Family dynamics can turn simple business intelligence decisions about buying beer for a Fourth of Mapping Out a Trajectory for Targeted Sales Using the Math of Predictive AnalyticsJuly party into a predictive analytics challenge.
At last year’s party, Uncle Chet downed a six-pack of Coors Light and his five still-at-home sons ripped through two cases of PBR. All things being equal, I’d pretty much know what to buy for them this summer.
But last fall, the middle son moved to Brooklyn and developed a taste for nano-brew beers, while his younger brother discovered cocktails during his semester abroad. And since the oldest kid’s DUI on New Year’s Eve…well, enough said. There might be a predictive analytics algorithm for stocking up on beverages using kind of data, but I haven’t found it yet.
So, how hard must it be for companies – faced with infinitely more variable data — to know what merchandise to buy or services to offer?
Retaining a loyal customer base – while also growing a business and attracting new customers – has long been more art than science. For a long time, business managers and their sales forces relied on their own observations and gut instincts to market to customers and to promote their services, but not anymore. As the science of market research practices continued to prove its value, companies swarmed to programs – and business tools – that helped them refine homegrown approaches. With the arrival of powerful computers and the ease of tracking large amounts of data, that research only grew more complex and more specialized over time.
Business Intelligence Targets the Past, Statistical Analysis Looks to the Future
Business intelligence, a catchall term for the reams of information that churn through a company’s databases every day, is a useful tool for seeing patterns and making informed decisions. Companies can track sales by zip code or bar code and can see how well sales fliers versus email coupons work.
They can note that size-12 shift dresses in paisley prints sold out in days in Indiana, but lingered for months on the discount racks of Brooklyn clothing boutiques.
It’s dependable data, for the most part, assuming things stay more or less the same over time.
But more recently, as companies contemplate global competition with the click of a mouse, the hazards of just-in-time delivery, fluctuating consumer demand and razor-thin profit margins, they’re not content to rest on past information. They want to predict the future, as accurately and quickly as possible. Thus the rise of predictive analytics.
I like my beer in cans
Predictive analytics allows companies to make refined decisions or plans, using the same sort of sophisticated analysis. For example, it lets corporations see that not only are paisley dresses a hit in the Midwest, but that its customers over 40 in Indiana love coupons in the Sunday paper, while the same customer in the same demographic in Ohio only wants the dress she bought online shipped for free. And that twentysomethings from Brooklyn who shunned those dresses at full price will snap them up at a twenty percent discount, provided that they are stocked next to the cowboy boots and leather jackets for ironic layering.
But the same tools that make predictive analytics so powerful – and precise – are also the reason that some corporations may shy away. These tools can require dedicated programs and even databases, specialized programming skills to input complex formulas and specific knowledge of advanced mathematics and statistics to input and interpret the correct data, barriers that can seem insurmountable to businesses that need to use their resources in the broadest and most efficient ways.
So for a long time, the added costs of such analysis outweighed its benefits. But these days, powerful new software allows even small and lean businesses to cash in on the untapped gold mine of information at their fingertips. They take the math out of the mathematical analysis, providing ways to integrate predictive analysis into multipurpose business tools that offer point-and-click formulas and allow broad data mining that draws from a company’s existing databases. Which means that even small and mid-sized corporations can take look to the future, rather than just reflecting on the past, all without breaking their quarterly budgets.
So when the nephew from Brooklyn shows up at the barbecue with his new girlfriend who interns for that social media start-up in SoHo and she’s wearing a paisley dress and hipster boots, you’ll not only know why – but also that she’s probably dying for a kale martini.
More Business articles from Business 2 Community: