Spend any time in IT and you're bound to hear theexpression, "You can't manage what you can'tmeasure." Metrics is the mantra of many data center managersor network architects. How else do you know who consumes whatresource how often and how that affects things such as WANperformance?
For chief marketing officers (CMO), however, useful metrics arehard to come by. Yes, like most executives, the CMO is buried instatistics: Same-store sales, inventory data, production data, CRMdata, transaction data, sales and promotions data and so on.There's lots of data, but little of it will tell you what yourcustomer will do next or why they did what they did, like abandonan online shopping cart.
Transaction data doesn't improve the customer experienceunless you combine it with some other data to learn something newand useful. The same is true if customer data is collected atmultiple touch points-Web, brick and mortar, mobile-but isn'ttied together on the back end or accessible to call center reps inreal time.
Take everything you know about your customers and prospects, though, and combine that with newdata sources ranging from social media to location and weatherdata, and you enter the world of big data. What matters in thisworld isn't so much the individual data points, but where theyintersect. It's the mashup that will give you the insights youneed to improve your next-best offer or understand why customers dowhat they do.
Use Big Data, Gain Competitive Advantage
"The ability to collect vast amounts of data on individualconsumers-their consumption habits, their preferences, theirinteractions with the company-and then analyze those data sets forpredictive behavior and proactively apply those insights both toyour existing customers and to customers coming into your callcenter or your website or your agents office, [that's] thebasis of competitive advantage in the future for the CMO becauseyou can provide a better experience," says Matt Jauchius, CMOof Nationwide Insurance.
While this is happening at a certain percentage of companies,most aren't embracing big data or spending the dollars requiredto achieve the state of the art.
A recent, frustrating encounter with a new cell phone providerproves this to be true, even for multibillion dollar players intech-centric industries rife with data that should make the customer experience memorable forall the right reasons.
After ordering a smartphone online, only to have that orderlost-which I learned only when the "overnight" shipmentnever arrived-I called customer service. That didn't go well,either, so I went to a retail store. I spent 45 minutes giving thesame information over and over to a very capable sales rep, who inturn had to deal with a rather inept back office, all to get one working smartphone when I had originallyplanned to get two.
In short, I had to work way too hard to give this carrier mymoney. Siloed product channels and data sets stifled cross-sell andupsell opportunities and left customer service reps unable toresolve problems or provide next-best offers or actions.
The moral of my story: Multichannel marketing is great only ifyou don't experience a problem with one channel and try toresolve it via another.
Big data technologies such as Hadoop can flatten the data silosthat support each of these channels and then feed the results tohyper-fast in-memoryanalytics platforms. These systems could inform call centerreps in real-time that I, for example, placed an order online,called three times and had an outstanding unresolved order, atrouble ticket and a separate account for my office phone allcausing customer service headaches at the same time.
Get to Know Your Target Audience
This is precisely why Nationwide, a 90-year-old company withmany databases and myriad compliance obligations, is spendingseveral million dollars on bigdata initiatives, Jauchius says.
But this is just one side of big data that CMOs need toconsider. Among the advantagethat big data bring to marketers, perhaps the biggest are theability get out in front of customers and prospects and to conductmore effective predictive and prescriptive marketing, says ElanaAnderson, vice president of IBMEnterprise Marketing Management.
"Marketing has long been on a quest to get to theindividual," Anderson says. "Smart marketers...havebeen trying to get beyond the demographic for a long long time. Ifyou're able to address the individual at an individual level,if you're able to sense needs or meet needs before the customeris explicitly saying, 'I have a need', that requires BigData and analytics in order to get to that point. We're seeingtremendous value with uses cases around that."
How an individual company will getstarted with big data depends on its use case, industry,available data and factors that depend on the outcomes it'strying to achieve. In general, though, prescriptive marketingcombines longitudinal knowledge of your customer with their largerpatterns of activity. You then combine that knowledge with thebroader patterns that affect your business-geography, demography,weather, social media activity or anything else you need-to get amore complete picture of your target audience.
Don't Sweat the Small Stuff; Let Analytics Sort It Out
These patterns will provide the insights you need to reachcustomers in novel ways, says Olly Downs senior vice president ofData Sciences for Globys, a big data analytics provider to thetelecommunications industry.
Globys has been doing big data for more than 15 years, Downssays. The difference between today and years past, he says, is thevolume of data and number of data sources. That said, thecapability to capture, store and analyze that data has reached aprice point that makes big data ROI achievable for marketing.
"By applying machine learning to big data, it's doingthe discovery for you," says Downs. This means you can uncover50 marketing scenarios you would never cook up on your own."That's the power. It's not about any individualscenario that's discovered, it's about being able tosurface many of these scenarios and act on all of them in a waythat's dynamic and meaningful."
For example, Globys does a lot of work in the developing world,where the average prepaid mobile customer generates 29 pieces oftransaction data per day, in the form of SMS messages, phone calls,top-up requests and so on. (A Facebook user averages just three, incomparison, and a Twitter user less than one.) All that daily dataequates to a lot of upsell and cross-sell opportunities.
With its recommendation engine, Amazon is the poster child forthis type of big data marketing. But Amazon isn't doinganything you can't, says Gartner Research Director BillGassman.
Most marketing departments already use big data. It's justburied in their analytics engines or customer experience managementsystems, Gassman says.
"There's so much more data to play with, and it'sjust so much easier to play with," he says, noting that thead-hoc query have advanced significantly since the term was firstcoined. In fact, queries can be done so quickly, and soinexpensively, that Gassman tells CMOs, "Don't worry abouthow it's done; worry about what you're going to do withit."
Allen Bernard is a Columbus, Ohio-based writer who covers IT management and the integration of technology into the enterprise. You can reach him via email or follow him on Twitter @allen_bernard1. Follow everything from CIO.com on Twitter @CIOonline, Facebook, Google + and LinkedIn.
Read more about big data in CIO's Big Data Drilldown.