“Voice of the Customer” initiative that gathered user sentiment data across multiple interaction points to provide a cross-sectional view of user experience with Bentley products. The program combined input from four sources: executive interviews capturing Net Promoter Scores (NPS) and qualitative feedback; digital surveys offering periodic sentiment trends and open-text analysis; transactional Customer Effort Score (CES) surveys assessing ease of interaction; and G2.com product reviews providing product-level sentiment insights.

I implemented the system through a series of workflows to automate survey distribution, data retrieval, and sentiment processing.

  • Orchestrated by Azure Data Factory pipelines to initiate sending CES surveys, and retrieve and process various sentiment data.

  • The pipelines interacted with Durable Azure Functions in Python through HTTP to initiate and retrieve Qualtrics surveys and to download G2 reviews using REST APIs.

  • The various customer feedback streams were processed with SQL to form a multi-component user sentiment index for use in a Voice-of-the-Customer dashboard in QlikSense.

flowchart image

These data streams collectively formed the basis for a multi-component sentiment index displayed in a QlikSense dashboard for product teams to track and act on user sentiment.

Techniques for generating continuous date ranges, pivoting denormalized data, and computing inline rolling averages are shown in this SQL example. This Python example illustrates an Azure Function that retrieves and flattens Qualtrics survey data for ingestion into the sentiment analytics pipeline.