Text analysis for anyone

Gavagai Explorer

Example use case

To illustrate the capabilities of Gavagai Explorer, here is a before-and-after example of how to gain business insight into a set of hotel reviews.

Eager to get started?

  • Download the data in a CSV format.
  • (File => Download as => Comma-separated values (.csv, current sheet)).
  • Sign up for a free trial in Gavagai Explorer.
  • Explore the data yourself: create a new project and upload the example file.
  • Read a tutorial on how to get the most out of the Explorer.
  • Get started with your exploration!

Before: unstructured qualitative text data

Imagine you are the Customer Experience Manager of Hotel Monaco in SF. Instead of sifting through all the reviews your hotel gets, you download the most recent ones and use the Gavagai Explorer to find out what the guests of the hotel are actually talking about.

We have downloaded 130 reviews and made them available in this spreadsheet. Click the link and have a look at the data (it will open in a new window). As you can see, the spreadsheet contains five columns, of which we are interested in the one named ‘review’.

The question we are asking ourselves prior to exploring the data is:

When the guests of the hotel are writing their reviews, what topics are they touching on and how often? How do they feel about them?

Answering this question will allow us to understand in a very granular detail which small or large aspects and initiatives are driving guest satisfaction/dissatisfaction, guest recommendations, booking levels, pricing and ultimately the profitability of the hotel. This is the kind of insight that analysts can now yield on a large scale, with levels of quality and automation efficiency that have previously not been possible.

After: structured and quantified results in just a few minutes

Here are our findings after about ten minutes of exploring the data. The top three topics among the reviewers are:

  • 75% are concerned with the room, most of them are positive, and mention terms such as clean and comfortable in conjunction with the room.
  • 58% mention the staff, with associated terms being friendly and helpful.
  • 44% mention the restaurant, with associated terms being renovation and closed.

Already at this point, we know that there is an issue with the restaurant that is impacting on the hotel's satisfaction rates, repeat business and, ultimately, its profitability.

Looking further down the list of reviewers’ concerns, there are a number of positive as well as negative aspects to take into consideration:

  • 26% talk about the interior decoration of the hotel.
  • 12% mention the bed.
  • 24% speak of the renovation.
  • 16% talk about the noise resulting from it.

As Customer Experience Manager, there are already clear points emerging of what drives guest satisfaction. Decisions can quickly be made about what is important to invest in based on quantified customer insights.

Here is the PDF version of the report for our session with the explorer, and here is the corresponding spreadsheet data of the same findings.

Note that the reviews used in this example are from December 2014. Since then, the hotel has changed name to The Marker San Francisco, and the restaurant is up and running (click here for the most recent reviews on TripAdvisor).

Calculating Charges

The system is set up to charge for the total amount of rows, no matter the content (text, symbols, emojis, etc), or if there are empty rows. The reason that the system includes empty rows (and charges for them) is that even empty rows might be significant in the statistics and therefore we leave it up to each customer to prepare their data as they see fit before uploading to the Gavagai Explorer system.