Few brands are as poised to take advantage of advertising automation as Netflix. The company was founded on the idea of managing and monetizing the long tail of movie titles in its vast library by using behavioral data to make movie recommendations to individual subscribers. Building rich datasets, and intelligently deploying them to drive their business, is foundational to the way Netflix operates.
Eight years after its founding, automation is also core to the way Netflix markets itself, and Netflix is one of a handful of sophisticated brands that have taken their advertising automation practice entirely in-house.
Rubicon Project’s SVP of Marketplace Development, Jay Sears, recently discussed the preconditions for bringing a sophisticated advertising automation program inhouse with Netflix’ Global Director of Programmatic Marketing, Mike Zeman. In a nod to Netflix’ movie roots, Sears laid out four models for companies’ relationships to ad automation management, using movie titles as metaphors.
The Great Train Robbery
“It’s all very good…unless something goes wrong.”
This is the pay-for-performance model. Here, advertisers pay their agency for a specific business outcome at a fixed cost. The actual costs of the media, data, and other elements needed to achieve that business outcome (leads, sales, app downloads, etc.) are opaque to the advertiser. Agencies take their margin based on the spread between their actual costs and that contracted fixed cost, and they win by delivering the desired business outcome to the advertiser while keeping their costs as low as possible.
“Show me the money!”
In this scenario, advertisers want a “visible line of sight” into how much of their budgets are spent on working media – the portion that actually winds up in the hands of media owners – and how much goes to other fees. The agency takes the lead on strategy and execution, but in addition to performance, insight into what elements comprised their campaign are one of the outputs brands expect from campaigns run in this model. Here, the watchword is “visibility”: the advertiser still outsources the work to the agency, but expects insights into the mechanisms that made their campaigns work (or not).
“It’s a little more complicated these days.”
Brands that work with their agencies per this model are advertising automation sophisticates. They go to market with great clarity about their goals in terms of media types, advertising formats, data, and other aspects of campaign setup, but they still rely primarily on their agencies to construct the campaigns, and to execute the strategy around those requirements. In essence, the brand is the brains behind the automation strategy here; the agency is the arms and legs. Brands that understand the landscape well enough to truly direct their ad automation strategy, but that lack the technical infrastructure and human resources to do the heavy lifting on execution, might find themselves adopting this very hands-on working relationship with their agencies.
“New…powerful…hooked into everything, trusted to run it all.”
Here, the advertiser operates their automated campaigns inhouse, as Netflix does. “When I came to Netflix, we were still in the Jurassic period of outsourcing buying decisions,” he said. The two factors that drove them to bring the decision-making and execution in-house? First, Netflix wanted the “intellectual capital” and insights that automation is uniquely able to provide. As a company whose core value proposition was primarily about its data-driven recommendations (at least until recently, when quality original programming took center stage), this was core to what Netflix valued in general.
And secondly, Netflix had the assets – the “tools, tech, and talent,” as Zeman put it – to do advertising automation well. “We have a large tech team that sits between engineering and marketing, and helps business requirements happen,” he said.