Designing with Data by Rochelle King

Author:Rochelle King
Language: eng
Format: epub
Publisher: O'Reilly Media
Published: 2017-11-09T00:00:00+00:00

Designing to extremes to learn about your users

Eric Colson had a great example from Stitch Fix where they evaluate the feedback from “polarizing styles” to better understand their customers’ preferences:

We have a notion of a “polarizing style.” This is a piece of merchandise that clients tend to either love or hate. There’s no one in-between; no one just “likes it a little”—it’s one extreme or the other. We detect which styles are polarizing by studying the feedback data on our styles and applying measures of entropy. It’s hard to predict a priori which styles will be polarizing. It’s also hard to know which clients will be on the love side or hate side for a particular style. But after the fact, it’s obvious—they explicitly tell us. And this can be useful information. Each client’s response to a polarizing style reveals information about their preferences. The people who hate the style tend to share certain preferences. Likewise, people who love the style share similar preferences. So, regardless of whether or not the client loved or hated the polarizing style we now know more about them. This means we can even better serve them on their next fix. It’s about setting yourself up for long-term success by continually learning more about your clients’ preferences.

It reminds me of what I learned about baseball during the 2014 playoff series. I am a passive fan of the San Francisco Giants. But of course I get more interested during the playoffs. I happened to watch a game with a friend who was much more of an expert on baseball. It was not looking good for the Giants. The opposing team’s pitcher was on fire. He had the Giants shut down—no one was hitting off him. But the tide turned after a “great” at-bat by the Giants. I didn’t initially see it as “great”—he had struck out! Yet as he returned to the dugout, all the players were high-fiving him. I turned to my friend and said, “He struck out. Why the high-fives?” He explained to me that it was a productive at-bat: he took up nine pitches (pitchers are only good for 100 or so). And each of those pitches revealed information to the next set of batters watching from the dugout. That was nine more times that they got to learn from watching the pitcher’s timing. They got to see his slider, his fastball, the curve—even the sequencing of pitches is valuable to learn. So he revealed information that would benefit the next batter. It felt very similar to the value we get from our clients’ feedback to various styles; we learn for the next time.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.