Automating the Ad Agencies

• Tweet
• Buffer

I spend a lot of my time with ad agency folks. And two recent conversations with senior people have struck me.

The first conversation was with a digital agency managing director in the U.K. We were discussing advertising's future, and he said:

Agencies don't spend nearly enough time strategizing for ourselves. We purport to be the strategists for our customers, but there is enough of a gray area in all the things we touch that even in online, there isn't a huge amount of accountability. We suggest a strategy, we decide how it will be measured, and we get buyoff from the customer. If it doesn't seem to be working as the campaign rolls out, we shift strategies and attack from another direction.

We try to keep our fingers on the pulse of advertising innovation on behalf of our customers -- but we don't hold our finger to the wind on our own behalf. We don't invest in research or strategies that let us evolve or be particularly proactive. Mostly this is because everyone at an agency is focused on winning, keeping and improving business for the next quarter's revenue. We have no idea how we'll differentiate from competitors in five years -- because we don't have the time to think about our own business.

I've had plenty of conversations with smart, even brilliant, people at ad agencies over the past few years that were very forward-leaning. And I've had many great discussions about the market's future . But this managing director was right: almost all these conversations focused on how to help their advertiser customers win. Not how the agency could win.

This is actually a large part of my work at Microsoft. I spend a lot of time thinking through the impacts of market changes on all advertising ecosystem participants, and I try to build models that will help them evolve and thrive as the ecosystem changes. Advertising agencies occupy a big part of my thought process.

Which brings me to the second conversation I had, this one with a senior executive at a U.S. digital agency. We were ruminating on my earlier conversation, and I asked him how agencies were going to evolve. He responded:

All the major agencies, especially at the holding-company level, recognize that the market is shifting and the existing models aren't going to work much longer. We get paid mostly on an hourly rate, which worked out as a proxy for value for the last hundred years or so.

But as media fragments, we can't keep shoving more people into the mix. At a certain point, this all needs to be automated. Our agency buys from hundreds of publishers a year -- significantly more than our traditional agency counterparts. The average ad network deals in thousands of publishers. Just logistically, we can't deal with trafficking, reporting, billing, and mundane issues like over- and under-delivery with any more publishers than we buy from right now. We need to automate the buying and campaign management process significantly. And once we automate things, if it actually works correctly, the billable hourly rate system simply won't work any more because the manual processes will be too efficient.

I asked him how the big five holding companies were going to figure this out, and he said something striking: "They're just going to watch the market and buy their way out of the problem. Why do you think WPP bought 24/7 Real Media? They're trying to figure out how to scale. How many publishers does the 24/7 ad network deal with? Can they apply that model to media buying?"

This fascinates me because it intersects with my thinking for the past year. How will ad agencies deal with automation? I've talked to almost all of the major digital agencies and a bunch of traditional agencies, and everyone agrees we must find a way to scale online advertising. Everyone agrees that traditional media will move to digital processes over time (even if the delivery isn't digital, the buying, selling, and measurement tools will use online ad models). And everyone agrees that even with the most cutting-edge tools for online media, we're not close to solving the problem yet.

Whether they figure it out internally or buy their way out of the problem, media agencies must keep an eye on the future. They must find a way to scale their businesses, grow margins, and add significant value to the ecosystem.

People who don't understand this ecosystem, especially those from a technology background, will rub their hands together gleefully and drool over the opportunity to disintermediate the agencies. But this is asinine. Agencies perform a valuable service to advertisers, and it isn't just about manual execution, even though this has been the accepted proxy for value from a billing standpoint. Agencies aren't to be trifled with. Antagonize them at your peril, technology startups.

The opportunity is not to kill agencies but to help them survive and thrive going forward. There's a lot of money in that business, and that sounds suspiciously like an exit strategy.