Analytics news and expert advice every Thursday.
Intentional Serendipity: How Marketing Analytics Trigger Curiosity Algorithms And Surprise Discoveries
With the wealth of data available to marketers, marketing analytics can unveil unanticipated breakthroughs that will bolster your strategy, says columnist Joshua Reynolds.
The most exciting phrase to hear in science, the one that heralds new discoveries, is not “Eureka!” but “That’s funny…” — Isaac Asimov
Microwave ovens. Velcro. Viagra. These and many other everyday scientific advances were discovered quite by accident. Serendipitous discovery plays a huge role not only in scientific advances, but also in major breakthroughs in marketing strategy.
With more data open to marketers than ever before, marketing analytics can now play an even bigger role in revealing unanticipated breakthroughs that propel your business. And even though planning for serendipity may sound like an oxymoron, there are ways to optimize marketing processes that boost your odds of hitting pay dirt.
Mining for these kinds of surprises is more important than ever for marketers. Earlier this year, a Deloitte study found 82 percent of CMOs feel uncomfortable interpreting consumer analytics, for example, even though 71 percent of marketers feel harnessing data analytics is one of the most important challenges they face.
Some “That’s Funny” Data Surprises
At my own company, Quantifind, we’ve used these techniques to unearth some interesting marketing data surprises of our own, including:
- Loud and Clear: A major telecommunications brand found that data overage charges drove more customer churn than dropped calls.
- What’s in a Nose? Analysis reveals that millennials prefer to use emojis without noses. Emojis with noses usually indicate somebody 35 or older.
- But Moooooom! A major fast-food chain discovered the best way to boost breakfast sales among teens was to offer moms better coffee, as many teens rely on mom for a ride in the morning.
- Hail the Pumpkin King. Analysis reveals that in the fall, pumpkin flavors make virtually all food and drink sell better, not just coffee.
- Stars in Their Eyes: In more than a few cases, there was a difference between who movie studios promoted as the big star of a movie and who audiences considered the big star, leading to big changes in movie trailers and promotions.
Of course, not every marketing analytics exercise results in a “that’s funny…” moment. But it is amazing what happens when you let your data whisper to you.
Finding these kinds of data surprises requires a lot of sophisticated natural language processing and complex data science. And that data science becomes most useful when the patterns and possibilities they reveal incorporate the thinking of human beings, who contribute the two most important algorithms in the entire marketing analytics framework — the curiosity algorithm and the intuition algorithm.
When you think about it, curiosity is simply an organic algorithm that’s processed in the human brain, the conclusion of which is “This is worth further study.” And while computers can process possibilities and conduct extremely sophisticated pattern-matching exercises at scale and speed, the human in the loop is best equipped to discern which patterns are most likely to lead to meaningful discoveries.
So the first function of a marketing analytics platform is to trigger the curiosity algorithm in the human, at the right time and place.
Similarly, intuition is simply another organic algorithm, the conclusion of which is “This is worth taking action on, even if I can’t articulate the rationale completely.” Just because sometimes we can’t see all the math that takes place in our own brain doesn’t negate the validity of those calculations.
So when it comes time to take that leap of faith between correlation and causation, when it comes time to decide whether or not to take action on the insights surfaced, nothing is better than human intuition to make that calculation.
So the second function of a marketing analytics platform is to trigger the intuition algorithm with as much intelligently correlated data as possible.
These two principles, then, form the basis for what we call “explanatory analytics” — marketing analytics that go beyond merely predicting outcomes or prescribing action and go that next step to helping people explore and understand correlations so they can actually change outcomes.
This is the foundation of the alliance between humans and machines that needs to be redesigned in marketing analytics.
In service of that, here are a few practical tips for making sure you and your team are set up to let your own data whisper helpfully to you and to help you discover as many useful surprises as possible.
1. Divide And Conquer Your Data Tasks
Machines excel at finding hidden patterns that are too difficult or time-consuming to discover by hand, such as the consumer language patterns in social media posts and customer service logs that correlate to sales.
The human brain excels at recognizing and probing the most interesting patterns presented to it — the ones that open new questions, create compelling stories and inspire novel strategies. Make sure you’ve got your humans in the loop at the right places.
Check to see that your marketing analytics systems take an explanatory approach that lets you explore possibilities, rather than a black box that spits out final answers that teach you nothing.
2. Get Clear On Where You’re Most Curious
The only thing worse than the wrong answer is the right answer to the wrong question. Get clear about what you want to explore.
What sorts of business outcomes are you looking to drive? Where are your biggest blind spots around that impact? And what sort of internal KPI data do you have available to serve as a divining rod for what external data does and doesn’t matter?
Once curiosity has set the stage, you can be more confident that your marketing analytics platform is working with the right data to produce more meaningful and actionable insights.
3. Let Your Data Suggest New Questions To Explore
Before you commit to one exclusive area of inquiry, take a moment to let data patterns bubble up on their own and suggest new lines of questioning.
Perhaps you think the best way to impact sales is to optimize creative for a certain demographic. But what if your more promising audience is in fact a demographic you’ve never considered before?
Or perhaps you think the best way to boost sales is to tackle the loudest complaints from your customers. But what if the complaint that’s actually hurting sales more isn’t necessarily the one that shows up most prominently in social media analysis?
Before you unleash the full power of your analytics platform on solving a specific problem, look for organically occurring data clusters that might indicate there’s something new and unexpected worth exploring.
4. Don’t Just Extract. Explore
Too often marketers expect to find a smoking gun in the social media verbatims. But as the old adage goes, if you interrogate the data long enough, it will ultimately confess to whatever you want it to tell you.
Truly serendipitous discovery is an iterative process of exploring unknown territory, not just confirming a hunch with a one-off view into the data. Better insights result from suspending your agenda and parking your preconceived notions long enough to let the machine spark your intuition algorithm.
So in your process, let your technology do the work of filtering unhelpful data and serving up the most promising data visualizations. Then make sure the humans in the equation are well equipped to explore the correlations and options that present themselves.
From there, you can find surprising new questions and correlations to explore, in partnership with the technology.
5. Look For The Shortest Leap Of Faith
Despite all the advances in data science, absolute truth remains elusive. Any analytics platform that promises to reveal actual causality is lying.
All marketing decisions are, ultimately, leaps of faith based on the best available insights. What marketing analytics platforms can reliably do, however, is drastically reduce how far marketers have to make the leap between correlation and causation.
And it’s human intuition that actually calculates that final jump. Assigning a cause to an effect, making the decision to take action based on what you perceive, these are all intrinsically human algorithms that our brains can do better than any machine.
So rather than expecting your marketing analytics platform to serve up concrete certainty that a certain marketing bet will succeed, look for the insights and data visualizations that give you the shortest leap of faith to make. Trust in your own intuition to take you that final step from insight to activation.
Some opinions expressed in this article may be those of a guest author and not necessarily Marketing Land. Staff authors are listed here.