Multi-Channel Analytics: Where do I begin?

A cou­ple of weeks ago, I wrote a byline arti­cle for DMNews​.com about the approach to take when plan­ning a multi-channel ana­lyt­ics ini­tia­tive.  I thought I’d share it here as well:

The num­ber of chan­nels avail­able to mar­keters for reach­ing their tar­get audi­ence has been grow­ing dra­mat­i­cally over the years.  It was not long ago when the only options mar­keters had were print, radio or tele­vi­sion.  Today of course, busi­nesses have a whole mul­ti­tude of chan­nels they can use to get their mes­sage across to cus­tomers; and on the flip side, cus­tomers also have a whole range of chan­nels with which to com­plete trans­ac­tions – and these aren’t nec­es­sar­ily the same ones.

Busi­nesses now real­ize that they need to employ mul­ti­ple chan­nels in order to effec­tively reach their tar­get audi­ence.  How­ever, merely hav­ing mul­ti­ple chan­nels does not trans­late into a true multi-channel strat­egy.  Busi­nesses need to avoid the costly and inef­fec­tive “every­thing to every­one” approach when putting together their chan­nels. They need to define a true multi-channel strat­egy where each chan­nel plays a core role and has a defined value propo­si­tion for the cus­tomer experience.

In order to do this, busi­nesses need to be able to under­stand cus­tomer behav­ior and pref­er­ences.  They need valu­able infor­ma­tion on the dif­fer­ent cus­tomer attrib­utes and how these affect pur­chase behav­ior, prod­uct attrib­utes, sales pat­terns and chan­nel pref­er­ences.  This under­stand­ing can no longer be achieved by look­ing at only one chan­nel – busi­nesses need to bring all this infor­ma­tion together from all the dif­fer­ent chan­nels in order to see the true big picture.

But be warned: bring­ing together all data from all chan­nels all at once is too ambi­tious a task and will likely result in fail­ure.  Busi­nesses should start with one eas­ily mea­sur­able chan­nel and aug­ment it with data from just one other chan­nel.  As tech­nolo­gies for web ana­lyt­ics have matured in the last few years, this is a log­i­cal place to start.  Most com­pa­nies with a web pres­ence should already be ana­lyz­ing cus­tomer behav­ior on the web.  The short term goal should be to aug­ment this analy­sis by adding data from one other chan­nel – for retail­ers, this would be in-store cus­tomer behav­ior and pur­chase pat­terns derived from POS sys­tems.  For other ver­ti­cals with heavy call cen­ter use, such as finan­cial ser­vices, travel or telecom­mu­ni­ca­tions, analy­sis of call data merged with web data would pro­vide quick returns.

By fol­low­ing this approach and then later adding addi­tional chan­nels once the two-channel analy­sis has yielded ben­e­fits, busi­nesses are more likely to develop a suc­cess­ful multi-channel strat­egy to retain and grow its customer-base.