Marketing and media planners today are in a better position than they were only a few years ago. If data is “the new oil”, then marketing ROI is surely one of the engines that need a high-octane version refined from this raw material. Data can be mined from a range of sources, including a brand’s own collection (CRM, transactions, customer behavior), and third-party sources (social media, data management platforms and customer data platforms).
It is often a bumpy ride. But marketers must be in the driving seat and keep their vehicle on track. Budgets must be optimized across more and more individual touchpoints and channels – paid, owned and earned. They need to balance strategic brand-building investments with short-term activation and promotional efforts. In real time, so with their hands on the steering wheel (that is, the keyboard).
When it comes to budget and media optimization, many brands turn to media mix models. Traditional media mix models are based on sales and marketing spend time-series data. Since CFOs typically want to understand what drives sales, they seek to understand its key components: (1) seasonality, (2) sales that are influenced by a their own and competitors’ marketing activity, and (3) the so-called “base sales”, i.e. the base revenue level independent of short-term marketing activity.
While traditional media mix models may do a good job of optimizing “incremental sales”, we often see them fail to explain the long-term effects of brand building. Consequently, more than one brand has led itself down a dead-end path of “activation and promotion-only” advertising, ultimately damaging brand equity and negatively influencing sales. It sounds obvious (Les Binet and Peter Field’s analyses have been well known for quite some time!) – but we often see it happen.
Competing for budget
So, how much should brands divide their budgets between brand building versus activation activities? Are all marketing channels equally effective for both? The answer, of course, can be found in data and not just gut feeling. We have seen second-generation marketing mix models effectively overcome the “activation only” trap, by also quantifying the medium- and long-term effects on brand equity.
The baseline of sales evolves over time. It is an expression of ad stock and brand equity. Our analyses in many industries and markets have shown that base sales and brand equity are highly correlated. This means that brand equity is a valid predictor of actual base sales. This is where touchpoint models come into play. They help explain which touchpoints (owned, earned or paid) drive brand equity and, ultimately, sales.
But marketers’ hands need to stay close to the controls while driving high-efficiency marketing – just like a Formula 1 driver’s hands need to stay on the steering wheel: Optimization of marketing ROI is not a one-off action, and needs to be adapted to changes encountered along the way.
Marketing activities evolve, customers change, and new data keeps pouring in. Luckily, today there is something comparable to driver assistance systems: Artificial Intelligence. We can use the latest machine learning tools to derive auto-optimization models that keep up to date with data streams as marketing activities evolve.
And, of course, the driver needs a dashboard on which to gather information and influence the machine: an intuitive optimization and simulation cockpit is essential for the high-octane marketer.
What’s the prize for successful high octane brand marketers? An average improvement of marketing effectiveness of between 20 and 30 percent in the first year. Holistic marketing mix models have passed the test-drive phase, not only in the automotive industry, but also in packaged goods, logistics and other sectors. The average ROI of second generation marketing mix models is disproportionately higher than that of simplistic approaches that merely take account of short-term media effects. Let’s drive!
Written by Markus Eberl,Senior Director,Kantar Analytics Practice