Don’t Do This! 7 Pitfalls When Deploying Analytics (Part II)

Pit­fall #5. Boil­ing the ocean

When deploy­ing ana­lyt­ics, you should aim to reach as broad as pos­si­ble across all your cus­tomer touch­points. You need to if you want to be con­sid­ered strate­gic. Your abil­ity to opti­mize and under­stand the cus­tomer life­cy­cle is directly related to how com­pre­hen­sive you can be with mea­sure­ment.How­ever, just because you go broad does not mean you should go deep. It is of much greater value to under­stand all your touch­points at a min­i­mum level that to under­stand a few touch­points at a max­i­mum level. I’m sure some peo­ple will dis­agree with this point vehe­mently – and I wel­come the feed­back. I’ve arrived at this con­clu­sion based on my own expe­ri­ence and this has worked for me and the clients I’ve worked with.

In the world of ana­lyt­ics, it’s bet­ter to go wide and shal­low, than nar­row and deep.For exam­ple, I talk with many peo­ple who want to mea­sure the day­lights out of some­thing like a shop­ping cart. They mea­sure every­thing: when some­one opens it, when some­one closes it, when they add or remove a prod­uct, the prod­uct size and color, whether they open a pop-up…all this data may be inter­est­ing, but it pro­vides them with a false sense of secu­rity because the gran­u­lar­ity of this data fails to reflect their broader cor­po­rate strategy.For exam­ple, their focus becomes incred­i­bly myopic and they become obsessed with minute obser­va­tions like why Fire­fox vis­i­tors con­vert at a faster rate than those that come in via Inter­net Explorer. And because they are focused on the weeds, they can’t see that their con­ver­sion rate for the broader site has been falling for the past 3 months and they aren’t doing any­thing about it​.In short, try­ing to “boil the ocean” keeps you from see­ing the big pic­ture and oper­at­ing on your core busi­ness goals.

Pit­fall #6. Mul­ti­ple ver­sions of the truth

Ana­lyt­ics suc­cess is all about build­ing a base­line for per­for­mance (your KPI trend), and try­ing new things to improve on this base­line. That’s it! That’s why I think it’s easy.  I know other blog­gers have argued that ana­lyt­ics is hard, but I’ve done this for a liv­ing and I can tell you that it’s not.  Sure, it can be hard – over time – to con­tinue to improve on your base­lines.  I’ll grant you that. But that comes after you’ve picked all the low hang­ing fruit and inno­va­tion becomes more crit­i­cal.  But to be fair, I can’t say I’ve ever met with a com­pany that has picked all the low hang­ing fruit. So from my per­spec­tive, it’s just not that hard.With that in mind, a crit­i­cal pit­fall occurs when cus­tomers try to use mul­ti­ple sys­tems to pro­vide this base­line for per­for­mance. In other words, they have mul­ti­ple ver­sions of the truth.  Most of you know what I mean.

Take a lead gen­er­a­tion site for exam­ple, like an auto­mo­tive man­u­fac­turer. They will often looks to total num­ber of leads, among other KPI, to deter­mine their suc­cess. But they will mea­sure this with 5 dif­fer­ent sys­tems. They have leads as mea­sured by Omni­ture, then leads as mea­sured by their data ware­house, leads as mea­sured by their email sys­tem, leads as mea­sured by their ad server, and leads as mea­sured by search. There may be even more sys­tems than that.The prob­lem with this approach is that you get mul­ti­ple ver­sions of the truth. And you waste your time try­ing to rec­on­cile all these dif­fer­ent sys­tems rather than try­ing to improve on your base­line.  Now – don’t get me wrong – you must make a best effort to under­stand why these sys­tems report dif­fer­ent results.  I’ve spent count­less hours doing this.  And often times when it comes down to it, there are just fun­da­men­tal dif­fer­ences in the mea­sure­ment approach (as we talked about in our approach on Data Migra­tion: Fools Erand).  Another exam­ple is with clicks.  A “click” as defined by an email provider will likely be dif­fer­ent than one pro­vided by Omni­ture which will be dif­fer­ent than that pro­vided by an ad server, which will be dif­fer­ent than a “click” reported by a search engine. That’s the real­ity of the world we live in, so just accept it. You can spend your time try­ing to solve this aca­d­e­mic chal­lenge, or you can spend your time improv­ing your busi­ness and beat­ing your com­pe­ti­tion. You decide.

You must have a sin­gle ver­sion of the truth, because your abil­ity to opti­mize your busi­ness is based on the rel­a­tive dif­fer­ence between points on the cus­tomer life­cy­cle. It is not based on the absolute rela​tion​ship​.In other words, if a search engine like Yahoo! says you have 1000 clicks, and another ana­lyt­ics provider puts you at 900 clicks, that doesn’t mat­ter as much as when you com­pare the 900 clicks as mea­sured by your ana­lyt­ics provider for a search cam­paign against, say, 700 clicks as mea­sured by the same provider for an email cam­paign. If you com­pare num­bers from the same provider against each other, you can see how dif­fer­ent cam­paigns are doing in rela­tion to each other. When you com­pare results from dif­fer­ent providers, you’re com­par­ing things not based on the same absolute rela­tion­ship. There­fore, your results are seri­ously skewed.

Pit­fall #7 – Not Teach­ing how to fish

In my own ana­lyt­ics career, some of the biggest gains we made from ana­lyt­ics were actu­ally from peo­ple out­side my ana­lyt­ics team. They came from other busi­ness units that we had trained to use ana­lyt­ics. Why? Because they often under­stood their busi­ness ques­tions bet­ter than any­one else, so they could inno­vate the most from the ana­lyt­ics data. In other words, they had the best con­text for the data​.To that end a crit­i­cal mis­take that peo­ple often make is not train­ing end users to be self-sufficient. Sure you can send users to pro­grams like Omni­ture Uni­ver­sity and con­duct inter­nal train­ing. That’s great and a crit­i­cal first step. But once you’ve done this, you should seek out the peo­ple that “get it” – and bring them into your inner cir­cle. When you iden­tify these power users and nur­ture them, they can become your great­est ally and drive some of the most sig­nif­i­cant gains you’ll real­ize from ana­lyt­ics – all with­out any incre­men­tal effort from you and your team. This might sound like a fairy tale – but it works – trust me, I’ve done it and I’ve helped other com­pa­nies like yours do it.

In Sum­mary

So there you have it. Seven crit­i­cal pit­falls mar­keters often fall into when deploy­ing ana­lyt­ics. Granted there are oth­ers – but if you go into your next deploy­ment with eyes wide open to these com­mon mis­takes – I have every con­fi­dence you’ll be more suc­cess­ful than ever before. If you’d like to talk about your unique sit­u­a­tion and require­ments, we’d be happy to do so. Just give us a call and we can work with you to realign your ana­lyt­ics deploy­ment with your strate­gic busi­ness require­ments and indus­try best prac­tices. It’s what we do best!