Leveraging Text Intelligence to understand Drivers of Customer (Un)Happiness
Imagine that you are responsible for a product. Over a period of time, your customers have provided feedback in the form of ratings and written reviews. The figure below shows how the reviews are distributed across the different ratings.
Now, from the figure it appears that most of your customers are happy, that is, have rated the product 4 or 5 on a 1 – 5 scale. It is also clear that a substantial number of customers are not happy at all and have thus rated the product 1 or 2. In between these extremes, there are a number of neutral customers.
As a smart person responsible for the product, you wonder about what drives the customers in the respective groups. In particular, you ask yourself the questions:
- Why are the unhappy customers unhappy? The answer to the question will provide clues of what to improve in future generations of the product, so as to gain new customers, and perhaps even win over dissatisfied ones.
- What are the neutral customers talking about? The answer to this question will help you mitigate the churn that you worry about from the non-loyal (aka neutral) customers.
- What strengths of the product are advocated by the happy customers? The answer can help you find successful features of the product to stress in future marketing campaigns. Perhaps a better communication around the product’s strength can take the focus away from its perceived weaknesses?
This post is shows how the Gavagai Explorer can be used to elicit the answers to the above questions from the review data at hand.
A good place to get product reviews is via the BestBuy API. We arbitrarily chose a product from BestBuy’s catalogue that had been reviewed more than 1000 times, and received an average rating fairly close to the middle of the rating scale: Misfit – Flash Activity Tracker. We downloaded all 1419 reviews available as of March 19, 2016. The distribution of the ratings is the one shown in the figure above, and the average rating of the Flash Activity Tracker is just below 3.4.
By using the Gavagai Explorer, we were able to analyze the 1400+ reviews of the Flash Activity Tracker in less than 30 minutes, and thus come up with answers to the questions Why are the unhappy customers unhappy?, What are the neutral customers talking about? and What strengths of the product are advocated by the happy customers?
Why are the unhappy customers unhappy?
To answer this question, we assessed the major themes related to features or customer experiences present in the reviews written by the unhappy customers. The red bars at the top in the above figure represent the top-5 drivers of these customers. It turns out that over 53% of the customers talk about how the device broke or otherwise stopped working. What’s more, the Explorer allows us to drill down to the actual data and obtain verbatim examples from the reviews under scrutiny:
- “Well, she only got 3 days of usage out of it before it stopped working all together.”
- “It only worked for a couple of days then just quit.”
- “This was a good product for the two days it actually functioned.”
Just over 40% of the customers in this group talk about the quality of the wristband. There is a large overlap between this theme and the above one, although this pertains more specifically to the wristband. Examples:
- “I really like this product but the wristband is a piece of junk.”
- “The first one was lost when the band broke, otherwise it worked fine.”
- “I have tried 2 different ones both broke within a few days and both watch bands cracked immediately. “
13% of the reviews touch on how the device syncs with external hardware, which is closely related to the app, a subject matter concerning just over 10% of the unhappy customers. Examples:
- “Also, syncing it with my phone, can be challenging.”
- “feels cheap and hard to connect to the app and sync”
- “They FINALLY updated the Android app so it’s comparable to the iOS app, but it crashes every other time I try to open it.”
Finally, the fifth theme represented in approximately 12% of the reviews is about the device’s ability to track sleep. Example:
- “readings were inaccurate esp. sleep tracking.”
- “It often logged me as sleeping when I was actually up and moving around.”
- “It keeps saying I slept an hour longer than I did every day.”
Unhappy customers are primarily concerned with the quality of the product. The above is a priority list of the top five things that require fixing to reduce future negative reviews, and likely loss of customers. Note that the two most prominent themes, that essentially have to do with the same thing, represent a very large portion of the reviews. As we shall see, this is not the case for the other two groups of customers.
What are the neutral customers talking about?
The top-5 driving factors among the neutral customers are shown as the yellow bars in the figure above. There are several differences worth noting between the unhappy group of customers, and the neutral ones. First, the largest themes in the reviews are not nearly as dominating as the corresponding themes are in the case of unhappy customers. While the neutral customers are still concerned with the wristband and the breaking of the device, 34% and close to 26% respectively, the themes have switched order compared to among the unhappy customers. Looking at examples from the reviews give the sense that the neutral customers are not as negative in their feelings in relation to the themes:
- “Only issue is the band and misfit device doesn’t fir [sic] properly and breaks the band”
- “Good tracker for the price wristband needs improving.”
