Programmatic

Marketing-Mix Modeling: A Cure For Short-Termism And The Obsession With Performance Marketing

Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Sara Owens, partner, West Coast practice lead, data sciences, at Wavemaker.

Today’s CMO has two responsibilities often at odds with one another: meet short-term revenue goals and build a brand. The first responsibility comes at the behest of shareholders, the second benefits long-term business health.

In fact, 40% of marketing leaders say their top challenge is driving revenue growth, which is evident in the prevailing performance-driven mindset. Yet despite the continued pressure on marketing to drive bottom-line impact, marketing budgets have remained largely unchanged, holding at 11% of overall company revenue in 2018-2019.

Today, the true job of the marketer must be to foster the long-term relationship between the consumer and the brand, a vital complement and balance to finance and sales. While adopted in theory, it does not hold true in practice. When you consider that compensation and shareholders’ measures of success are tied to achieving short-term revenue profitability, balance is lost and the scales tip, fueling the obsession with performance marketing to drive immediate sales.

If this short-termism is the pervasive disease our industry is facing, could marketing-mix modeling be its cure?

A statistical analysis that uses historical sales and marketing data to quantify the sales impact of various marketing activities, marketing-mix modeling has been in the measurement solution toolkit for more than 30 years. Until now, marketing-mix modeling’s key strength is its mastery of measuring every marketing activity’s impact on sales.

But in the era of big data, is it possible we’ve overlooked its ability to connect brand health to overall business growth? Have we been so blinded by attributing an action to every dollar invested in media that we’ve traded insight for a more efficient cost per action?

One would think that if data is used tirelessly, the market would want to ensure that it’s working to the best of its ability at all touchpoints. But that’s just not the case, especially as found in the measurement of the priming stage, aka the upper funnel. Biases toward brands are formed here that can influence purchase behavior, but there’s been very little innovation in measurement in this stage. We’re still relying on expensive survey-based brand health trackers that don’t capture media exposure and ad effectiveness studies that don’t connect to business outcomes on their own.

How did we get here? An over-reliance on data without the proper investment or training in how to use it the right way.

By over-reliance, I mean focusing on the immediate impact. Most measurement solutions are focused on the active stage of the consumer purchase journey – the lower funnel. There has been a lot of innovation in attribution, from automatic content recognition technology that connects television exposure to online sales, to the use of mobile location data that connects digital media exposure to in-store visits.

If harnessed correctly, an annual implementation of marketing-mix modeling can become the CMO’s most valuable measurement solution to justifying marketing’s investment and sales impact. Modeled alongside ad effectiveness results, marketing-mix modeling can prove the value in priming-stage metrics, such as lift in awareness and brand affinity, and quantify how driving lift in brand health leads to business growth.

Why then are only 41% or brands leveraging it? The answer is simple: They aren’t incented to adopt it. Relative to other measurement solutions, marketing-mix modeling is too costly and takes too long to build. Facing down short-term revenue quotas, brands would rather invest those dollars in working media.

It’s time we reject short-termism and reembrace the health of the brand and the longer view. Invest in marketing-mix modeling to glean insights that connect consumers, build brands and grow the business. While easier said than done, there are some ways to help smooth this transition:

  1. Hire the best: Unless you have PhD statisticians that deeply understand marketing and media on your payroll, do not build an in-house solution. Partner with marketing-mix-modeling specialists who have a deep understanding of what the data represents.
  2. Put in what you what you want to pull out: Do not underestimate the time and precision needed during data collection. It is easy to miss data during this first critical phase and impossible to ignore when the results come back. Take your time and involve everyone who plays a role in executing media and other marketing activities.
  3. Measure twice, cut once: This is an old proverb borrowed from carpentry. In our case, check the quality of data twice, model once. This goes hand-in-hand with the above. Remodeling because important data was wrong or missing blows up cost and time.
  4. Include everything, even the kitchen sink: Not sure if a marketing activity is relevant? Include it anyway. Include everything you have data for – the math will tell you if it is relevant. Hold a brainstorm with your team and partners to ensure you don’t forget anything.
  5. Promote it and use it: Don’t let marketing-mix modeling be the thing you did once that dies in a PowerPoint presentation. Socialize the results in your organization. Use the results to guide conversations and decisions. Hold your agency partners accountable to living by it.

Follow Wavemaker (@WavemakerGlobal) and AdExchanger (@adexchanger) on Twitter.

This post was syndicated from Ad Exchanger.