Spend Matters welcomes a guest post from Paul Blake, a sales manager at GEP.
Earlier this week, I met with a prospective client to talk about spend visibility, big data. and all that good stuff. With the best of intentions of course, and having applied a very great deal of resource and brain power (and inevitably cold hard cash), this client had built a program of data capture, manipulation, and transformation, based on all sorts of rules and manual interventions all designed to make sense of their global spend patterns. I got the impression that they felt that had created a bit of a monster that they'd yet to get anything really, you know, useful out of.
Perhaps it's a familiar feeling for you as well?
“What we need now,” they said, “is something to bolt on to the end of this process to make sense of it all, to give us the visibility we need.” Cue the little voice in my head: “problem is…you don't want to start from here.“
Making sense of a vast volume of data is as much about where you start from as about how you get there. Just as a reputable plumber might say "No! For pity's sake, don't 'start it off' for me!" a solution designed to make sense of bulk raw data typically works best if it's fed what it expects: bulk…raw…data.
If your spend data is in a bit of a global pickle, then there are absolutely folks out there who can straighten it out and craft the metaphorical silk purse from the sow's ear. When we start from the very beginning, it is a good deal easier to deliver the visibility and business intelligence required. Look at it this way: every piece of data that is "cleansed" out of your system is a piece of data you will never get back. And you wouldn't believe how important some of those little nuggets can be. A good deal of "signal" can be lost when correcting for "noise," to mix a metaphor.
Expending effort on tidying up your historical spend data yourself is mostly useless, and is far more costly than getting the experts in to do that for you. I would say that, of course, it's my job, but that notwithstanding, you should be focussing all that effort and expertise on the systems, processes, and solutions that change the very nature of the raw data that you're ADDING from now on. Stop feeding your big data pool with messed-up, chaotic information and make sure what’s going in is ordered and accurate.
The notion is to transform your operational purchasing landscape so your raw data looks just like data that has been through a spend analysis process - right from the get-go. Over time, the un-cleansed data will be steadily replaced with ordered, accessible, reportable data - available, at source, at the touch of a button.
The trick is knowing that you need to start from here, not there, and a smart investment in the right spend visibility solution could save a small fortune, and that, my friends, is no joke.
Finally a word of caution: beware the plumber who DOES say "can you just start it off for me?" This is, of course, another analogy that I'm sure doesn't need any explaining.
For more interesting thinking on procurement, visit the GEP Knowledge Bank.