Database Reference
In-Depth Information
doing a lot great work especially in the area of product recommendations.
Flash sale sites have extremely high product turnover so they're dealing with
the cold start problem for every product every day.
Gutierrez:
Are there other types of data that you think are ripe for plucking
given that you have the big department store model?
Shellman:
There's lots of exciting new data coming from our recent acquisi-
tions, HauteLook and Trunk Club. It's smart that Nordstrom recognizes that
there isn't one type of customer and provides different shopping experiences
through the full-line stores, the Rack, HauteLook, and Trunk Club. It's going to
be really interesting to see how customers interact with all these channels and
it's going to pose new challenges. Measuring behavior across all those channels
will be difficult, but lends itself to so many opportunities to cross-promote.
The most interesting types are data are those collected for one purpose
and used for another. For example, one of our developers, Jason Wilson, had
a really cool idea to look at what was purchased when people asked for
gift receipts. Then you could make recommendations for the most gifted
products for an upcoming holiday. It's an approach in the same spirit as Our
Stylists Suggest. Just let people do what they're doing and use that behavior to
enhance our services.
Gutierrez:
How could someone starting out learn more about the fashion
retail industry from a data science point of view?
Shellman:
Conferences are a great place to learn and meet people working
on similar problems. Twitter, honestly, is a really great resource. It's really easy
to find and follow people from Rent the Runway, Warby Parker, Zulily, and
other companies to see what they're up to and what tools and technologies
they're currently into. Another great place is company blogs. MailChimp and
Etsy have really great blogs.
I would also suggest that when you're in a conversation and have the opportu-
nity to learn, ask lots of questions and don't worry about if you're perceived as
not knowing what's going on. Then ask follow-up questions. A good medium
for this type of exchange is emails. I'm not against cold emails at all. Don't
worry about not knowing the language. Coming from biomedical sciences,
I didn't have any experience in retail, so there was a lot of stuff that I didn't
know and I just asked people to explain it to me. If you go into every conver-
sation with that attitude, people will tell you a lot of stuff, and it's just your
job to absorb it.
