Twitch Plays Hard to Get with Advertisers

by Michael Dowd
Jun 14, 2017

For three days in May, tens of thousands of people crowdsourced the control of a giant, animatronic squid. Eight independently-controlled arms struggled to work in concert as it attempted to run a pizza parlor and man a production line. And it was all part of a brilliant creative campaign by Old Spice. How in the world did marketing get to this point?

This story starts with a platform called Twitch, though similar stories can be found for Snapchat, Reddit, Facebook Live, and a handful of other youth-dominated media. Twitch is often labeled as “people watching people play video games.” While this is functionally true, it demonstrates a shallow understanding of the 100M user platform that Amazon purchased for a billion dollars in 2014.

Twitch is about communities, united by a shared interest in a particular video game or “broadcaster.” Its value comes as much from interacting with the broadcaster and other like-minded members as it does from the gameplay itself. And those members communicate through chatrooms in a language that is thoroughly and deliberately obtuse to outsiders.

In 2014, a Twitch broadcaster built a system into his chatroom that parsed through comments and translated them into crowdsourced player movements in the video game Pokémon Red. After 16 days and the collective contribution of 1.16 million community members, they beat the game. The idea became a phenomenon, and it spawned several dedicated channels used for social experiments, including the currently-active StockStream, in which a user is letting Twitch crowdsource the allocation of a real $50,000 stock portfolio (it’s up about $2,300 or 4.6% after a week and a half, compared to 1.0% for the S&P 500).


While there are native advertising opportunities within Twitch, using them without proper context can expose brands to the wrath of a skeptical and extremely vocal community. Last summer, Bomb Pop ran a pre-roll ad across the Twitch network, presumably inspired by the thought “the kids are on Twitch nowadays, and kids like Bomb Pops.” The self-unaware, 30-second spot of children on a playground stood out like sore thumb against a backdrop of action-packed video game trailers and energy drinks, and the community predictably sighed.

The advertising world is very good at buying ads based on demographic and behavioral profiles. In a few years, 84% of display advertising will be bought programmatically because it yields proven results. Yet my concern is that we are optimizing toward an artificially constrained inventory at the expense of exploring new platforms and marketing opportunities that are not so easily manipulated and which represent the communities of choice for a new generation.

I understand how we fell into this trap. We as marketers preach to our clients about disruption and the need for digital transformation, not realizing that we ourselves are at risk of being disrupted by focusing on short-term performance over longer-term innovation. Allocating budgets to DSPs that can buy programmatic inventory is safe and effective; doing something groundbreaking is risky and uncertain.


When Old Spice first started to explore Twitch, it was a welcome relief from teed-up gaming campaigns and facepalm-worthy demographic ads. It was clear that they had taken the time not only to look at the data about Twitch, but to understand how its community operates. They understood that Twitch is at its finest when its users can rally behind a cause because they had seen Twitch Plays in action.

So in 2015, Old Spice dropped a real person in the middle of a forest and had the community determine his – admittedly, heavily scripted – actions for three days. You can read the articles about it on Ad Week or Digiday, but those are predictably myopic, declaring it as a win for targeting an elusive demographic. If you really want to know how and why it worked, I recommend this post from the notoriously brand-unfriendly Reddit.

Then they followed it up two years later with a giant, octopodal robot. It was ridiculous theatre. It really didn’t work, in the strictest sense. Food flew everywhere, tentacles had literal minds of their own. But it was also glorious, and it deserves respect from the marketing community. Not the kind of respect that comes from admiring their ability to get in front of Millennials, but the kind that comes from their continued resolve to opt for audience-relevant creativity at a time when automation is so readily available.

Keep buying your programmatic display ads. They work really, really well. But take a cue from P&G and Old Spice and spend some time immersing yourselves in the platforms of the future, where modern ad strategies fall flat, and where true, creative-driven marketing can shine once again.

Photo Credit: Dan Silva

Michael Dowd is the Executive Director of Digital Strategy at Maark, where he coordinates new product development and execution for our clients. Mike brings with him eleven years of experience in agency-side digital marketing, during which he provided guidance on marketing technologies, platforms, and strategies for more than a dozen Fortune 100 companies across the B2B, retail, and automotive industries.

PRAM Organization Aims to Guide the New Privacy-First Digital Ad Landscape
The Partnership for Responsible Addressable Media is made up of advertising groups, tech platforms, and advertisers. But what will it do?
What’s in a Name? Your Entire Brand
Alexandra Watkins’ book, "Hello, My Name Is Awesome" turns the highly subjective task of naming a brand into a more objective exercise.
First-Party Data is More Valuable Anyway
Third-party cookies are disappearing, but marketers could always do far more and far better things with their own data.