Last month, as reported in Ad Age by Jack Neff, Unilever lowered the boom…
“Who’s Next to Fall? Unilever’s Massive Job Cuts Put Other Marketers on Notice” Analysts say more downsizing on the way…”
Hold tight: Unilever’s announcement last week that it’s slashing more than 800 marketing positions is an acceleration of a widespread downsizing of marketing departments, analysts say.
Of course, the last thing marketing and research should do is “hold tight”; this is a wake-up call to change approach, culture, and to be held accountable for profit results.
THE PROBLEM: WE ARE DRIVING FORWARDS LOOKING BACKWARDS
Marketing research mostly delivers insights about marketing productivity that come from the rear view mirror which have decreasing usefulness in a rapidly changing digital world. Most brand trackers report historical relationships between brand attributes and brand equity measures that are slow, expensive and have limited value for driving measurable Marketing ROI, as currently designed. Most conduct marketing mix modeling in ways that require 2+ years of back data and only work on levers with lots of spending over that timeframe…this puts data-driven marketing decision-making 3-4 years into the PAST, and tells marketers nothing about new approaches like geo-fenced offers on smartphones precisely targeted by customer and third-party data.
MARKETING RESEARCH MUST REINVENT PRODUCTIVITY PRACTICES
The Marketing research function should commit to a “productivity” practice. Their mission would be to leverage data-driven approaches in a digital, social, and mobile age that drive measurable increases in marketing ROI. Leveraging data science, this practice will empower the marketing enterprise to improve ad targeting, create relevant matching of communications to users, create ideas that spread, provide real time feedback for course correction and deliver against moments of opportunity (e.g. Oreos tweet during the Superbowl blackout, Walmart stocking the right things in advance of a storm).
The mindset and question set for this practice will be different. For example…
| Traditional productivity research: questions about the past | New productivity analytics: predictive questions |
| Campaign assessment | Campaign optimization |
| What is the profile of our current customers? | How do we identify lookalikes for our current customers for ad targeting purposes? |
| What brand attributes define our positioning? | What targetable behaviors, preferences and conversations in social and digital data identify those most likely to respond to our advertising? |
| What media expenditures most correlated to sales? | What message delivered on what screen, in what context, to which user is most likely to generate positive response? |
| What is the mix of products people buy on given shopping trips? | What should the retailer overstock for the upcoming storm? |
| How effective is mobile advertising? | How do I connect what I learn forensically about users from computer behaviors to deliver messages on their smartphones? |
The Productivity practice will cultivate research approaches that are ultra-lean, lightning fast, and as automated as possible because you are mining through massive data sets for unexpected patterns to crack the code on customer retention, acquisition and responsiveness to communications and offerings. This team will leverage digital data as much as possible, supplemented by micro-surveys towards the singular purpose of improving marketing ROI.
Ad effectiveness needs to move towards becoming self-measuring and reported in near-real time, so marketers can adjust campaigns in-flight, requiring data extraction and analytics to be as automated possible. Two approaches for self-measuring are tagging ads and matching to conversions or creating a single source of ad exposure and shopper data. It can be done!
Wherever possible, set up an ad campaign as a naturally occurring experiment. In digital, this means treating ads as bundles of creative and targeting elements, tagged accordingly, and matching combinations to conversion rates so the marketer can improve campaign effectiveness in-flight. The big data challenge includes storing all these results for retrieval to inform upcoming campaigns.
With productivity research, consumer segmentation uses different approaches (for more on reinventing segmentation, click here); your goal is to create groupings that not only have meaning but also lead to ad targeting to those most likely to respond. This will lead marketers to include leveraging Facebook/Twitter/Google interest profiles, shopping process identification via clickstreams, 3rd party data used for targeting and customer/website generated cookies. Productivity research will lead you to combine on-going experiment results, attribution analysis with marketing mix regression results into a holistic and up to date view of what works.
In a white paper, HP offered, “Leading organizations…succeed using two simple principles: act strategically and use data to drive your execution”. Marketing research does a good job of supporting marketing strategy but fails on driving execution. Our trackers and campaign assessment methods are too slow, costly, and not comprehensive enough. It is time for marketing and research to partner to leverage big data and predictive analytics to drive up marketing productivity.