Normalize Your Corporate Data. Experience the Same Truth


No alternative facts please -

In order to get the most out of Big Data and AI corporations must implement and nurture a data-driven corporate culture. Creating that culture takes time and abundant leadership to establish, nurture, embrace and grow. Becoming data driven is not just an issue of spending money in analytics software and graphical displays, it is a thoughtful, immersive and iterative process that depends on trusting the data. So how do we get to do that; trust?

As part of my blog series on how to “Transform Your Corporate Culture to Best Use Big Data”, here are three suggestions on how to develop trust in corporate data by normalizing it across the organization.


Raise your hand if you’ve ever been part of an interdepartmental meeting where various teams are jointly analyzing a production line, project or companywide program, and someone says “that is not what my numbers are saying”, or “that is not the budget we were turned over to execute”. You can lower your hand now. Every single time those words are uttered, there are tens to hundreds of man-hours wasted by teams debating, in public and private, who has the correct facts, trying to establish a baseline for “truth”.

This dynamic is neither efficient nor trivial. It is not efficient in that oftentimes, the differences needed to erode trust are rounding errors in magnitude, but demand hours of painstaking detective work to resolve. It’s not trivial in that said detective work is usually entrusted (as a reward of course) to your “rock stars” whom you should be keeping busy doing the critical work of the corporation. Rarely ever does anyone rounds the bases and updates the team to say “the error was…”, but you can count on the fact that during plenty future cooler “anecdotal Olympics” one department will say about the other “yeah, their numbers were all messed up”. And with that, trust in data is lost.

These situations happen when corporations forego the process of defining the dataset they need to collect, the databases of record where they will reside, and the formulas and methods that convert it into useful information, or business intelligence. While it is perfectly fine to allow corporations in early formative stages or initial growth mode to use the database tools that their key employees are comfortable using, it is absolutely insane not to normalize data corporate wide, and integrate these databases at key milestones of corporate life growth.

So on to developing trust in data -

First

Starting with your strategic plan and corporate goals in mind, identify the Key Performance Indicators (KPIs) that will provide the factual basis to direct corporate decisions and validate failure or success calls. Once you have them identified, define KPIs using mathematical formulas – not just words. Why is that? You see, Math is an exact science, and formulas are a universal language understood by all (ok, most). Wordy definitions combined with interpretation or ambiguity can lead to “alternative facts” making their way into your corporation, and again, trust dies with them.

Second

Review what data and in which form you must collect and store it for use. By defining the KPIs mathematically, you will know, as example, if storing distance traveled and time elapsed is more convenient or accurate for your corporation than storing speed or velocity data. You can calculate speed from time elapsed and distance traveled easily. You can go in the other direction as well, but some assumptions need be injected, and there again trust in data dies a bit more once again.

Keep in mind that your SMEs are that for a reason. In the absence of specific direction on how to collect and process the data, you may have as many formulas as individuals in your team or corporation. You will not be able to please everyone so consider input from your SMES, and publish a final decision so everyone uses the same identical formula, and source data to evaluate the KPIs.

Third

If you want to establish a data-driven culture; make source data is “king” and establish one database of record, through which all analytics will flow. If you have heard the terms “data lakes or data warehouses” you are in the correct zip code. They are closely related but not identical architectures. That said, they both attempt to establish one true source data set.

“But I use Deltek, Salesforce, Dynamics and INFOR databases in my corporation” – All the more reason to work on that data lake or data warehouse Johnny. The idea is not necessarily to delete any of these data sets, but rather to carefully research, understand and decide where from does the data that feeds the KPIs originates and how to feed it to the other data sets. In that process the decisions to “weed out” some databases will happen as a byproduct.

As I mentioned before, it takes substantial effort and leadership to establish, nurture and grow a data-driven culture. In order to develop trust in the data corporate leaders must choose with zero ambivalence to operate under one common set of source data, one data warehouse/lake, and one published and common mathematical KPI definition set to set your baseline of “truth”. Without this, every corporate BI tool will never be fully trusted, and your data-driven culture will likely be ever doomed.

////////////////

As a reminder, I recommend that you read the following Harvard Business Review (HBR) articles and NewVantage Partners’ survey. Good stuff for sure and they will certainly help “normalize” the data as you read these blogs.

HBR – “Is Your Business Masquerading as Data-Driven?”

HBR – “Companies Are Failing in Their Efforts to Become Data-Driven”

NewVantage Partners – “How Big Data and AI are Accelerating Business”

2 views0 comments

Recent Posts

See All