The 35 year old Huffy bike that I bought for $5 in 2002 finally broke:
It needed a new lower bracket, which costs $10 (and several hours) to replace. So I decided to buy a new version of the same bike. Huffy.com referred me to Target.com as the only place that would sell the same one, and after realizing that a new bike would only cost $148 (including taxes and free shipping to my door, worth $70), it took no time to decide.
Surveys waste our time
After taking just 1 minute to complete the e-commerce I volunteered to waste 10 more minutes on a 50-question customer satisfaction survey. It was 95% a waste of time because they didn’t ask the right questions. They used traditional survey techniques to codify buyer patterns and reduce the online buying experience into a set of variables with 3 to 5 possible answers. They asked such things as, “was the site layout helpful?” and “were you able to use our search feature to reduce the choices?” and “did you buy something that was only available in our stores?” and “how likely are you to shop online at target.com again?”
The one question they should have asked but didn’t was:
What was the main reason you bought the product?
And this should have been open text. I would have written the whole story very briefly:
I wanted a bike that only target sells. I followed a linked from Huffy, the manufacturer. You offered free shipping and a reasonable price.
And if there was a second question, it would have captured even more of the story without the 50 questions:
Q: We noticed your order has a special discount of $70. How much did this offer affect your purchasing decision?
A: A lot. That was a 40% discount, and I get the bike delivered to my door in two days.
Like a lot of complex data “systems” – people focus on the wrong questions and discount the emormous value of meta data.
- Target’s tech department already knew that I came from Huffy.com.
- Their tech team also knew how much of a discount they offered me (on my receipt).
- But the marketing department didn’t get this information. They didn’t use it. They didn’t ask what impact my being referred from Huffy and offering the discount had on my behavior. Instead they played 50 questions.
The real story was lost in the cracks in between two departments inside Target corporation. They wasted my time and missed the story.
Big data is all about making connections among the data you already have. I am spending this year on a storytelling project in the hopes that do-gooder organizations can act smarter than Target by starting with the obvious open-ended questions, mashing in existing knowledge, and learning more, faster.
Read an older story about “the bike“.