The Data-Driven CMO: ‘No More Burn to Learn’

Written on
Sep 28, 2011 
Omar Tawakol  |

ADOTAS – Over the past 20 years big data and analytics has had a transformational impact on changing the way we do business. Our culture has increasingly become a data democracy or an environment where individuals have access to unlimited information from diverse sources and the power to make informed decisions with greater responsibility.

Nowhere has this been more prevalent than in the media space where a reliance on a “burn to learn” approach has been utilized by media planners and analysts for some time. In this practice, a media planner may have a hypothesis on a particular audience and due to various exigencies like client budget and competition for campaigns the need for certainty is paramount before a commitment can be made.

So in order to see the results, they might spend a little money on a little media to just get a sampling of what’s to come. This form of testing is employed in a myriad of industries and is a reliable safeguard in managing potential risk.

This approach may be effective in getting real data so that the first step in the media buying process is a confident one, especially when millions of dollars are at stake in selecting the right set of inventory or audience. However, there’s an alternative approach in achieving this goal but it’s rarely employed.

Why? Because it challenges the everyday minutia of our lives that has increasingly become more busy, more distracted and more demanding.

If done wrong, using the “burn to learn” approach can be a losing proposition.  Here’s a quick look at the costs that it take to run this type of test:

Speed Kills

In this case speed kills in a positive way.  If testing can be done in a quick and efficient way then that’s ideal.  But if test results take time then the lack of speed kills.  It kills client enthusiasm, team motivation, chances to iterate and time to market.  So in order to extract any value from this type of approach, speed is essential.  Without it, the risks go up.

Money Is Scarce

Budget dollars for testing is only marginally easier to get than expanded budgets. Why? Because it’s somebody else’s money and gambling with somebody else’s money is risky especially when you don’t have any data to support your bet.

Testing Inventory vs. Audience

This is flat-out wrong. At its core, advertising needs to touch actual people, so testing proxies will distill little information about an audience.

You Could Be Wrong

Using a small sample audience doesn’t paint an accurate picture of the types of results a media planner might get at greater scale. Testing a small sample will only result in data on a small sample. It’s difficult to extrapolate data to begin with so it’s vitally important that the ends justify the means in this process.

The next revolution of audience targeting eliminates this practice of “burn to learn” by giving media buyers data on a large enough scale that will enable them to make better and informed decisions before spending media dollars through analytical solutions. With analytics, media buyers can better understand which first party audiences need to be targeted, eliminate any costs to model or simulate multiple audiences in real-time and make audiences go live when appropriate.

While a true data democracy might be a stretch for our industry in the near term, putting a stop to the “burn to learn” approach is within our grasp today and good first step.

As Chief Executive Officer at BlueKai, Omar is responsible for the overall management, growth, and vision for the company's groundbreaking exchange business. Prior to founding BlueKai, Omar was the Chief Advertising Officer of Medio Systems, the leading provider of mobile search and advertising solutions for carriers such as Verizon and T‐Mobile. He joined Medio from Revenue Science, where as Chief Marketing Officer and General Manager he was instrumental in founding and growing the company's leading behavioral targeting businesses, which provide services for many of the largest Internet publishers. There he was actively involved with key industry leaders in defining standards for behavioral targeting.

Omar earned Master of Computer Science and Industrial Engineering degrees from Stanford University and a Bachelor of Engineering degree from MIT. While at Stanford, he was a computer science researcher at both the Stanford Logic Group and HP Software Labs. Omar's research on formalizing content was published in the American Association of Artificial Intelligence. Omar is also an active speaker and commentator on topics such as behavioral targeting, mobile advertising, and consumer privacy. He was listed in Media Magazine's Media 100 for 2008.

Reader Comments.

I think we can and will all agree with the headline on this article but the body does not tell us anything we did not already know. The titles implies that BK has figured out a way to assure clients or partners that with their data; there is no more “burn to learn” time. I am sure BK can agree, data speaks louder than words.

Posted by JK | 12:06 pm on October 1, 2011.

Leave a Comment

Add a comment

Tags: , , , , , , and
Article Sponsor

More Features