Skip to main content

Retailers should use more data to improve their customer relationships

Rachel Wheeler Archive

Retailers are already well aware that data analytics can help them improve their marketing strategies and attract new customers. By analyzing their target demographics, companies can find better ways to entice consumers. The more challenging endeavor, though, is improving the relationships retailers have with existing customers.

In theory, it makes sense that stores should reach out to their loyal patrons by offering them deals, discounts and personalized recommendations that fit their needs. In reality, it rarely works. Coming up with specific ideas for each customer is a difficult task - retailers have too many customers to address each one in person, and analytics don't yet have the precision to find good answers.

Retail Week recently analyzed this problem, citing Amazon as an example of a retailer that hasn't yet figured things out. The site tries to recommend products that, according to data analysis, the customer will appreciate, but often the tips are too vague or inaccurate.

The phrase everyone's using is "single-customer view." Grouping customers into crude categories based on age, race and income level isn't enough - people demand individualized attention. It's unknown whether that ideal will ever become a reality, even with the increased power of data analysis, as Deloitte partner Richard Small explained to Retail Week.

"A single-customer view has been on the 'too hard to do' list for a long time," Small said. "My first conversation about this with a brand was in 1998. Each project pointing in that direction is individually valuable. It's pointing toward something that might never happen, but if you get to 80 percent, you are well ahead of the competition."

Retailers still have a lot of work to do before they can give customers the personal attention they demand. Gathering more information and improving data quality are positive steps, but more innovation is still necessary.

Comments