Read your reviewsas a whole
Find recurring praise, criticism and unanswered feedback, with the original reviews behind each theme.
A star rating gives you a starting point. The review text tells you whether guests are talking about the welcome, the room, breakfast or a problem that remained unresolved. This tool brings those signals together so you can explore the experience guests describe.
How the review analysis works
Enter the hotel’s public website. We identify the property and look for its Google listing. When more than one listing could match, you check the name, address and website before continuing.
The tool requests up to the latest 100 Google reviews, ordered from newest to oldest. You see the collected sample before the text analysis finishes. Usable texts are read in independent blocks, so themes and sentiment appear progressively.
The dashboard lets you move from an overview to a theme, a subtopic and the original words behind it. You can filter the review feed and download the analysis as JSON, including its source, dates and completion status.
You can request the dashboard by email while the analysis runs. After the request is confirmed, you can close the page. We email the results as an HTML dashboard and JSON data, with any incomplete analysis clearly marked.
What the sample covers
The preview uses recent Google reviews. Its dates come from the records returned: the latest 100 might cover weeks, months or a longer period, depending on how often the hotel receives reviews.
The number collected is separate from the all-time total on the Google listing. Coverage is shown when that total is available. Reviews with only a rating stay in the sample; texts up to 20 characters are excluded from AI analysis. Every eligible text in the sample is processed.
The report shows exclusions, shortened texts and incomplete blocks. A failed block keeps the analysis partial and can be retried. This sample describes the reviews collected here, rather than every guest or every review platform.
Ratings, sentiment and themes
Google rating and Lumina sentiment answer different questions. The Google rating is the listing’s average on a five-star scale. Lumina sentiment is an AI interpretation of the themes discussed in the sample, on a ten-point scale.
A review may mention several themes and express a different tone for each. An overall positive review can still criticise the room or the check-in. Select a theme to inspect its subtopics, counts and exact excerpts.
A theme needs at least five mentions to show a score. Rating and sentiment trends use three-month moving averages with at least five supporting reviews per point. Missing points remain missing. Monthly volume counts the reviews in this collected sample.
Use the excerpts and links to assess the interpretation. AI classifications support your reading of the feedback; they are not independently verified facts about a stay.
Check public replies
See how many relevant analysed texts have a public reply from the hotel, and how many do not. The denominator is stated beside the percentage: rating-only, short, unprocessed and irrelevant texts are outside it.
Open the reviews with a reply, those without one, or negative and mixed texts without a reply. These selections help you inspect the feedback. This tool does not send replies or measure response times.
Continue with Lumina Analytics
This public tool previews one part of Analytics: reading a recent Google review sample. Inside Lumina, your team can work with Google, Booking.com and Tripadvisor, and compare the hotel with competitors.
Reviews also provide context for what the hotel communicates. A recurring question about breakfast, a praised spa experience or a room detail can help your team decide what its website and campaigns should explain more clearly.
Explore the other public tools to see how Lumina prepares hotel content from the hotel’s own material.
Questions about the review analysis tool
What do I need to start?
Your hotel’s public website. You can start without an account. The analysis runs on desktop or tablet; the example and this guide can be read on a phone.
Does the preview analyse every hotel review?
It requests up to the latest 100 Google reviews. Every eligible text in that sample is analysed. The dashboard shows the period covered and, when available, the total reviews on the listing.
Are Booking.com and Tripadvisor included?
This preview collects Google reviews. Booking.com and Tripadvisor analysis is available inside Lumina Analytics.
Why can a star rating differ from the sentiment?
The star rating belongs to the original review. Sentiment is an interpretation of its text and individual themes. A guest can give a high rating while mentioning a problem, or describe both positive and negative experiences.
Can I read the original reviews?
Yes. The report shows exact excerpts, the analysed text and links to the original Google reviews. You can combine theme, subtopic, sentiment, rating and reply filters.
What can I download?
A JSON file containing the analysis, sample, themes, reviews, dates, source and completion status. A partial analysis stays marked as partial in the file.
Google
Booking.com
Tripadvisor