Get your own Little Black Book.

Build your own personal news stream. Discover the latest work created that interests you, share your favourite stories and follow your favourite people and companies

Already have an account?

Opinion and Insight

Huge's Jon Gibs on Why This Year's SXSW Was the Most Relevant Yet

Huge, 2 years, 7 months ago

VP Data Science & Analytics at Huge discusses three experiences that stood out for him at the festival

Huge's Jon Gibs on Why This Year's SXSW Was the Most Relevant Yet

When I typically come to SXSW, my immediate feeling upon landing in Austin is: “this won’t end well.” The combination of beer, tacos, BBQ and shiny new digital things is a pretty deadly combination for me. Most of the time, the shiny stuff is interesting – but rather irrelevant to my day-to-day work or my clients. It’s fun, educational and a good way to meet friends who share a common love for new trends in interactive.

This year was different. There weren’t a lot of shiny new things. No new startup stood out. The parties were crowded and the streets were rowdy, but the tech itself was relatively unimpressive.

It was probably the most relevant SXSW I’ve been to.  

My area of interest is the intersection of data and creativity – how data can inform the creative process, making better, more seamless, experiences for users. In this vein, there is a specific idea that we’ve been working on called Anticipatory Design. The core of this idea is that user experiences across all channels should foresee user needs and adapt accordingly. The problem is, I entered SXSW with no idea what this really meant. I wanted to basically use the four days as one giant workshop. I wanted to bounce ideas around with smart people and see what stuck. I wanted incorporate new ideas and codify my thoughts into a cohesive approach.

Rather than going to whatever session had the most cliché-ridden, click-bait worthy title (The Optimized Big Data Internet of Things) or wait in long lines for the most desirable parties, I used my need to work out this concept as my guide.  I only went to sessions, parties and meet-ups that seemed relevant to this topic. My partner in crime, Sophie Kleber (our Executive Director of Product Innovation) and I, were on a mission. Sophie took the design part and I took the mathy bits. Each day, we would go to sessions and every night we’d come back to our rented bar, Midnight Cowboy, to hash our ideas, have a few (or more) cocktails and just generally work it out.

There were three experiences that really stood out for me and influenced my approach to both the nature of the idea and our eventual solution to the problem.

1. The 9am, Tuesday morning session by Netflix about their learnings from 10 years of testing, drove two core ideas home, a users behavior tell more about them than almost any other data point.  They walked through a series of tests and how their results have influenced the long-term product roadmap.  They described their one time attempt to allow users to rate movies with half stars, which resulted in users actually rating fewer movies.   Next they discussed how the demographics they collect are more or less useless for predicting what movies people will watch then their behavior.  Finally there was the biggest insight of the talk, that user ratings don’t have a particularly correlation to what movies people will watch in the future.  Movie ratings are aspirational, people rate highly what they think is supposed to be good, rather than what they actually like.  In essence, behavior trumps all other factors in predicting what people will like.

2. A few drinks with Dries Buytaert, the creator/project lead of Drupal and the co-founder and Acquia. We spent most of the time discussing their new-ish testing/personalisation platform Lift. I’m yet to use Lift, so I don’t have a specific option on the product, but two ideas we discussed really stood out.  That there could be a testing/personalisation platform that could be built in an open source framework – a platform that was adaptable to all data sources.  Indeed, the business process for content development can be as important as the technology itself.  We pushed on the idea sometimes a truly adaptive experience requires a change to a more adaptive business process.

3. An hour and a half breakfast chat with Katherine Bell, the editor of HBR.org.  We were prepping for the capstone of the four days, a backyard idea throw down that me, Katherine and Sophie would have later in the evening in along side 30 or so of our favorite people. We talked about the growing need to develop the right content for all types of users. That the effectiveness of content shouldn’t be based on the number of clicks an article gets, but rather that those clicks need to be taken in context relative to other similar content. Whither click-bait. Personalisation means knowing your users and anticipating their needs not just making pretty power point slides with persona descriptions on them.

The four days wrapped up with a lively discussion on Tuesday about the idea of Anticipatory Design in the backyard of Midnight Cowboy. It was an open discussion with 30 or so smart people from a range of backgrounds, from data science, to planning, to creative, to UX, to ad-tech. Sophie and Kristen discussed what we had learned during the week.  We drank cocktails, bounced ideas off each other, off the audience and even some folks from competitive agencies.  I’m not quite sure we figured everything out, but we were a lot closer than we started. 

SXSW can be a giant waste of time and a boondoggle. This time wasn’t. Having a specific idea to workshop allowed us to take advantage of the interdisciplinary nature of the show and ultimately get a ton of value out of it, even if there weren’t any shiny objects.

Genre: Strategy/Insight