Think about your com­pe­ti­tion for a moment. What have they got that you don’t? That’s the age-old ques­tion, isn’t it? Per­haps the bet­ter ques­tion to start with is this: what assets does your orga­ni­za­tion cur­rently have that could be bet­ter managed?

In a ground-breaking study of ana­lyt­ics com­peti­tors, Thomas Dav­en­port, President’s Dis­tin­guished Pro­fes­sor of Infor­ma­tion Tech­nol­ogy and Man­age­ment at Bab­son Col­lege, and his col­leagues at the Work­ing Knowl­edge Research Cen­ter stud­ied 32 orga­ni­za­tions that “made a com­mit­ment to quan­ti­ta­tive, fact-based analy­sis.” In other words, these orga­ni­za­tions have taken obvi­ous and prac­ti­cal steps toward Big Data man­age­ment. These orga­ni­za­tions are data vir­tu­osos. They’re sim­ply excel­lent at han­dling their data. Dav­en­port assessed:

At such orga­ni­za­tions, vir­tu­os­ity with data is often part of the brand. Pro­gres­sive makes adver­tis­ing hay from its detailed pars­ing of indi­vid­ual insur­ance rates. Ama­zon cus­tomers can watch the com­pany learn­ing about them as its ser­vice grows more tar­geted with fre­quent purchases.”

Is the sta­tus of data vir­tu­oso attain­able for your orga­ni­za­tion? Absolutely.

The activ­i­ties that made those orga­ni­za­tions pow­er­ful ana­lyt­ics com­peti­tors today are the same activ­i­ties that can make your orga­ni­za­tion an ana­lyt­ics com­peti­tor, too. Dav­en­port explains three key marks found in these ana­lyt­ics giants:

  1. Top man­age­ment had announced that ana­lyt­ics was a strate­gic priority.
  2. The orga­ni­za­tion had more than one ini­tia­tive involv­ing com­plex data and sta­tis­ti­cal analysis.
  3. The orga­ni­za­tion man­aged ana­lyt­i­cal activ­ity at the enter­prise (as opposed to the depart­men­tal) level.

In this three-part series, we will look at break­ing down this seem­ingly insur­mount­able task into prac­ti­cal steps that will help your orga­ni­za­tion begin to move closer to man­ag­ing Big Data in bet­ter ways.

Step 1 | Cre­ate a Clear Plan

This step starts with busi­ness objec­tives and leads to busi­ness out­comes. This seems intu­itive and obvi­ous, but many orga­ni­za­tions have yet to link their busi­ness objec­tives to their busi­ness out­comes as they relate to har­ness­ing Big Data. In the same way that strate­gic plans can cre­ate a com­mon pur­pose and a com­mon lan­guage through­out the orga­ni­za­tion, the same phe­nom­e­non can occur with a Big Data plan.

Be Sure to Include the 3 “Great Plan” Components

In “Big Data: What’s Your Plan?” the McK­in­sey Global Insti­tute writes that any great plan will include three key ele­ments: data, ana­lytic mod­els, and tools. These three ele­ments are defined as follows:

1.  Data

How will you assem­ble the data that you already have? How will you inte­grate it across teams, depart­ments, or func­tions? You need to think through those ill-devised silos that exist within your orga­ni­za­tion. How can you break down walls, inte­grate essen­tial func­tions across teams, and more? Where is col­lab­o­ra­tion nec­es­sary? Select just two areas to get started in.

Best Prac­tice: Harrah’s

For exam­ple, Harrah’s, named an ana­lyt­ics leader in an early study (2006) by Dav­en­port out of the Work­ing Knowl­edge Research Cen­ter, chose to focus on mar­ket­ing and cus­tomer ser­vice. As a result, they focused much of their “ana­lyt­i­cal activ­ity at increas­ing cus­tomer loy­alty, cus­tomer ser­vice, and related areas like pric­ing and pro­mo­tions.” Gary Love­man, Harrah’s chief oper­at­ing offi­cer (who would even­tu­ally become its CEO), began work­ing on the “data” com­po­nent of the plan by bring­ing in “a group of sta­tis­ti­cal experts who could design and imple­ment quan­ti­ta­tively based mar­ket­ing cam­paigns and loy­alty pro­grams.” That’s cer­tainly one way to start: hire out­side experts and bring them inside.

