For even those com­pa­nies who are most advanced in online test­ing & opti­miza­tion, one large part of their sites usu­ally remains rel­a­tively untouched: cus­tomer sup­port.  I’ve been try­ing to fig­ure out why the response is usu­ally lack­lus­ter at best when I sug­gest run­ning tests on cus­tomer care and knowl­edge­base pages.  One rea­son seems to be the lack of own­er­ship within mar­ket­ing.  Ever heard of a cus­tomer sup­port marketer?

Another rea­son may be because it’s harder to quan­tify the ROI.  There’s no straight­for­ward way to mea­sure how effec­tive a test is.  We could look at how many peo­ple don’t go to the phone num­bers page, from which we would extrap­o­late that they found the answer online.  In con­junc­tion with that, we could look at the num­ber of page views to under­stand how deep within the bow­els of FAQs and arti­cles the cus­tomer had to nav­i­gate.  Does that really tell us if what we’re chang­ing on the site is work­ing though?

Let’s say we then track them into the front-line of phone sup­port: IVR (Inter­ac­tive Voice Response), the auto­mated phone tree that we are all too well aware of.   That could be our neg­a­tive con­ver­sion event.  But the big dol­lars in cus­tomer sup­port reside in the call cen­ter.  Most com­pa­nies would be per­fectly happy if cus­tomers could ser­vice them­selves through the web­site and IVR.  So per­haps IVR is the first stage in the fun­nel, and live cus­tomer sup­port is the sec­ond.  As some­one who lived and breathed call cen­ter met­rics and per­for­mance in a pre­vi­ous life, I am still blown away by just how much every inter­ac­tion with a call cen­ter agent costs, and how much can be saved by reduc­ing AHT (aver­age han­dle time) by even one sec­ond.  We’re talk­ing a sav­ings of mil­lions of dol­lars for any enter­prise company.

If those fig­ures aren’t enough to get peo­ple to start test­ing, what about the straight­for­ward goal of reten­tion?  Reten­tion is not just about send­ing out emails to exist­ing cus­tomer lists in order to upsell.  As some­body who sells software-as-a-service on a sub­scrip­tion basis, I know that a large part of my job resides in keep­ing my clients happy.  Actu­ally, it’s about more than just being happy.  I want them to be con­tin­u­ally delighted by how much test­ing & opti­miza­tion improves their work, whether it’s in the imple­men­ta­tion, the test design, the strat­egy and exe­cu­tion, or the stel­lar results they get to evan­ge­lize.   Tra­di­tional cus­tomer sup­port can be that as well.  It’s not most of the time, but hey, when it’s good it’s great.

The cus­tomer sup­port sec­tion is a per­fect place to test because your cus­tomers are most likely try­ing to shout the goal of their vis­its at the top of their lungs.  Look at what brought them to your site in the first place.  Did they do a search for a spe­cific prod­uct in con­junc­tion with “help”?  Prob­a­bly safe to assume they own it.  Are you show­ing the most read arti­cles for that prod­uct on the page then?  Or bet­ter yet, are you deliv­er­ing the answer to their ques­tion upon land­ing?   If you can pro­vide the answer, then you have a huge oppor­tu­nity to cross-sell and upsell on top of it.  For exam­ple, maybe I’m look­ing for tips on how to extend the bat­tery life on my lap­top.  If you can take me to a page with sug­ges­tions, I might also be enticed by the extended-life bat­tery you’re rec­om­mend­ing for me.  Addi­tion­ally, you could also pro­file me as this spe­cific prod­uct owner so that you can tar­get and sell me more effec­tively once I go back to the shop­ping side of the site.

My guess is that most peo­ple are think­ing that these ideas would be nice if mar­ket­ing and oper­a­tions talked to each other, but right now they’re on dif­fer­ent sides of the house.  I’m think­ing that it might be time for oper­a­tions to get their own set of web marketers…