Case Studies — Real Analysis, Identity Concealed

Studies built from real reviews of real hotels in the Saudi market. The numbers come straight from our analysis — but the client's identity, location, and every identifying detail are fully redacted.

What the Two Studies Share — and Where They Diverge

Both studies come from the same market and the same two review sources (Google Maps and Booking.com), and both windows close at Q1 2026. Yet they describe two opposite failures: the first property is silent in front of its guests; the second talks a great deal and says nothing. The table below sets the figures published inside the two pages side by side.

Read side by side, they show what neither page shows alone: moving from a 2.0% to a 35.1% response rate did not move the second property into a healthy position, because only 4.1% of its replies were written for a specific guest about a specific problem. Response rate on its own — the single metric most commercial monitoring dashboards put on screen — can hand a management team a false sense of safety. And the rise in high operational risk flags from 6 to 23 is not a verdict that the second property is worse: its review base is more than four times larger and its window far longer, so what surfaced there would have stayed hidden had only two years been read.

What You Read Inside Each Study

Both pages follow the same structure so that they can be compared line for line:

• The full five-level star distribution before any average is quoted — because the shape says what the average hides
• Review tone split across four states: positive, negative, neutral, mixed
• The split between Google Maps and Booking.com, and what it says about which channel deserves priority
• Every complaint mapped to the operational department that owns it, instead of a flat "the guest was unhappy"
• Operational risk (safety, health, conduct) kept separate from routine dissatisfaction that improves gradually
• An explicit methodology section, and a "Study Limitations" section stating plainly what the numbers cannot prove

Seven minutes of reading for the first, eight for the second; neither assumes an analytical background.

How We Decide What Is Published and What Is Withheld

One rule governs it: publish what teaches the method, withhold anything that could lead back to the property. So these pages carry percentages and aggregate counts, and never a property name, a neighborhood, a specific city, a room rate, a star class, or a direct review quote. Direct quotes are withheld precisely because they are searchable word for word, which would re-identify the property and the guest who wrote them — and that guest signed nothing with us. Staff names appearing inside comments are stripped from every published output, even when the mention is praise. Study dates give the closing quarter of the analysis window, never the day. The same rule governs client reports: the report belongs to the property alone, and nothing from it is published without written consent and redaction. How data is handled is set out in the privacy policy.

Why Only Two Studies?

Because every published study needs explicit consent to appear even in anonymized form, and because reading hundreds of reviews line by line before a classification is accepted is slow by nature. Publishing ten quick studies would have been easier — and less honest. What is here is two real cases with their numbers exactly as they came out, including the ones that do not flatter: 47% of the second study's reviews carried no clear language marker, and that is written into the "Study Limitations" section on the page itself rather than buried. A third study is added when a case deserves publication and consent allows it.

Where to Read Next

To apply what the two studies show to your own property, start with the guide to responding to negative reviews; it turns the difference between a templated and a personalized reply into written steps. For the terms used here — bimodal distribution, templated reply, operational risk — see the glossary. For coverage in your city see service areas, and for practical questions on scope, timeline and pricing see the FAQ. For the full methodology, the service limits and what is explicitly not offered, see the methodology and founder page.