‘Crap optimized on crap’: The agency challenge with data

“It’s crap optimized on crap.”
Speaking to a room of hundreds of ad executives, Rishad Tobaccowala, strategy and growth officer for Publicis Groupe, noted in his keynote speech that “the way we think about data is all wrong” to the point where major companies still have the mindset of gathering all the data. But the reality is, except for Facebook and Google, no company has good data. The quantity of data doesn’t matter — instead, it’s the insights that are gleaned from the numbers by real people solve problems.

“We become so data-obsessed that the whole digital marketing doesn’t work because it’s just crap optimized on crap,” said Tobaccowala. “If you think data is the new oil you are wrong. The whole idea of refinement is a big deal.”

Clients aren’t organized well.
Data quality aside, siloes within client organizations are another big challenge faced by media agencies. Large companies typically have various in-house teams that own different pieces of data without talking to each other, while they expect their agency partners to see across that and assemble those datasets together, which is becoming detrimental to client strategy, said Sharona Sankar King, evp and head of marketing science for BBDO.

“It is less about technology and more about culture,” she said. “It is really important to break down the siloes between the teams.”

Tom Miller, vp of data solutions sales for agency Merkle, echoed the same sentiment. “The challenge is to bring people into the same room, have centralized data and make it useful for everyone in the organization,” he said. “It all comes down to [profit and loss statement] ownership: If I go to a client and say ‘Give me that piece of data and you could be successful,’ they may not want to because they feel that they are letting something out.”

Find talent who can see insights.
All the data in the world, even the best data, isn’t worth much without a human being to extract value from it.

If your sales data shows that 80 percent of your consumers buy one or two units of your product, while 20 percent buy four or five units, there are two ways to look at the data: You can analyze the numbers without thinking their implications for advertising, and say “The majority of people don’t purchase more than two units,” or the person who can apply the numbers to advertising would say “Let’s target the 20 percent who buy the most” and divide your product portfolio.

“Finding a person who can extract an insight from data applicable to advertising is different from finding someone who is good at numbers,” said Elizabeth Crawford, svp of strategy and insights for Match Marketing Group.