At the MRS Data Analytics and Insight Conferences 2020, we shared how we helped Verizon speed up their Yahoo NPS analysis with Artificial Intelligence.

To acquire feedback on users’ experience, Yahoo / Verizon frequently asks an Advocacy question (NPS), which includes an open ended question. Traditionally, these open-ended questions in text format are analysed and categorized manually by insights teams supporting the marketing teams of each page or theme. The feedback is organized in topics and used to inform the theme owners about the users’ perception toward the web pages and on improvements that can be made.

When Verizon reached out to SKIM, we proposed a faster and more reliable process of feedback using Artificial Intelligence, allowing to shorten the time to insights for the marketing teams. The algorithm used inspects the structure of the language to understand what a given text is about. The so-called Syntactic Analysis breaks up the text into a series of sentences and obtains the topic of the sentence. In order to select the most relevant topics in collaboration with Verizon, the translation of the AI analysis into insights was proceeded by human supervision of the AI process.

In our presentation we shared how this innovative process allows for a much faster and more reliable feedback process from raw data to insights, and how it opens doors for Verizon to perform more analyses at bigger scale.

Date and time: Thursday, 27 Feb 2020 at 14.55
Location: Radisson Blu Edwardian London

Giacomo Sartori
Giacomo Sartori, Data Science Manager
Michael Hetherington, Director