How a leading enterprise used AI-driven NLP in Mobile App Reviews Analysis

In this case study, we show how a home security provider improved the ratings and reviews of its mobile app by leveraging advanced analytics in text mining.

Mobile applications are increasingly becoming a serious source of revenue for businesses. With hundreds of thousands of online apps and millions of smartphone users globally, app stores themselves contain an ever-growing amount of ratings and reviews. This alone makes the platforms a vital source to analyze and understand.

However, the dynamic customer sentiment and drivers behind app rating scores make it difficult to extract valuable insights.

Our client operates across multiple industries, including consumer electronics. As a technology-driven enterprise, the company has been featured on the Fortune 500 list and is well-known for providing innovative home security solutions.

One of their core products is a mobile application for home security control, which has thousands of reviews across various app stores. Like a TV remote control, the app allows customers to monitor home-protection alerts and change relevant settings with a push of a button, whenever and wherever. Both iOS and Android users can download the app.

Our client presented us with the opportunity to collect and analyze their growing stream of app ratings and reviews. The company noticed a slight decline in the average star ratings of their Home app.

Their marketing department sought the support of Wonderflow’s technology to investigate the situation further. Our team was mainly tasked with:

• Identifying the proper technical way to extract all relevant reviews from multiple app stores where the client sold their app
• Performing a deep analysis of the consumer feedback data to uncover accurate user sentiment and the drivers behind declining app ratings
• Performing a detailed run-down of all similar mobile apps and brands to compare the client’s product against them
• Report practical insights for the client’s various internal departments within a matter of weeks

Wonderflow set out to collect data across various app stores where the client sold their app. A stream of valuable new insights was uncovered from the massive database of reviews.

Our platform automatically analyzed the vast stream of structured and unstructured texts using a specific AI technique called natural language processing (NLP). The system was also able to classify each relevant feedback by sentiment type (neutral, positive, negative).

Specifically, we accomplished the following main tasks for the client:

• Aggregated and processed dozens of mobile app reviews
• Extracted relevant insights based on most-talked-about topics by users. In turn, our client discovered specific app features that users enjoyed and disliked.
• Performed an in-depth competitor analysis, comparing the client’s app against every other similar home solution app and brand, granting the client a comprehensive 360-view
• Translated the wealth of insights into a more easy-to-understand manner for our client’s various departments to easily access and share, plus all through a single platform

As a result of leveraging the Wonderflow platform, our client has designed various effective integrated strategies involving all stakeholders. They’ve also introduced a short feedback loop after every software update. Collecting and analyzing data across different app stores, including the client’s competitors, has also given our client a 360º view of the market and the real drivers behind low user satisfaction. In turn, the client was able to promote consistent knowledge throughout their entire organization.

 

Wonderflow respects the client’s wish to share limited information in this case study. To learn more about our technology, check out our solution pages tailored to those needing support in advanced Ratings & Reviews analysis.