[ Summer 2010 ]
Breaking the Silos in Media Measurement:
An Interview with MediaVest's Yaakov Kimelfeld
As MediaVest's Senior Vice President, Digital Research and Analytics, Yaakov Kimelfeld stays on the leading edge of innovative approaches to integrated media measurement. His goal: To gain the clearest picture of ways to improve return on his clients' advertising investments. Kimelfeld is a thought leader in emerging media research and analytics, accountability metrics, and bringing discipline to the management of ad spend. His approach is open-minded yet uncompromising, and his knowledge of analytical approaches to media integration is unmatched.
Kimelfeld took the time to speak with A:I/R earlier this month about the promise and perils of "multichannel" research, and how he is helping clients rethink their assumptions about measurement to understand the converging media universe.
How do you approach the task of making sure that you are always providing clients with innovation and the highest quality in cross-media?
We evaluate our clients' needs and then constantly scan the outside world for what is available, evaluate emerging players and develop our proprietary methodologies. We always strive to optimize campaigns based on key business objectives; instead of analyzing a single channel every time and just looking at those numbers, we break the silos and create a multichannel "big picture."
How well do you think the integration of digital and traditional channels is going, in terms of planning and research?
Across the industry, there is still a long way to go. Companies tend to follow a traditional marketing perspective, viewing every medium as a separate thing. Many agencies have been built around specific silos – you have a division for print, a division for TV. Three years ago MediaVest began actively de-siloing its structure and cross-training its buyers to more closely buy and plan media based on consumers' habits. We have an internal educational division for all junior employees called Connection College; essentially, they learn to plan any campaign, offline and online. Of course, there will always be specialization; but we try to instill a universal knowledge of the tools for planning, measurement, and analysis into all our employees. We realize that we should plan not within specific channels but around specific consumer behavioral patterns.
Can you say more about the integration on the client end?
Advertisers understand the value of integrated marketing – the necessity of looking at levels of engagement and brand recognition, instead of just exposure metrics and measuring ROI on marketing spend. Unfortunately, media measurement convergence remains largely a theory. There are so many ways to analyze things and to measure, but people do not always know how to pull these strands together. Not many people think about proper attribution.
It sounds like clients want customized currencies – which is a bit of a contradiction.
Well, it used to be a single-currency world dominated by Nielsen ratings. But people realize it's not working anymore. And, for all the talk about engagement, it is very far from being a currency.
The most frequent request we get from our clients is to establish a common performance metric across platforms for intelligent cross-media budget allocation solutions. Essentially, we are asked to create a unifying currency, a "media Euro" of sorts, to adequately appraise the value of different media.
Today it seems to many like an impossible task; however, as all media becomes increasingly digital, then the U.S. currency will be such "common currency": the advertising will become fully accountable for its contribution to transactions, revenue or brand equity.
Until then, however, when you want to be strategic and gain the advantage of being the first in some kind of emergent media, you have to prepare to withstand a battle with old-timers. The burden of proof is always on a new, emerging medium. The only way you can scientifically prove it is to run a model. You can have anecdotal examples – how many fans your campaign generated on Facebook; you might even know that more people bought the product. But it's all "one-off"; the only way you can compare is to do a model, and a model will always be late, and it will be too general to actually optimize the spend that is being wasted now. By the time the model is ready, it will likely be obsolete.
Clients would love for us to produce online metric that is similar to a GRP on TV – and such metrics do exist. What you really should do, however, is to start measuring TV as online, and not online as TV. That is where things are going with addressable TV, with mobile TV, iPad, iPhone – all the new ways to serve video ads. TV had already become just one of the ways to serve video.
Do you feel the industry is seeking accountability and ROI in the right ways? Is there room for improvement?
There will always be room; there is no such thing as too much accountability (laughs). Everybody sees it as a great thing; everybody is pursuing accountability; but there are different types of accountability. How my particular channel or campaign performed may be completely different than multichannel accountability; the results for a particular channel may be disputed. The new accountability is separating all spurious or confounding factors, and looking only at the results of the campaign. Accountability should be a rigorous statistical exercise, and the best type is a test-versus-control experiment.
Unless you are accurately measuring the multichannel effort, there is no way to know if having more channels is actually better for the business: there could be a disconnect between message and the media, or one channel may be cannibalizing another channel.
The undisputed king of such metrics is "Cost per Revenue Generated," i.e., to what extent the marketing spend on a particular channel, platform, program, campaign generates positive return for your client's business. It is not readily available, though. Today we rely on econometric modeling to uncover what media investments are turning profits and how the different channels work together. While this currently serves as the most accurate way to allocate budgets by media, traditional econometric models are slow. And with the media developing the way they are, it is very hard for an econometric model to catch up and produce actionable recommendations in time. Real-time performance data, when not available, is the biggest gap in the media planning cycle.
What do you see as the obstacles clients face in taking data that they've acquired and putting it into action?
The industry is used to surveys; and while surveys continue to make sense for many applications, new methods continue to evolve – but they are trickier to navigate. Today's emerging customer and audience behavioral data is complex and granular; it does not always make immediate sense, and the abundant data can become a liability. It comes from different providers, which are very fragmented. Vendors come to clients to impress with their ways to measure iPhone, or rich media. It is tempting to concentrate on a small picture coming from a certain platform/medium – but it kills the grand measurement design. If you do not have all-encompassing measurement plans – how you measure the entire campaign, and analyze it based on client-specific objectives – all of these cool, shiny tools are in vain.
Just saying "multimedia" assumes that you measure every thing in a different way. We need to think about "convergent media" – how we can scale that. It is not even "channels" anymore; it's instances of connection.
How is MediaVest using the KN Dimestore platform?
KN Dimestore allows us to measure the power of a campaign when the immediate behavioral indication of success is unclear. Of course, clicks are not a good measure. KN Dimestore is a good way to tap into what is going on and optimize the campaign before it is over, as results are available real-time for ongoing monitoring of campaign performance. It also has obvious advantages over traditional large branding studies – Dynamic Logic types of studies – in certain situations. The technology is very smart.
What do you see as the other big challenges that you are tackling?
Today, no matter how large you are and how much you mass market and flood the airways with messages, you cannot get the undivided attention of the target audience. The world became so fragmented, and people began to do so much multitasking. With this, everyone must become serious about accountability when it comes to marketing. You can actually find very granular audiences on TV now using set-top-box data; and on the Internet, it goes without question that you can find your audience. But there are no standard ways yet to look into granular audiences and compare campaigns' performance across TV and Internet and mobile on a national basis. So this is my task – to make that happen for our clients.
