Unleashing buried data to meet customers’ unique needs

Our client is a large national eyeglass wear retailer with more than 700 locations and 8,000 employees across the United States. This client needed to better understand its customer base and identify niche opportunities to increase customer engagement and drive profitable revenue.

Client challenge

Our client’s marketing team needed help obtaining and making sense of transaction data from its customer base; however, this data was not easily accessed. To access and use meaningful data, the team needed to solve two significant problems – each at an opposite end of the data and information spectrum.

At one end, the client’s IT team had one large customer database backup file that contained more than seven years’ worth of data, but the file and data were unusable by the marketing team. The data was difficult to access, and extracting the data was beyond the marketing team’s existing skillset. Even if team members could access the data, they could not use it.

At the other end of the spectrum, the client had recently completed a project with a Big 5 firm in order to glean more meaningful insights from its data. However, this firm delivered a decile analysis that did not provide the level of detail the client needed; the segmentation was so high-level that it did not deliver any actionable insights on segment behavioral patterns to inform the team’s tactics. Even after a significant project investment, this team still did not have a strong understanding of the differences in customer segments to help target their marketing efforts.

The marketing team was left without much direction as to how best to engage with its customers. Caught between too much detail with no summary and a summary with no detail, this team had no way of bridging the gap. To solve this dilemma, our client chose to invest in a secondary project with Turnberry.

Our solution

Our client worked with Turnberry to develop a comprehensive solution that provided a strategic overview of its customer base – along with the necessary details for each segment – to differentiate marketing tactics and messages based on the needs of the different segments. The client needed Turnberry’s help in two key areas: accessing usable data and developing meaningful customer segmentation.

This team had seven years of data stored in an unusable backup file. Our team was able to reverse engineer the backup file into component parts, and then design and calculate additional features based on customers’ historical behavioral patterns for use in segmentation development.

After ensuring the client could access its useful data, we developed a customer segmentation of its 14 million customers into six distinct segments that would still be manageable operationally. Turnberry consultants used Mini Batch K-Means and Birch methods to create these customer segments. We then delivered a demographic and behavioral profile of each segment, along with high-level key performance indicators (KPIs) and marketing opportunity recommendations.

Turnberry was able to leverage our deep knowledge in data science and advanced analytics, along with our expertise in data engineering, to extract the maximum value from what had been an aging and unusable data repository.