Digital Ad Buying Has a People Problem

The main problems, according to NextMark CEO Joe Pych, are around the very basics of executing, implementing and reporting a plan. That means these high human costs aren’t even associated with how to put together an awesome media plan.

“People are spending too much time doing grunt work and copying and pasting from one system to another,” he said. “There’s a lot of low level work that should be automated.”

Pych ran his figures by a several digital agency veterans, who concurred they were directionally accurate. One thought it should be higher, between 10 and 12 percent, according to Pych. The caveat is the typical agency “it depends.” All campaigns are different, so mileage may vary. But overall, there’s an overall figure for transaction costs, and Pych feels like at this level.

“It’s death by a thousand cuts,” he said. “At the end of the day, there’s a lot of people scurrying around doing very, very basic things.”

There are, of course, many trying to solve this inefficiency problem. Google has made it a priority. MediaOcean was formed on this basis. Agency holding companies see their trading desks for models of how they can use technology to get leaner.

The unspoken truth is it involves fewer people. Agencies currently employ armies of junior people who are spending much of their time in Excel spreadsheets. For all the whiz-bang tech afforded by digital media, all roads typically lead back to the humble spreadsheet. All that cutting and pasting is expensive. Typically those preaching efficiency pretend that automation won’t mean agencies employing fewer people. That’s not the case. Automation in every industry means a shrunken but more specialized workforce. The same should happen in advertising.

A few weeks ago, I was told by a digital media executive that one of the holding company’s automated arms housing its trading desk was more profitable for it, with 300 people, than its enormous media agency network with over 10,000 people. The future of agencies is most likely somewhere in between, but closer to the automated unit’s makeup.

Assuming Pych’s analysis is accurate — you can dive into the specifics here — it leaves the question of Google’s figure. It is often bandied about without citation, but it would appear that Google is using a figure arrived at by investment research firm ThinkEquity.