Advertising ROIs, as measured in marketing mix models, are marginal on average. Anecdotal evidence suggests that this may be getting worse. There is an opportunity to use emerging ROI related analytics and metrics to guide better media/advertising decisions, reversing this apparent decline. Such new capabilities could enhance marketing in many ways, but a better “guidance system” for programmatic buying is certainly one of them.
Sequent Partners has done a lot of work with various methods of measuring marketing ROI. Over the past few years we have done a number of public and proprietary studies of the landscape that have put us in close communication with all of the important modeling companies. Both our direct experience with modeling and the interviews we’ve conducted with every major modeler confirm that advertising ROIs are marginal.
That’s not to say that there aren’t some great campaigns that do well. Some industries fare better than others and some media perform better than others. But it is a common situation that advertising investments often need to be justified with a variety of arguments. For example, one favorite is - “when you count the long-term effect, the sales effectiveness of advertising is really twice what was estimated by the model and the ROI twice as high, as a result.”
Most recently, we have spoken with a number of experts from the advertiser, agency and modeling worlds who suggest that this situation may be getting worse. They often cite “the lack of fundamentals” as a potential cause. Poor creative, creative poorly paired with media environments, poor control over scheduling and reach/frequency and unsure targeting based on look-a-like models are all contributing factors to this ROI decline. We only take these as hypotheses, but judgment suggests they may not be wrong.
When you consider the rapid move toward programmatic and other less comprehensive forms of automation, this issue earns an exclamation point! While computerized buying and allocation is unquestionably a necessity in today’s micro-fragmented media world, the algorithms have to be smart enough and fueled by metrics good enough to replace human judgment — or else the result will be worse.
Modeling has been used to measure marketing ROI for decades. Most often it is used on an annual basis, its insights used to shift money around inside the marketing budget. A model may conclude that one medium wasn’t a good investment, the brand should allocate those funds to a different medium next year. Useful. But it reduces the role of the marketer to a fund manager, buying and selling assets. It often results in over-investment in one medium, driving it to the point of diminishing returns. Marketing used to be a creative function! The real opportunity should be in finding insights that enable the marketing team to make each medium the best investment it can be. (Especially when, from a strategic, or channel planning, perspective they know that each medium plays a different role to move the consumer through the purchase journey.)
We are at the beginning of a new generation of ROI analytics. Single-source data, attribution modeling, big data, agent based models and other tools deliver granular insights at the level of households, and eventually consumers. Modeling insights at the market or store level were great for marketing activities that happen at those levels, like promotion. But advertising is about consumers. These new solutions could guide: targeting, creative strategy, media selection at the media vehicle level, and the effective joining of brands with creative with media and life context to deliver the right message to the right consumer in the right moment.
It seems clear to us that moving from market or store level data; from mix modeling to attribution modeling; integrating media and the vehicle level and creative at the execution level, will enable marketers to move from measuring ROI to managing ROI for better business outcomes.
The most immediate frontier for this evolution is programmatic buying. With the newest innovations in measurement and analytics enlisted, programmatic can be expected to produce better and better ROIs.
If not, a different three letter acronym will prevail – CPM — where cheaper is better and quality goes unnoticed and underutilized. That’s a race to cheap, low quality media that will only deliver an acceptable ROI when it is cheap enough. Prices will drop, budgets will be reallocated to more effective marketing factors. A gruesome future for media and agencies and a huge missed opportunity for advertisers.