I am not the eas­i­est per­son to live with.   I think my hus­band would be the first to tell you.  And while I am at home on bed rest in the final stages of preg­nancy – I think the high main­te­nance fac­tor could be a lit­tle higher than nor­mal, even for me.   But my cur­rent state of health has not made me shy about what I ask him to help me with.  If he goes to the gro­cery store on my behalf, he is usu­ally armed with a thor­ough, defined gro­cery list.  If he goes to Old Navy to pick up some t-shirts for our daugh­ter, he may call me a few times to ver­ify his strat­egy before pur­chase.  Heck, he does not even go to Sub­way with­out clear require­ments on what kind of sand­wich I want – my mood can change depend­ing on the day and while the Ital­ian sub worked well a few weeks ago, the only thing that will work today is a veg­gie on whole wheat.

You may be ask­ing, what’s your point Kiran?  (Other than the fact that your hus­band is a very, very patient man).  And Kiran, why the inor­di­nate fas­ci­na­tion with lunch meat? (see pre­vi­ous post “No Free Lunch …”).

My point is that if I ulti­mately want a deliv­er­able (in this case my six inch sub) that meets my expec­ta­tions, I bet­ter be able to com­mu­ni­cate those expec­ta­tions clearly.  I can have no qualms with my hus­band if he brings me home some­thing that I am not in the mood for if I don’t tell him what it is in the first place (My hus­band will tell you I have not always fol­lowed this logic in the past, but I am going to ignore those sit­u­a­tions for now to make my point).

So it’s inter­est­ing to me that I am find­ing more com­pa­nies who are shy­ing away from the very basic, fun­da­men­tal step of defin­ing require­ments as they imple­ment expen­sive and expan­sive web ana­lyt­ics solu­tions.   Or that I am encoun­ter­ing with greater fre­quency, very large com­pa­nies that are will­ing to com­press this very nec­es­sary piece of any suc­cess­ful imple­men­ta­tion down to a bare min­i­mum in order to “kick start” the good stuff.

I am not sure what the dis­con­nect is and I cer­tainly don’t under­stand the rush to imple­ment with­out clear require­ments.  Is there a grow­ing com­fort level with the knowl­edge that ven­dors bring to the table that makes com­pa­nies more will­ing to hand the reigns over?  Is there a belief that these things will “work them­selves out” as you imple­ment?  (I am not ask­ing rhetor­i­cally.  If your com­pany or some­one you know exhibits any of the symp­toms described above — I would like to hear your take on this).

The real­ity is that bypass­ing this nec­es­sary part of the soft­ware devel­op­ment life cycle can lead to very expen­sive reper­cus­sions for an enter­prise down­stream.  Here are just some of the rea­sons you can’t afford to take this process for granted.

Where is your voice?

The most obvi­ous piece of this is that your company’s vision must be accu­rately reflected in what­ever solu­tion you imple­ment.  Soft­ware alone does not give you answers – no mat­ter how much you spend on it – if you can’t clearly define the ques­tions you are seek­ing to answer and the goals you are look­ing to drive.  Often­times, while com­pa­nies may have some sense of the larger, dri­ving goals (increase con­ver­sion, max­i­mize ad spend), the indi­vid­u­als who have been given the man­date to imple­ment the solu­tions are not given proper guid­ance from the exec­u­tive level down on the strate­gic vision.  Under­stand­ing the exec­u­tive vision from the top down will help you for­mu­late the tac­ti­cal busi­ness require­ments you should be going after.

Some­times you think you know, but you just don’t until you get it all down and you have sig­noff from your exec­u­tives.  Most often, a mid-level man­ager will be tasked with the imple­men­ta­tion piece.  They will work ardu­ously with the ven­dor on imple­ment­ing their per­ceived vision of the ana­lyt­ics solu­tion, only to find that there is a big dis­con­nect with what they ulti­mately deliver back to the exec­u­tive level.

So find your voice – and make sure it’s in har­mony with the voices of your exec­u­tives or you can fall flat.  Don’t just rely on the fact that you can com­mu­ni­cate and ver­bal­ize their needs.  Doc­u­ment every­thing and have con­fir­ma­tion from exec­u­tive level down.

And remem­ber the famous words of Patrick Swayze in Dirty Danc­ing — “Nobody puts (insert YOUR  name here) in the cor­ner.”  Ensure that your voice gets heard and that it con­tin­ues to res­onate with the doc­u­men­ta­tion and com­mu­ni­ca­tion sys­tem you build around it.

No blue­print

The game of tele­phone is one we are all famil­iar with.  Brandy told Eddie, Eddie told Jane and Jane told Shawn.  By the end of it “Michael wants a pair of new jeans” becomes “Michael wants a pair of pants.”  No big deal, except if you are the one tasked with buy­ing Michael his present — now he has a pair of khakis he really didn’t need and they sit in the back of his closet.  Poor Michael.

