skip to Main Content

Making BI Count BannerThe Counting

Early in my BI career, I learned that telling friends and family I worked in “Business Intelligence” wasn’t very helpful. Most found the phrase meaningless and a few were relieved (or disappointed) to learn it did not mean I was involved in corporate espionage. So, I developed other go-to answers: “I help companies analyze data”, or “I help companies turn data into knowledge”, or “I help companies count.”

This last answer, while a bit facetious, was frankly true for many of my clients. The explosion of mobile and networked technologies meant that more and more business activities were leaving a digital trail. Some groups, like finance, had always dealt with data, but for other business functions, the ability to measure and analyze and be held accountable for their actions at this level of granularity was new and at times daunting. To ease their anxiety, I’d say something like “Let’s start by making sure all the things that count are being counted.”

I still use that line but in a different context. Now it’s a caution to people who have become enamored with the latest buzzy innovations trumpeted by bloggers and consultants and marketers without really understanding if that’s what their business needs.

I created a simple graphic to help structure these conversations:

Mathematical Complexity

Yes, it’s a two by two matrix, and yes, those are generally deployed to convince you to spend money getting to the upper right. That’s certainly where the excitement is. Machine Learning! Bayesian Analysis! Predictive Analytics! That’s what’s getting everyone excited, and no wonder, because those techniques do offer some truly revolutionary potential for organizations to better plan and execute and improve.

But are you ready for Predictive Analytics if you aren’t already pretty good at estimating next year’s results based on three years of historicals (Forecasting)? Are you ready for Machine Learning if you don’t have a good grasp of seasonality and other key sensitivities (Correlating)? The answer may be yes. Maybe you can leapfrog right to the cutting edge. But most organizations are better off mastering the fundamentals first, and that means getting the job done in that lower left quadrant.

Here’s a simple two-question test:

  1. Are you accurately collecting and aggregating the data that describes your core business activities?
  1. Can you quickly get that data in the hands of the people who can use it to make a difference?

Many organizations don’t realize how much power there is in being able to answer both those questions with a confident ‘yes’. Or they overestimate the cost and complexity to do it. Focusing on modern, self-service tools and processes that empower people and reduce time to knowledge is always a good way to make your BI investment count.

Microsoft’s platform is the world’s most fluid & powerful data toolset.  Get the most out of it.

No one knows the Power BI ecosystem better than the folks who started the whole thing – us.

Let us guide your organization through the change required to become a data-oriented culture while squeezing every last drop* of value out of your Microsoft platform investment.

* – we reserve the right to substitute Olympic-sized swimming pools of value in place of “drops” at our discretion.

Tom Williams

Tom has always gravitated toward data and analysis, whether working in finance, marketing, or even as a freelance writer. He joined a Wall Street bank right out of college, discovering a love of spreadsheets that eventually landed him a spot on the Excel engineering team at Microsoft. After leaving Redmond he worked at two start-ups and as an independent consultant before trying to make it as a writer.

Even as a writer, his greatest success came in technical subjects. So he took that as destiny and returned to his roots in Business Intelligence and Data Analytics, eventually reuniting with his old Excel colleague Rob Collie to open a Seattle office of PowerPivotPro. Tom’s hands-on work as a BI consultant coupled with his broad business background makes him uniquely suited to assist clients at all levels of scale.

This Post Has 9 Comments
  1. Tom, this is an excellent post. The grid summarizes the world of data analytics so concisely. Thanks!

  2. Love this article and chart. Thank you for sharing. And I’m going to try out your “I help companies count” line on friends and family because I have had the same problem in conversations.

  3. Thanks for the great article and for creating this 2×2. This matrix should guide a lot of conversations around BI/analytics roadmaps.

    One mistake I see a lot of people making is assuming that greater mathematical complexity produces more value.

    Many businesses are ignoring a lot of value-creating opportunities by over-investing in predictive analytics without a clear idea of how they will generate objective value from such investments. Meanwhile, they completely ignore opportunities to reduce costs or increase revenue by focusing on mere “counting.”

  4. The urge to bypass the fundamentals is challenge for many because it is considered to be the “boring” stuff. As this points out so well is that if you don’t master fundamentals first the results of the more “advanced” techniques will likely be highly flawed. Great article!

  5. I feel the missing piece in this discussion is lack of quality/reliable data. You can have the best Power Pivot/BI reports but if its rubbish in its rubbish out. Everyone wants to get to the pretty charts, but if the quality is not there you must go back to basics. One way I have found to address this is as you say go back to basics and build a model line by line, include checks everywhere and summarise these on 1 overall net check number/document every step until it makes sense. This is where the value add not the pretty charts!!!

  6. Would be great if P3 could run some articles on the different types and give some methodologies.

    What are some examples of predictive analytics (outside forecasting) and how would use the tools to go about them?

    That would be really helpful for a newbie like myself. 🙂

  7. Definitely without solid basic’s in capturing & recording business actual performance data, any analysis forward looking insight potentially floored.
    Love the simple and expressive 2×2 model.
    Would be interested in future articles regards Predictive Analytics.
    Really enjoyed the article ~ thanks David

Leave a Comment or Question