Madison Avenue’s Odd Couple: Data and Creative Making Beautiful Ads Together


ADOTAS – The Yin and the Yang, The Artist and The Scientist, Venus and Mars. Any one of those descriptors can be aptly applied to the marketing industry’s newest odd couple: the data scientist and the creative director. Long inhabiting different ecosystems, data and creative can now be found coupling up and making beautiful ads together.

Surprised? That’s understandable. As long as Madison Avenue has been the mecca of advertising, numbers guys and creative types haven’t really had much reason to cross paths, never mind collaborate. But that’s changing quickly.

Never the Twain Shall Meet, Until…

It goes without saying that digital technology has dramatically altered the advertising landscape.  Its reach and diversity is eclipsed only by the amount of data it can provide about consumers and their behaviors. Marketers, especially CMOs, have been reveling in this ever growing treasure trove of data (even if they don’t always know what to do with it – but we’ll come back to that later). Creative, on the other hand, could give a hoot. They’ve always cut a wide berth around numbers and the geeks that crunch them.  And why wouldn’t they? Creative’s stock in trade is magic – not mathematics. They’re the Don Drapers of advertising: swilling a few martinis to get the juices flowing then deftly tossing out pithy phrases, dazzling images and catchy tunes. But lately, the data guys have begun to offer up analyses that  really matter to the creatives. And now the Mad Men want to hang out with the Math Men.

Mr. Marketer: Tear Down That Wall

Most marketers still measure and analyze campaign effectiveness one medium at a time – in silos. These siloed results are good at informing marketers of what has already happened and providing a basis on which to suggest potential future trends. But the approach doesn’t take advantage of the full range of data available today and it lacks visibility into the entire market dynamic. It doesn’t measure the way different mediums interact – how an email campaign might prompt a Google search, for example – or how outside variables, such as competitors, can impact campaign performance. Siloed measurement and data can tell creative if an ad performed well, but lacks the more nuanced data providing insights into why.

Recently, data scientists have created methods to tackle the why questions. Using new, more sophisticated algorithms, they can break down silos and seamlessly integrate all data sources (first-, second-, and third-party) to get a complete picture across mediums. A recent Harvard Business Review article calls the new measurement process “analytics 2.0”, and siloed measurement “analytics 1.0”. The article’s author, Wes Nichols, says “marketers who stick with traditional analytics 1.0 measurement approaches do so at their peril. Those methods, which look backward a few times a year to correlate sales with a few dozen variables, are dangerously outdated.”

New, more predictive, comprehensive algorithms and methods can measure dozens of consumer touch points such as which channels influenced a user to convert or buy vs. looking only at the channel where a conversion took place. These methods can break down silos to measure across all channels and mediums, making it possible to gauge performance across a range of variables, such as campaign types (e.g. brand awareness, reach, social, or website), or audiences, all in real time. Instead of looking backwards, this approach allows brands and agencies to run “what if” measurements to help predict likely outcomes. The data can then provide insights to inform resource allocation, creative direction and other campaign elements – all on the fly – to boost conversion rates and ROI. Let me describe an example.

Things Aren’t Always What They Seem

As part of our work with a major U.S. apparel brand, we measured a campaign they were targeting toward rugged, outdoorsy men, aged 18-35, with a high sports affinity. The campaign’s media buys and creative were all very “man-centric”. The creative images included a panoramic photo of mountain climbers and another photo that was a clean, simple close up of a rugged looking man.Instead of simply measuring performance against the intended target segment (a siloed approach), we measured the campaign across a wide variety of audience segments including those outside the target. Much to our surprise we discovered that the intended audience (male, ages 18-35) did not perform very well (e.g. make purchases) but that another segment that had not been actively targeted — married women — performed very well! Our measurements allowed the client to test segments being actively targeted with its media buys as well as those not being targeted. The resulting data helped not only inform decisions around new media buys, but also realigned the campaign’s creative.

Since women, rather than men, were viewing the ad and making the purchases, the creative team eliminated the panoramic image and fine-tuned the rugged looking man photo to better appeal to married women. The media buy and creative changes fueled by the measurement data helped increase conversions 9x, and resulted in $3M+ incremental revenue. This underscores how data and creative make ads more effective together. Data can inform the creative types that women, not rugged men, are doing the buying. BUT it can’t intuit what visual or phrase might appeal to those women – that’s where the “magic” still comes in.

A Match Made in Heaven

Comprehensive measurement and analyses provides marketers with a host of predictive tools and insights, including data that can inform creative as never before. Sure, creatives will still be the ones with the magic, but now they will look to the numbers guys to learn a new trick or two.  Suddenly the Data Scientists and the Creative types of the world don’t seem to be such an unusual pair. In fact, the Mad Men and the Math Men may become a match made in heaven.


Please enter your comment!
Please enter your name here