CURIOSITY, INTUITION and DATA:
DESIGNING a NEW ALLIANCE between ANALYTICS and MARKETING
By Joshua Reynolds
Head of Client Consulting and Marketing, Quantifind
Josh Reynolds is the moderator of &THEN Inspiration panel “Marketing NeXt: A look into the Lens @ 2017 and Beyond”. Click here for more
Marketers are in a dysfunctional relationship with data. We expect the wrong things from data, like certainty of success and silver-bullet growth strategies. We’re suffering from stimulus overload from looking at too many marketing analytics dashboards. But we’re not getting the insights we need to drive our businesses forward, so we keep looking. And we keep feeling overloaded.
Meanwhile, executives expect marketers to tap into rich streams of data all around them and discern with certainty what’s happening with the business and discover ways to grow revenue. According to a recent study by the Fournaise Group, three out of four CEOs think CMOs lack business credibility because they’re not talking about revenue. According to Deloitte, two out of three CMOs say using marketing analytics is one of their top priorities, but seven out of 10 CMOs say they’re still struggling to glean the benefits of marketing analytics.
It’s a vicious cycle, and it’s time to get off the merry-go-round. It’s time we run an intervention for our addiction to data and redesign the alliance between marketers and analytics.
It starts with recognizing that analytics platforms aren’t the only source of intelligent computing. The human brain remains one of the fastest, most flexible, and most potent forms of computing on the planet. When you think about it, a marketing analytics platform is really the integration of more than just data sets and visualization engines. To be truly effective, a marketing analytics platform is the integration of data analytics, visualization, and the human operating system manifested in marketing decision-makers.
In other words, the computer is only half of the analytics platform. We, the human operators, are the other half. And when we think of it that way, we can begin to eliminate the dysfunction from our partnership with data.
To be even more specific, there are two algorithms that take place within human cognition, and they are two of the most important calculations in all of marketing: curiosity and intuition. These are proven, mathematical calculations that neuroscientists can now observe taking place across synapses, axons, and neural membranes. While curiosity and intuition are still “black box” algorithms, and the IP behind them hasn’t yet been opened entirely, they are no less valid than calculations taking place in binary-coded silicon circuitry.
Curiosity is an organic algorithm, the output of which is the conclusion that “this is worth further study.” Human brains are uniquely designed to recognize meanings in shapes and patterns, and we’ve been programmed for countless generations to have an innate sense of when it’s worth it to spend a portion of our finite, mortal lives exploring something in more detail. This ability to see patterns and trends and shapes and anomalies and outliers is something that a computer can be programmed to do—and in some cases has been—but the decision to look for clues lies exclusively within the human domain.
For example, Quantifind recently worked with a Quick Serve Restaurant that wanted to grow its breakfast sales. So the brand developed a new breakfast menu item aimed at teens. The new food item tested well with teens, and the creative content and campaign materials tested well, too. But when it launched the new breakfast menu item, sales didn’t grow. Traditional analytics revealed nothing helpful or insightful, only a confirmation of what they already knew—teens, despite voicing an appetite for the new food, weren’t coming in to buy it.
But when Quantifind (my employer) used its explanatory analytics platform to explore what was happening, it discovered a new signal in the noise—not around teens, but around moms. And within the moms in the data an interesting pattern emerged, one that caught the attention of the human operator—a particularly strong signal around coffee. Turns out moms didn’t like this restaurant’s coffee. And when human curiosity was triggered, a further exploration revealed that teens relied on moms for a ride to the restaurant, but moms wouldn’t go there because of the coffee. So the restaurant launched a new coffee menu, leading to an increase in both breakfast and coffee sales.
This is just one example of what’s possible when we think of analytics platforms and human cognition as two halves of the same data system.
Therefore, the first function of any marketing analytics platform is to trigger human curiosity. That means the analytics platform needs to filter out everything that definitely doesn’t have any measurable connection with success or growth, and that’s where synthetic computational power shines the most. Computers can process ungodly volumes of data at blinding speeds, weed out the junk, and generate a subset of data correlations that have the greatest likely chance of having some bearing on the business. Then the analytics platform must generate intuitive data visualizations that allow data-savvy marketers to quickly discern between potential game-changing insights and wild goose chases.
The other critical computation that takes place in the mind of the marketer is the intuition algorithm, the output of which is the conclusion that “this is a causal relationship” and “this is worth taking action on.” No data platform, no matter how sophisticated, can ever determine cause and effect with certainty. The best a data platform can hope to achieve is to systematically eliminate everything that definitely does not have a causal relationship. It must measure the strength of the correlations that remain, and then offer marketers a line-up of suspects of causation. Beyond that, it’s up to human intuition to decide which suspected causal factor to go with, and what course of action to take next. That means the analytics platform needs to go as far as it can to explain not just predict likely outcomes, and offer clues and choices as to how to change those outcomes.
For example, another of Quantifind’s clients was looking to increase opening box office sales for a soon-to-be-released military movie, which by all accounts should have been a smash hit. It had an Academy Award-winning duo of director and producer, an incredibly powerful and topical storyline, and broad market appeal across multiple demographics – in particular conservative Americans who liked a good story about soldiers. But an initial screen through a time-tested predictive analytics solution indicated the movie was going to open to much lower box-office numbers than was otherwise expected. Unfortunately, the predictions offered no clues as to why the movie was going to open up to poor numbers, or whom the movie studio needed to win over in order to get more butts in seats on opening weekend.
Quantifind’s explanatory analytics solution, however, soon revealed that while the movie studio had expected conservative America to line up to watch soldiers acting heroically, somehow the movie can become hyper-politicized, and was seen as agitprop for the current Democratic leadership. And the only way to come to this conclusion was to remove all the factors that definitely didn’t have any connection to ticket sales and then line up all the remaining possible explanations and let an intelligent human brain take it from there. The movie studio used this insight to adjust the movie trailers, take the attention off the political elements, put more focus on the heroism of the soldiers, and the movie opened up 41% higher than originally projected.
All this because of intelligent human-in-the-loop computing where both machine and marketer knew how to play their respective roles.
The second function of a marketing analytics platform, then, is to trigger human intuition. That means the strength of the correlations and the reasoning behind why certain insights and findings must be served up transparently to the marketer. Black box analytics platform that simply say, “trust our math” are missing the point—analytics platforms must stimulate the organic math that takes place in the form of the intuition algorithm. The analytics platform must find, filter and visualize the voice of intelligent data, and use that to inform an inspired, intuitive marketer who can put those findings into action.
So ask yourself… how can you adjust your own relationship with data? Where are you and your analytics platform getting along? Where can you redesign the alliance? And how can you optimize your marketing analytics solution to trigger your curiosity, tap into your intuition, and unleash the full business potential of you and your team? The marketers who figure out how to get into a healthier relationship with data are the ones who will lead the industry in 2017 and beyond.