Treat data like it’s scarce and expensive

I’m a big pro­po­nent of plat­form think­ing. Cre­ate a plat­form, open it up respon­si­bly, and enable peo­ple to cre­ate using the platform’s capa­bil­i­ties. It’s a way to lever­age tech­nol­o­gy that’s dif­fer­ent than in the past because it requires IT to let go of con­trol (but not of gov­er­nance) and let oth­er peo­ple do the work.

This new atti­tude requires a change in the sto­ries told, the pri­or­i­ties set, and the val­ues to which an orga­ni­za­tion aspires.

It often seems like I’m push­ing rope though, when I try to con­vince peo­ple of this future. I know a lot of resis­tance has to do with incen­tives and mea­sures of suc­cess. For exam­ple, ROI is not always a good mea­sure. If you’re build­ing a plat­form you have to invest and it can’t be the first project to use the plat­form that bears the full bur­den on their denom­i­na­tor (for the math-chal­lenged, I’m refer­ring to the “I” in ROI).

Sim­i­lar­ly, there can’t always be a return at the lev­el of gran­u­lar­i­ty defined by the project. In any case, I’m wan­der­ing.

I want to tell you how to think about the data in your orga­ni­za­tion.

It’s not meant to feed a silo any­more. In fact, orga­ni­za­tions that lock their data in silos are doomed. Yes, doomed.

Benchmark to manufacturing

If you don’t read and lis­ten to every­thing Horace Dediu cre­ates, you’re miss­ing out.

Read his lat­est arti­cle on Tesla’s man­u­fac­tur­ing capac­i­ty, and what we can infer from the elec­tric vehi­cle mar­ket. I’ve read it three times.

Here’s a quote, and I think it’s the most rel­e­vant part, but it’s hard to tell because the whole arti­cle is so damn good:

So why, after a few years of Mod­el S pro­duc­tion, is Tes­la run­ning at one fifth the speed of a BMW or Toy­ota plant? One obvi­ous answer is that the com­pa­ny is not run­ning 3 shifts but per­haps 2 or even 1. But why would they shut down the line when demand is infi­nite? If it is finite, is it real­ly only 25k at most per quar­ter? That’s a pal­try num­ber con­sid­er­ing the size of the mar­ket they oper­ate in. BMW alone ships 2 mil­lion vehi­cles a year. Tes­la X/S top­ping out at 100k a year sug­gests a seri­ous lim­it to their oppor­tu­ni­ty.

Once the man­u­fac­tur­ing capac­i­ty is built, it should be run to the ground (assum­ing demand exists) so as to max­i­mize yield and reduce cycle time.

Data locked in silos is not “run to the ground” and you can’t pos­si­bly tell me that data is not of inter­est out­side of the silo. I know I’m shift­ing ver­ti­cals, but look at the way agri­cul­ture uses data… no mat­ter how much they have, they want more.

To max­i­mize your data you’ve got to break it free from appli­ca­tion silos; if you don’t, you’re com­peti­tor will. Click To Tweet

We need to think of data as scarce and expen­sive just like man­u­fac­tur­ing capac­i­ty. Once it’s “cre­at­ed” it needs to be used as much as pos­si­ble.

Consider Amazon’s legendary AWS genesis

Jeff Bezos said, make every­thing an API. And so it was. Now we have AWS a huge busi­ness in its own right.

Why wouldn’t any com­pa­ny cre­ate a data plat­form? No one has ever said “Ok, I have all the data I need” — right? In fact, we strug­gle mov­ing data from spread­sheets and sales­force into pow­er­point to make it look pret­ty. Of course, once the data is copy-past­ed, it’s sta­t­ic and has an expi­ra­tion date.

A plat­form would make data eas­i­ly acces­si­ble. While some data is more com­plex to secure (like PII), much is rather straight-for­ward. Espe­cial­ly when the right gov­er­nance pro­ce­dures are in place.

I don’t know what would be cre­at­ed in every instance, but I guar­an­tee that com­pa­nies would be sur­prised by the inno­v­a­tive­ness of their employ­ees. In fact, with data avail­able, they’d be more moti­vat­ed to cre­ate inter­est­ing per­spec­tives… per­spec­tives that I have no doubt would be action­able.

Don’t let it happen, make it happen

Here’s the thing. Often the project (get­ting some­thing run­ning) is the final step from a team per­spec­tive, when it should be the first. A data plat­form wouldn’t be suc­cess­ful when it’s run­ning, only when the data is being used wide­ly.

That implies that to make a plat­form suc­cess­ful, some evan­ge­list (at least the begin­ning) needs to go out and pro­mote data avail­abil­i­ty and lis­ten to unmet data needs to improve the plat­form. To break down even more silos.

I would argue, this needs to be an aggres­sive proac­tive activ­i­ty. It’s not that data APIs are cre­at­ed to make it easy to share data… but that data shar­ing oppor­tu­ni­ties to improve the busi­ness have to be sought after. Because like in man­u­fac­tur­ing, if you’re not being as effi­cient with your man­u­fac­tur­ing capac­i­ty as you could be… some­one else will and will win.

If you’re not using the data every­where it might be use­ful, your com­peti­tor will. They’ll have an insight, or a capa­bil­i­ty that you don’t sim­ply because they’re max­i­miz­ing the util­i­ty of a resource they’ve already paid for, while you’re just let­ting it stay in-silo.

I’m fin­ish­ing up this post as I watch some of the ear­ly pan­els as I pre­pare for my own at CA’s 2017 Gov­ern­ment Sum­mit, and I heard a great quote from (delight­ful­ly artic­u­late) pan­elist Tar­razz­ia Mar­tin a gov­ern­ment tech­nol­o­gy expert that might help (my addi­tion in paren­the­sis):

Data needs to be treat­ed like an enter­prise asset (like a man­u­fac­tur­ing plant) #API #SmartC­i­ty Click To Tweet