If, in read­ing this, you are already feel­ing ill-prepared and over­whelmed, let’s pause for a moment. In “Com­pet­ing on Ana­lyt­ics,” Dav­en­port wisely coun­sels: “You will have to under­stand the the­ory behind var­i­ous quan­ti­ta­tive meth­ods so you can rec­og­nize their lim­i­ta­tions.” How­ever, “if you lack back­ground in sta­tis­ti­cal meth­ods, con­sult experts who under­stand your busi­ness and know how ana­lyt­ics can be applied to it.” There’s noth­ing wrong with ask­ing for out­side help.

2. Ana­lytic Models

Can you iden­tify key areas or func­tions in which ana­lyt­ics will cre­ate value for your orga­ni­za­tion? In those cases, what key staff will need to use these solu­tions? As you scale up, or increase the use of ana­lyt­ics, how will you mon­i­tor con­sis­tency, accu­racy, and more? How will you man­age and assess qual­ity? The McK­in­sey Global Insti­tute report explains it this way:

A plan must iden­tify where mod­els will cre­ate addi­tional busi­ness value, who will need to use them, and how to avoid incon­sis­ten­cies and unnec­es­sary pro­lif­er­a­tion as mod­els are scaled up across the enterprise.”

Best Prac­tice: Walmart

In Davenport’s study, Wal­mart is listed as a sup­ply chain exem­plar, along with Ama­zon and Dell. These orga­ni­za­tions use sta­tis­tics and mod­el­ing for func­tions such as sim­u­lat­ing and opti­miz­ing sup­ply chain flows and reduc­ing inven­tory and stock-outs. “The most pro­fi­cient ana­lyt­ics prac­ti­tion­ers don’t just mea­sure their own navels – they also help cus­tomers and ven­dors mea­sure theirs. Wal-Mart, for exam­ple, insists that sup­pli­ers use its Retail Link sys­tem to mon­i­tor prod­uct move­ment by store, to plan pro­mo­tions and lay­outs within stores, and to reduce stock-outs.”

3. Tools

For this part of your plan­ning, it’s imper­a­tive that you work through the ques­tion: build or buy? You must sit down and weigh the pros and cons of build­ing your own sys­tems in-house or pur­chas­ing solu­tions from out­side sources. What are the trade-offs?

No mat­ter what you decide (to build or buy), the key here is intu­itive, user-friendly tools. The tools that you choose must be acces­si­ble and under­stand­able for the every­day end user. Yes, your teams and staff will be trained in ana­lyt­ics (at the most basic level), but most of your key staff won’t oper­ate at the expert level. Keep that in mind. Select or build solu­tions that enable your staff to work smarter, not harder. Again, the McK­in­sey Global Insti­tute explains:

What’s needed are intu­itive tools that inte­grate data
into day-to-day processes and trans­late mod­el­ing out­puts into tan­gi­ble busi­ness actions: for instance, a clear inter­face for sched­ul­ing employ­ees, fine-grained cross-selling sug­ges­tions for call-center agents, or a way for mar­ket­ing man­agers to make real-time deci­sions on discounts.”

For exam­ple, data man­age­ment plat­forms (DMPs) or cen­tral­ized intel­li­gence engines are often a good start­ing point for orga­ni­za­tions look­ing to opti­mize their audi­ence or customer-facing processes such as mar­ket­ing, loy­alty pro­grams, cus­tomer ser­vice, and more. Accord­ing to For­rester Research, these essen­tial tools cre­ate “a uni­fied view of the cus­tomer, regard­less of channel.”

It All Starts with a Plan

I know I promised seven steps and I’ve only deliv­ered one. Those next six steps will fol­low in the com­ing weeks. For now, focus­ing on plan­ning your Big Data man­age­ment plan is a great start in mov­ing your orga­ni­za­tion forward.