Big data gone bad

Big Data, Data Analytics, Artificial Intelligence were all hot topics at the 2018 NACUSO Network Conference. Credit Unions are sitting on piles of member data that has been largely untapped. When you think about all we could know about our members, it’s both exciting and frightening.

I remember the first time I heard about building a data warehouse.  I pictured a huge metal warehouse with towers of printed data (cuz this was in the 90’s when we still printed green bar reports).  I was the VP of Marketing at a very high tech credit union and our IT guy was determined to build a data warehouse. So I asked this question, “IF we could indeed have easy access to all of our member’s data … what are we going to do with it? In other words, what do we want to know?” He couldn’t really answer me at the time.

I think that’s the danger in having too much information. If you don’t know what you’re going to do with it or why you need it in the first place you could have your data geeks uncover a statistic that actually goes against your business model and could damage your reputation.

Nordstrom is synonymous with great customer service, mostly centered around their legendary no hassle return policy. I read the book the Nordstrom Way many years ago and it confirmed that the story was true about a Nordstrom employee taking back a set of snow tires – even though they don’t sell tires. The book also said that Nordstrom knows they have customers that “borrow clothes” but the loyal customers that pay full retail and don’t abuse the policy offset that cost. Nordstrom solved their own problem of returns by opening the Nordstrom Rack. Where you can shop Nordstrom fashions at a discounted price.

I had my own amazing Nordstrom experience years ago when I was dating a guy that wanted a red sweater for a Christmas party. Nordstrom did not have his size. But Nordy’s sales people will do everything they can to accommodate so he called around to other locations in the State. None of them had the sweater. The sales clerk took down his phone number and said he would continue his search and have an answer by the end of the day. Later that night he called and asked for my boyfriend’s address. He said he finally located the sweater and would gladly deliver it on his way home. When we answered the door he handed my boyfriend a Macy’s bag. They carried the same brand at the time and had a red one in his size. Amazing. Loyalty for life.

Enter big data, gone bad. A good friend of mine is a very loyal Nordstrom shopper. She’s the CEO of a credit union and dresses very professionally. She has a personal shopper, who will call her when new styles come in. She’s been served champagne on shopping trips. She’s a profitable promoter. The holy grail of retail. Until recently, that is. She also shops the Nordstrom Rack online. And some data geek started tracking the return rate of online shoppers. According to the scolding email she received her return rate was, in their opinion, higher than average. This is what they threatened to do:

“If your rate of returns is not lowered, we may have to deactivate your online shopping account. We value you as a customer, and want to do everything we can to ensure that doesn’t happen.

In order to improve your shopping experience, we have installed a number of new tools online, such as customer reviews and an improved sizing tool to help ensure a more accurate fit. We’re sorry if you’ve experienced issues with your past purchases, but we’re hopeful that this updated set of tools will help you avoid having to make as many returns in the future.”

What really kills me about this threat is they admit that they don’t have the best tools online to insure a perfect fit … but the data is there and a graph shows she is one of the biggest “offenders” of their return policy differentiator.

What is also sad about this situation, is Nordstrom’s did not recognize what a very loyal Nordstrom shopper my friend is, and how much she buys from Nordstrom’s separate from her online purchases from Nordstrom Rack, where they had a problem with her “above average returns.”  One clear thing we must do when using big data, is take the entire relationship into account. How long has the member been a member of your credit union? How extensive is their relationship, or has it been in the past? We need to be sure to recognize member loyalty, while finding ways to help our members and improve service and convenience with the big data and data analytics tools we now have available to us.