decorative yellow lines on background

His focus on customer obsession was honed during his Netflix years as the company pivoted from DVDs to Streaming. In a riveting talk to kick-off Opticon19 Thursday, September 12th at 8:30 AM, he’ll share the secrets of consumer science he and his colleagues developed over those years and how you too can become customer-obsessed. 

He was kind enough to do a video interview from France where he was finishing up a speaking engagement. And with that, it’s time for this installment of ‘Ask an Experimenter!’

EA. TLC. Netflix. Chegg. What are you doing right now?

My purpose is all about teaching, coaching, mentoring and advising.

Right now I’m about 3 things – writing and giving talks and workshops all over the world.

What will you be talking about at Opticon19?

I’m going to share the personalization journey of Netflix. And it’s kind of fun because everybody knows it kind of worked out for Netflix. And within that journey, I’m going to bring the notion of customer obsession to life. I’ll also help people understand the balancing act of delighting customers in hard to copy margin enhancing ways. 

The fun part too is I’ll do a bunch of pop quizzes and ask folks in the room what they think the data will suggest. And often the answer is the exact opposite of what their intuition tells them to do. 

How would you define customer obsession?

Customer obsession means putting the customer at the center of everything you do, to figure out how to delight customers in hard to copy margin enhancing ways. If the Product Manager can really get consumer insight, which is incredibly hard, and bring the voice of the customer into the building and keep that voice alive, that is incredibly powerful to do. 

I’ve had two transitions. 

I went from satisfying customers to now I try to delight them. If you can spend most of your time delighting customers and do it in hard to copy ways that help to build the business, then you find yourself in this rare space where you don’t really feel like you are competing with anyone. 

The other transition is I used to be very competitive. Thinking about competitors is distracting. Thinking about your customer isn’t. You can’t overstate the power of that. 

What role does experimentation play in allowing companies to feed that customer obsession?

Basic scientific method is you form a hypothesis, you test it, you get results. The tricky part is sometimes you get results that say one thing yet when you apply human judgment and strategy you realize you should do something quite different.

How has the field of experimentation grown during your career?

It went from the basics of designing, executing and interpreting A/B tests into a world I can hardly fathom. Today’s it’s a world of machine learning – Netflix likely has 1200 classifications describing 1200 different movie taste profiles and there are 100s of algorithms being tested in parallel today. 

What is an experiment whose results shocked you? 

At Netflix, for many years on the non-member page — the page that helps folks sign up for a one month free trial, the more complicated page designs kept winning. Usually, you expect the simpler designs to win. In order to make it look like there was a lot there– to communicate value– we had to put a lot of stuff on the page. It took a long time for the Netflix brand to demonstrate the value instead of putting a lot of stuff on the page and we could finally execute the simpler, more elegant non-member page that you see today.

In your talks you mention things that are “fugly” but successful and the blend of presentation layer and algorithm work that fuels success – what’s your advice to Product Managers who are trying to balance these two camps?

The hard part of visual design is keeping stuff simple. If you keep stuff simple for a customer that tends to be best. But keeping things simple is hard. Stuff that looks good is typically simpler and easier. Generally, simpler and easier tends to win in an A/B test.

What does “0ut-experiment. Outperform,” mean to you?

Whether you have a successful or unsuccessful experiment – you are still learning. Generally those companies who can iterate faster will learn faster and those who can learn faster will outperform. 

We had 8 hypotheses about what would make Netflix a great product, and 3 of them succeeded. The success of Netflix is essentially high cadence failure. 

If you haven’t signed up yet for Opticon19- you can still get tickets here.