hammer, bolt and nut

During the Thanksgiving break I had the pleasure of reading two great articles written by Tatyana Shumsky in the Wall Street Journal.

The first article, Stop Using Excel, Finance Chiefs Tell Staff, was published on 11/29/17. It caused such a response that the very next day, Shumsky wrote a followup piece called Finance Pros Say You’ll Have to Pry Excel Out of Their Cold, Dead Hands. As of this post, both articles garnered a combined 600 comments.

As a data governance practitioner for over a decade, these two articles and subsequent comments reminded me of conversations I have had many times with business and technical people. What I took away from these articles has nothing to do with the limitations of Microsoft Excel from a capability and integration standpoint nor the possibility of retiring Excel from your organization (which will never happen BTW). Rather, it is related to understanding why people ‘WANT TO’ or ‘HAVE TO’ use Excel in the first place.

Understanding the root cause as to why Excel or Access (or any business-managed system) is being used is the key to changing behavior and enacting change in the organization. Data Governance teams should be working with business people to inventory and understand the use of these business-managed systems for all the reasons brought up in Shumsky’s articles. Outlawing the use of these tools will only drive people to other solutions or the use of rogue laptops outside the jurisdiction of the technical team. People will find a way.

So, let’s start here. (This is by no means an exhaustive list.)

Why do people ‘HAVE TO’ use business-managed systems such as Excel?

  • Functionality – Other system(s) do not have the capabilities as required by the business.
  • Analytics – Other system(s) do not have the reporting capabilities as required by the business.
  • Access – People are not granted rights to use the system or related analytics.
  • Time – People cannot get the data in time to act.

For countless reasons not outlined here, the lack of functionality and analytics in these system(s) force the business teams to improvise their own solutions. All too often this means not partnering with their technical counterparts. If you’re Les Stroud, then improvisation is a good thing. But if you’re doing analytics for the annual report, then improvisation is a bad thing. With proper planning and partnership between business and technical people, these types of issues can be mitigated.

The more insidious reasons that are harder to remediate are related to why people ‘WANT TO’ use business managed systems.

  • Change Management – People do not like change. They want the familiarity and comfort of processes, tools, people, and content even as systems or business models change. I have worked with and for companies that did not utilize functionality, or worse, they intentionally disabled it, when implementing a new system simply because no one could envision doing anything in a new way.
  • Trust – People do not trust the content that they are being asked to use. They may not trust it for legitimate reasons. Or they may not trust it for an irrational reason — such as a corporate urban legend! For these people, trust is only created when they control all aspects of the data. I recall working with someone making daily, multi-million dollar decisions in an Access database which resided on their laptop and was one spilled cup of coffee away from disaster. When I mentioned that this data was available in the data warehouse, the person answered without hesitation, “The only data I trust is my data” And that was that.
  • Data Sorcerer – there is always that one person who has the ‘magic’ data consisting of data ingredients from disparate sources that get mixed together in their data cauldron. The Data Sorcerer will tell you not to bother with the data in the data warehouse because their data is better. The problem here — and this is a big, big problem — is that the Data Sorcerer works outside the jurisdiction of IT which means their data cannot be replicated (or verified) by anyone. And yet, the Data Sorcerer becomes the ‘trusted’ source of this content.

These issues are real as evidenced by Shumsky’s articles. Data Governance teams should be attacking these issues daily. The problem is not the use of Excel and other business-managed systems. The problem is failing to understand WHY people use these solutions and then doing something about it.