotel/lodge group illustration; avoid confusing independent Sarova UK procurement with Kenya group. No complete portfolio or room total claimed.

S27 — Horwath HTL. India Hotel Market Review 2024. Reference: 2024. Retrieved 2026-10-15. Chain-affiliated room supply passed 200,000. The very broad UN stock of 2.84m rooms cannot be treated as chain supply. Used as a scale check; no paid data or registration.

S28 — Chile Subsecretaría de Turismo. Guía del inversionista 2026, Pacific Alliance accommodation comparison, p31. Reference: Chile 2021; Colombia and Peru 2024; Mexico 2019. Retrieved 2026-10-15. Visually checked PDF p31: Chile 120,545 rooms / 8,243 establishments (2021); Colombia 335,295 / 19,484 (2024); Peru 329,340 / 28,050 (2024). Broad lodging; scope factors explicit. Mexico's 2019 comparison is not used. Different years are not interchangeable. Peru uses the 2024 stock of 329,340 rooms and 28,050 establishments. Its 60% / 75% / 90% property-scope factors are unvalidated model assumptions applied to this S28 broad-lodging stock, not percentages from S20 or from the source itself.

E13 — Motel One. Company. Reference: Current 2026. Retrieved 2026-10-15. 111 operating hotels, approximately 30,000 rooms in 14 countries. Excluded under 100-property/20,000-room boundary; potential inclusion under 200/40,000 subject to parent/control checks. Not a core prospect count.

E14 — Lemon Tree Hotels. Q3 and 9M FY26 earnings presentation, p4. Reference: 6 February 2026. Retrieved 2026-10-15. 130 operational hotels and 11,772 rooms, separately from 129 pipeline hotels. Excluded by proposed 100-property cap, potentially included at 200 properties. Later changes not independently mapped.

Country scenario explorer

Published stock and modeled target measures are separate. Changing the case changes scenario outputs only. Search includes excluded and unquantified markets. Derived room, buyer and revenue displays use approximately two significant figures. Published stock facts keep their source precision; full arithmetic remains in the workbook, CSV and JSON.