drishva.ai

Home / Case studies / Multi-brand group

Case study

A 129-hotel group, and the entity problem nobody had measured

Multi-brand, asset-light, 72 destinations. The scale that made the group commercially efficient was also what made it difficult for an AI model to resolve.

The group

An Indian hotel group founded in 1994 and listed in 2025, operating in the mid-market segment across upscale, upper-midscale, midscale and economy tiers. At the time of assessment: 129+ hotels across 72 destinations, over 6,000 keys.

What makes the group commercially distinctive also makes it structurally complex for AI visibility. It operates through five different models — owned, leased, revenue-sharing, managed and franchise — and its properties trade under multiple international franchise brands rather than a single unified name.

That structure is efficient. It is also, from a machine's perspective, a set of loosely-related entities rather than one recognisable group.

Why this group was a useful first engagement

Three characteristics made it representative of the problem we built the firm to solve:

  • Multi-brand. Properties trade under several franchise names, so group-level authority does not automatically flow to any individual hotel.
  • Multi-tier. Economy and upscale properties compete in entirely different query sets, in the same cities.
  • Multi-model. Franchised and managed properties have different levels of central control over digital surfaces — which changes what can be executed directly versus specified.

What we measured

A representative sample of properties across tiers and city types, against the standard methodology: 25 queries per property, four engines, five named competitors per property, run from the relevant source markets.

Query categories included destination-plus-segment ("business hotel near Hyderabad IT corridor"), amenity-led, proximity ("hotel near Mumbai airport"), occasion, and price-band variants.

Measurement in progress

The full baseline is being completed. Share of Model figures, per-property breakdowns and the competitive delta will be published here once the first measurement cycle is complete and the group has approved publication.

We publish numbers when they exist and are verified — not before. If you would like the methodology in the meantime, it is set out in full on the Share of Model page.

What the assessment surfaced

The entity fragmentation problem

The most significant structural finding was not about content. When a group operates under several franchise brands, each property's identity is anchored to the franchise name rather than the group — and the group itself has no consolidated machine-readable presence tying them together.

The practical consequence: reputation and citation accumulated by one property contributes nothing to another. A group with 129 hotels can have the machine-readable footprint of 129 unrelated independents.

The unclaimed surfaces

Consistent with every assessment we have run, the highest-authority surfaces were the least contested — national and state tourism board listings, secondary map platforms, and structured entity records. These are free, permanent, treated as authoritative by models, and were substantially unclaimed across the portfolio.

The proximity opportunity

Several properties hold genuinely strong positional advantages — airport proximity in particular — that were well documented in review content and weakly expressed in structured, extractable form. This is Expression in the CITED framework: the fact exists and guests confirm it, but it is not stated in a way a model can quote.

What this tells a group of any size

Three findings that have held across every assessment we have run.

  • Scale does not produce visibility. A group with 129 hotels is not automatically more visible than an independent. If the entity signal is fragmented, scale works against you — the same effort spreads across more, weaker records.
  • The highest-value surfaces are the least contested. Tourism boards, structured entity records and secondary map platforms are free, authoritative and largely unclaimed. This is not a criticism of anyone's existing agency; these surfaces sit outside every standard digital marketing contract.
  • Existing strengths are usually unexpressed rather than absent. Properties frequently have real advantages that guests confirm in reviews and that no structured content states in a form a model can extract.

Published with the group's context described in general terms. Named results will be published on completion of the first measurement cycle, with the group's approval.

Run the same assessment on your group

Five questions, four engines, your properties. We will send you exactly what came back.

Request a free visibility check