Zappos Labs’ Will Young finds silver lining in feature failures
August 27, 2013 | Russ Martin | Comments
As director of Zappos Labs, Will Young heads a team that runs experiments in e-commerce.
Housed in its own office separate from the e-tailer’s main San Francisco digs, Young and his nimble team of 11 test new ideas to see what will pick up traction with Zappos customers.
For every idea that sticks, there are dozens of misses—big ideas that didn’t pan out once tested on consumers. By running A/B tests in which some Zappos.com visitors see new features and others don’t, Young and his team are able to see what ideas drive engagement and sales.
If a feature fails to achieve its intended output, Zappos kills it. But even those disappointments are useful for teaching the retailer more about its customers. Without testing new ideas, Zappos would risk missing out on the next big thing in online shopping, says Young.
“You never know what’s going to hit,” Young says. “Take a big macro trend like mobile—obviously it’s super important—but even four or five years ago we weren’t positive that people were going to shop on their phones. You have to explore things early and capitalize on trends before you’re left behind.”
In the lead-up to Vancouver’s Grow conference earlier this month, where he spoke about promoting innovation within companies, Young led Marketing through four big initiatives Zappos shelved, why they didn’t work out and what the company learned from them.
First off, tell us what Zappos Labs does.
“We’re a small team of 11 people of a 1,300-person company. A little research and development team within Zappos. Our mission is to explore new technologies.
What’s the advantage of having a space away from the rest of the company?
“It’s good to pull yourself out of the work environment. To create a space to not get pulled into day-to-day fires. Sometimes people create innovation teams that sit right beside what used to be their day job and they will keep getting sucked back in. ‘Oh no, the search page is broken. We need your help on it.’ If you create a different physical space, that helps.”
Lead us through your process. How do you test a potential product?
“The majority of our projects are launched as a microsite, a separate site where we can have our own space, somewhere we can do a lot of quick changes without breaking the core check out. We also have a customer list who have said they love to try new stuff. We use them to get an initial spike in traffic and then we’ll promote microsites on the homepage or on landing pages.
We’ll usually run an A/B test [a kind of experiment in which one group is affected and another isn’t] where 1% of our customers see it to make sure it doesn’t break. Then we’d go from 1% to 50% to see whether the feature should keep living on the website.”
What’s it like to report back to management when an idea flops?
“We’re lucky because Zappos leadership really treats our team as research and development and they don’t necessarily say, ‘You need to generate x million in sales.’ Through that lens of research and development, it’s easier to report back. When I report to management on a project that didn’t work, they might challenge how much time we spent to figure out it didn’t work. Those are the discussions, not just ‘Why did you build that?’ but ‘Could you have learned that faster?’”