Marketing ROI: time for a 'no excuse' policy.

At the age of “real time marketing”, “marketing automation” and other “big data”, at a time when technology is supposed to propel marketing into the era of accountability and predictability, most marketers still have trouble answering the fundamental question of marketing ROI. Surveys after surveys, this disgruntling fact gets confirmed and several root causes can be identified. This post explores one of the main ones:  the perceived complexity and challenges of marketing mix modelling (MMM), a proven approach to help companies identify the drivers of their business. This post explains why this perception is quickly getting outdated and will soon appear as a wrong excuse for lack of accountability.  Indeed, a new technology enabled approach to MMM now brings untapped opportunities to marketing departments genuinely interested in assessing and optimizing their ROI. In addressing the key challenges of traditional modelling, this approach has the potential to bring full accountability at the core of the marketing practice. It’s time for marketing to notice, step ahead of the curve and drive a no excuse policy to marketing ROI – anticipating on the expectation of the C-suite.

The litany of 'gut-feel' marketing.

Year after year, different surveys point to the consistent lack of understanding of marketing ROI by most marketing departments. Just a few examples:

•      In December 2012, a research amongst UK marketers showed that they were throwing away billions of pounds a year in spend by relying on gut instinct and guesswork, rather than data and rational decision-making.
•      In that same year, a report identified that 70 per cent of CEOs believe marketers lack credibility, because they cannot prove the business impact of marketing.
•       At the end of 2013, the annual CMO survey run by the AMA has shown that only 1/3 of US marketers can prove the marketing impact quantitatively.

And so goes the litany of gut feel marketing. Despite reports after reports alerting to growing discontent and mistrust among the C-suite, the bulk of marketers continue the same old “spray and pray” approach. And pinpoint to the challenges and complexity of approaches like marketing mix modelling to justify the lack of a rigorous ROI measurement.

Looking at the excuse: shortcomings of traditional marketing mix modelling

Marketing mix modelling (MMM) is a well-established methodology to rigorously assess and model the impact of important marketing variables (e.g. media investments, promotions, pricing…) on key performance indicators (e.g. volume sales, customer acquisition, qualified leads…). For savvy modellers and marketing departments (like consumer package goods), the approach is proven, well understood and drives actionable insights when combined with business acumen.  Yet, outside of what can be described as an insiders’ circle, the reputation of MMM is not quite at par with the value it is supposed to deliver.  As a matter of fact, quite a few marketing departments that have genuinely embarked into MMM projects have been somewhat disappointed. The root causes of the shortcomings usually trace back to a combination of different factors:

•      Time required by the end-to-end process, which often involves several weeks of data gathering and validation prior to a few additional weeks for modelling.
•       Lack of flexibility, as any further question or new data set will require another few weeks of modelling after the first results.
•       Lack of transparency, as model assumptions are rarely fully disclosed or understood, leading to sometime limited internal buy-in.
•       Low scalability: due to the above challenges, modelling gets done only once or twice a year at best, primarily for major brands and/or regions. Its potential impact on marketing optimization remains therefore limited and in the end its cost-to-benefit ratio gets questioned. 

     Despite the robustness of its methodology and its potential positive effects, MMM in its traditional form has some way to go to become part of a standard marketing toolkit.  So far, its impact has remained somewhat limited to the marketing teams that can afford the resources, skill sets and budgets to overcome its challenges. And for other marketing departments, perceived shortcomings or actual experience have provided reasons no to move.

No more good excuse: the technology enabled approach to MMM  

So, how to address the above challenges? As often, technology can step change the way  old questions are handled. A great illustration is the modelling application developed by marketing software firm marketingQED. Armed with sales data and details of recent marketing activities, the marketingQED software can run thousands of potential models in a matter of minutes. It uses data visualizations to show when and where marketing activities are most effective, giving the user, even if non expert,  the power to adjust and forecast accordingly. What would require several days, even a few weeks, now becomes available after a few clicks and the time for a coffee. So how does this new modelling capability transform the MMM experience and output capabilities? Let’s look again at the above issues one by one:

• Time required: getting the appropriate modelling output in just a few minutes means that insights can be obtained straight after data validation. Combined with a streamlined data management process, ROI optimization gets close to a real time, always on capability, in line with today’s marketing expectations. 
• Flexibility: multiple insights and questions can be explored, thanks to the speed of modelling. Modelling is no longer about one unique overarching Truth, always arguable, but about fruitful conversations among stakeholders, which drive buy in and alignment.  
• Transparency: all assumptions can be spelled out, different ones can be explored if necessary and get assessed via sound business logic, because the modelling exploration is always conducted in the presence of the final marketing user.
•  Scalability:  modeling requirements can scale up to different levels of complexity because the marginal cost of each new modelling project lowers along the learning curve

So time to move on and realise that yesterday's acceptable reason is becoming today's wrong excuse.  Companies that engage into systematic assessment and optimisation of their marketing ROI will ripe very significant benefits, as proven by several studies. Until recently, this incentive was somewhat counterbalanced by the perceived cost in terms of time, resource, skill-set and yes, money. One big roadblock is now out of the way: ROI modelling has the potential to become a marketing routine, independently of the type of industry, level of expertise or budgets. This is a great step towards accountability. For marketing, a 'no excuse' policy to ROI measurement and optimisation should now be the only acceptable policy