We've all heard the saying "it's hard to teach an old dog new tricks" but when it comes to marketing, those who are willing to learn, and eventually apply the new tricks are the ones who are best positioned to turbo charge their marketing performance.
Changing how you think about, execute, and measure your marketing efforts is not an easy task but the rewards can be significant.
Scientific marketing is not about adding an element of science to your current marketing processes; it’s about using science to completely change how you engage with your customers. It’s about leveraging new technologies and techniques to understand your customers’ behaviors and intentions to systematically improve how you influence profitable behaviors.
And yes, it’s about transforming how you think about, execute, and measure your marketing efforts, which in turn, is changing up the status quo. So as we say “out with the old and in with the new”, let’s pay homage to some of the words that are quickly becoming obsolete in today’s age of marketing, science, and technology:
Very few would debate the sheer genius that lies in the minds of marketers (okay, I may have a biased opinion) but at times, intuition is more about assumptions based on gut versus insight based on real results – especially when tasked with trying something new.
The problem with intuition is that it limits the thinking of the marketer. This isn’t a dig at the marketer – we can’t feasibly know every possible opportunity to act on the troves of customer data that we have (although that’s the expectation some are challenged to live up to). The alternative is to swap science for intuition, and suddenly you have marketing that’s driven by data versus hunch.
Out with intuition, in with science.
Those of us who have been around for a while have witnessed some significant improvement in how marketers experiment, exploit, and measure. The shift to digital channels has increased the speed in which we can learn what works and what doesn’t work, but the problem for most marketers is that they assume a sequential approach to understanding what they should do before they actually do it.
When you have thousands or even millions of customers, there are hundreds or thousands of treatment combinations and contexts that will ‘work’. And by the time you apply learnings from your experimentation to the exploit phase, there’s a good chance that what you previously learned no longer applies. What you really need is experimentation capability to simultaneously test any number of marketing strategies, offers, messages, and contexts in ways that allow you to explore all possible learnings with statistical significance while at the same time exploit the marketing interactions that impact performance against long-term business KPIs.
Out with sequential, in with simultaneous.
Tying to the sequential process mentioned above, marketers have long been known for selecting and then scaling champions to drive marketing success. After all, if you can determine the best offer for a specific segment of customers at a set time, then bingo – you’ve hit the jackpot!
This practice assumes that all people and behavior are static, and needs and preferences don’t change over time. And believe me – based on my wife’s constant ‘need’ for a new pair of shoes, I know this is not the case. Behaviors, needs, situations, and contexts continuously change – especially in the digital world – so the practice of scaling champions should be put to rest. Instead, what marketers need are tools that allow for dynamic experimentation where scaling means continuous adaptation of messages relative to current performance results.
Out with scaling champions, in with continuous adaptation.
Most marketers rely on business intelligence teams to build predictive models related to customer behavior and opportunities to upsell and cross-sell. Yet, in many cases BI teams or the tools they utilize lack the ability to provide insight into sequences of customer behavior that lead to purchases, decisions or events that impact customer profitability. And in most cases, the process for marketing experimentation and feedback is very slow.
What marketers need is access to powerful behavioral prediction models without being dependent on the project output of business intelligence teams. They need systems that can perform unsupervised discovery of customer behavior to identify unique attributes, both targeted and untargeted. And they need a system that learns and refines marketing experiments according to the rate at which customers behave and respond, rather than at the rate of when business intelligence teams can complete analysis projects.
Out with dependency, in with freedom to independently discover, test and learn.
Many companies have adopted technologies to advance the sophistication of analytics being applied to their customer data but unfortunately, much of the onus is still on the marketer when it comes to actually applying the output of those efforts.
The problem with this approach is that as mobile marketing is brought to scale – think thousands of different treatments being delivered to millions of different customers in hundreds of different contexts – it becomes impossible for humans to interpret the learnings, let alone apply them quickly enough to have a sustainable impact. Instead, marketers should leverage the power of machine learning to automatically and continually refine offers, and eliminate the guesswork in determining what to present to whom, when and where.
Out with manual, in with machines.