The more explicit, the more detail you can pro­vide – the bet­ter you are arm­ing every­one to ensure you get what you are look­ing for.  It also makes it that much clearer to iden­tify areas that the solu­tion in ques­tion may not work and that alter­na­tives may need to be iden­ti­fied or explored.  But with­out that blue­print, you are more likely to open up the pos­si­bil­ity of error in inter­pre­ta­tion and deliv­ery.  Why take the chance? Like Michael’s khakis that are sit­ting in the back of a closet, you don’t want the solu­tion you are work­ing so hard to build to become rel­e­gated to the back of people’s minds because it never was relevant.

No legacy

It’s a fluid world.  Change is every­where.  No mat­ter how much you think your boss will be at your com­pany for­ever or no mat­ter how much you want to believe that your key ana­lyst just loves work­ing for you – you can’t count on con­ti­nu­ity just through your resources alone.

I was recently on a call with a For­tune 500 com­pany where I was fight­ing resis­tance on this exact thing – try­ing to get com­mit­ment to have the busi­ness own­ers actu­ally doc­u­ment require­ments.  My con­cern was dri­ven by many fac­tors, but mainly that they were putting them­selves in a posi­tion of weak­ness ver­sus a posi­tion of power.  They brought up the fact that they expected to have the same staff on hand and also the same Omni­ture con­sul­tants to ensure the legacy remained.  Con­tin­gency and risk plan­ning were obvi­ously not being con­sid­ered here.

Lis­ten up every­body – don’t ever make this mis­take!  If it’s the bur­den of col­lect­ing and doc­u­ment­ing the require­ments – out­source it.   We all know where our vul­ner­a­bil­ity points in our orga­ni­za­tions reside due to key resources that we just can’t afford to lose.  Turn your vul­ner­a­bil­ity into an asset and just make sure you doc­u­ment the liv­ing day­lights out of every­thing.  Peo­ple leave, peo­ple get sick, peo­ple have acci­dents, soft­ware changes own­er­ship within an enter­prise.  Don’t ever put your­self in a posi­tion of weak­ness by tak­ing any of these pos­si­bil­i­ties for granted.

How do you eval­u­ate your deliverables?

If you haven’t doc­u­mented clearly what the man­date dri­ving your end deliv­er­ables are, how do you assess the suc­cess of your imple­men­ta­tion?  What are you mea­sur­ing?  Is it just that it pro­duces some cool, nifty new dash­boards and reports that you never had before?  (I really hope you are defin­ing suc­cess as more than that).

I was recently on a call with the direc­tor of an ana­lyt­ics orga­ni­za­tion and he was express­ing dis­sat­is­fac­tion with some of the tools out there today.  He was quick to point out which solu­tions would not work for him, but could not express WHY very clearly.

If you can’t express that vision or the “why”, you are set­ting your­self up for dis­ap­point­ment right out of the gate.   Per­haps, ulti­mately, there truly is no solu­tion that will work for this gentleman’s com­pany.  I find that hard to believe.  What I do believe is that he was respon­si­ble for an orga­ni­za­tion that could not pri­or­i­tize and item­ize all their require­ments in a way that would ulti­mately allow him to deter­mine which soft­ware would meet the major­ity of their needs.  Every­thing seems so over­whelm­ing until you can fac­tu­ally present it on paper.  It gives you the oppor­tu­nity to fil­ter out the “noise” and the mean­ing­less dri­vel that some­times deters this process from hap­pen­ing.  When I say “noise” — I mean all the require­ments on the “nice to haves” that nobody was ever able to present a log­i­cal, busi­ness dri­ven expla­na­tion for hav­ing.  You will find that once you elim­i­nate that dis­trac­tion, it becomes much eas­ier to hone in on what’s important.


So yeah – as you may have guessed — I am pas­sion­ate about this and I really think it’s one of the most avoid­able pit­falls to any imple­men­ta­tion (ana­lyt­ics or not) that a com­pany pro­ceeds down.  It may not be the eas­i­est thing and it may take time to really develop a strat­egy for how you col­lect and doc­u­ment every­thing and ensure all crit­i­cal stake­holder involve­ment – but it is the foun­da­tion for every­thing you do to drive your imple­men­ta­tion forward.

You know that Allen Iver­son clip where he goes off on his coach for being mad about his skip­ping prac­tice too many time with his Detroit Pis­tons team­mates?  “I mean, we’re talk­ing about prac­tice.  Prac­tice peo­ple!”  Well instead of the word “prac­tice”, which was said 20 times through­out that dia­tribe,  replace it with the word “require­ments.”   Yeah — that’s how strongly I feel about this.

Don’t under­es­ti­mate the value of the time you invest in this piece of your project.  If my hus­band won’t buy me a sand­wich with­out clear require­ments, why would you arm your ana­lyt­ics team and ven­dor with any­thing less?