- “Just be careful with the band.”
A new theme made the top-5: just over 20% of the neutral customers talk about the price. Examples:
- “Good for Price.”
- “Considering the price this is a fine entry level tracker.”
- “Simple easy to operate and low cost.”
The fourth largest theme is the app covered by approximately 15% of the reviews. Examples:
- “The actual tracker and accompanying app were really nice.”
- “IPhone app works great and I could pair app with device in less than 5 minutes.”
- “The App itself could be easier to figure out and get around in, such as figuring out how many steps I want to attain in a day.”
Finally, the sleep tracking abilities of the Misfit is covered by close to 14% of the customers. Examples:
- “It was a nice tracker, but I was a bit unsure how the sleep quality is tracked when there is no way to change it to sleep mode on the device itself, only in the app I found it.”
- “I especially enjoyed the sleep tracker.”
- “Not the most accurate sleep tracker.”
While the neutral customers are concerned with the quality of the product, they are less so than the unhappy customers. In addition, the neutral group brought up the device’s low price as something very positive.
What strengths of the product are advocated by the happy customers?
The final group of customers, the happy ones, is represented by the green bars in the figure above. Compared to the previous two groups of customers, the top-5 themes present in the reviews written by happy users are evenly distributed: there is no single biggest thing that warrant your attention. Furthermore, even though the themes expressed in this group is largely overlapping with that of the neutral customers, the contents of the themes is more positive.
The device’s sleep tracking abilities are brought in 24% of the reviews. Examples:
- “Sleep tracking is surprisingly good.”
- “My only complaint is that I’ve napped during the day a few times and it doesn’t log that.”
- “Good product for anyone beginning the process of tracking their steps and how long they are resting.”
Almost as many reviews, 23.5%, talk about the price of the device. This is clearly a selling point! Examples:
- “Great fitness band for the price!!”
- “Very good price too.”
- “Great Watch For The Price.”
Again, almost as many as in the previous group, 23%, talk about its ease of use. This is the new theme on the block. Examples:
- “Great, easy to use product.”
- “Simple to use, accurate..”
- “This is a quick and painless tracker!”
The black sheep of the family remain in top-5; 14% talk about the problems with the wristband. Examples:
- “The only reason it lost a star is because the band broke and the device seems like it will fall out.”
- “The wristband broke within a day but I used the clothing clip with no problem.”
- “All the bands that break, break in the same general spot.”
Finally, in the fifth place with 13% of the reviews covering it, the app remains a theme of concern. Examples:
- “This product has great app on Google play store”
- “Works great with the Weight Watchers App but now they are having a problem with connection.”
- “Love this tracker, the app is very nice in iOS.”
The reviews originating from the happy group of customers highlight two positive features of the product that can be used for future communication: its price, and its ease of use.
This short exposé into the 1400+ reviews of the Misfit – Flash Activity Tracker has brought our fictitious product manager up to speed with what he has to do:
- For the unhappy customers – improve the quality of the product. This can be achieved in several different ways, e.g., by improving the quality assessment at the assembly factory, using different materials and toolings, or simply by providing a sense of a better quality and ship extra wristbands, free of charge, with each purchase.
- To retain the neutral customers, make sure the sense of quality is improved, emphasize the device’s low price in marketing, and make sure to roll out frequent updates of the app for each platform. If it isn’t already, consider making the app open source and invite the community to contribute.
- The happy customers are already happy, but you can learn from them to change your marketing. The device’s low price and its ease of use are key selling points, as the testimonials of your most satisfied customers show.
About Gavagai Explorer
Gavagai Explorer enables individual analysts to single-handedly process amounts of data that would otherwise require tens or hundreds of analysts doing manual analysis.
Gavagai Explorer turns qualitative and unstructured text into quantitative measures by automatically identifying and ranking common themes, detecting associative expressions significant for each theme, as well as by scoring themes against multiple dimensions of sentiment, such as positivity, negativity, and skepticism.
Gavagai Explorer is the Text Intelligence tool of choice for analysts who want to gain rapid insights into large text collections, such as answers to open-ended survey questions, NPS follow-up investigations, output from Customer Experience touchpoints, product reviews, and social media discussions around brands. The Explorer relies on the vast language knowledge continuously learned by Gavagai’s Semantic Memories, currently available in 20 languages. Gavagai Explorer is available at https://explorer.gavagai.se