Analytics for Finance-Part IV of IV-Future

Analytics for Finance-Part IV of IV-Future

Welcome to this article that is part of a four-article series for Finance where we explain how to build a culture of analytics, how to build insights and foresight, and sharing of some cases. In combination with the Business Partnering Institute, our aim with the series is to further push the finance function in its on-going transformation towards being a strategist with the business.

Part IV of IV: The Art of Turning Data to Insight Explained

We’ve reached the highlight of our Analytics Series where we describe how Finance can drive more return on investment on our analytics efforts by building a new culture, creating insights, and now making the future happen through foresight. Foresight is described in business literature as…

“…the ability to make a successful outcome happen…”

…hence while it does include forecasting and predictions through advanced analytics, it is just as much the ability to make the predicted outcomes happen or not. Good outcomes we want to maximize and bad ones we want to mitigate or eliminate. Foresight is what binds the WHAT MIGHT happen and the HOW to do it, together.

Let’s make the future happen!

Through predictive analytics you can today, given all that you know, make predictions about how your business will develop should you take no actions to change it. On top of that you can then build different scenarios and make models to predict WHAT MIGHT happen if you were to attempt making certain changes.

However, just because something is predicted to happen and seems like the best outcome, you cannot allow yourself to autopilot all your decisions. There are more considerations to make like…

  • Do we have enough resources e.g. capital, people, etc.
  • Do we have the right capabilities e.g. people with experience of having done this
  • Can our systems and processes deliver it

All of this shouldn’t arrest you from becoming prescriptive about HOW to make the desired future happen. It’s simply a sanity check of the predicted outcomes.

What the future looks like

The future of analytic decisions is powerful and let us demonstrate with two examples.

First, a Fortune 500 company, used advanced analytics to predict the propensity of sales deals to close in a quarter. From history extracted in the CRM system, the analytics built a profile of the characteristics that indicate when a deal will close, and when it will not.

In a particularly interesting example, the analytics revealed that while a specific customer was actively interested in buying, the salesman did not have the characteristics to close the deal with that profile of deal size. When the deal did not close in the quarter as predicted, the salesman on the account was changed the very next quarter and the very next month the sale closed.

Now, the original salesman was not incompetent but simply had not the skill to close a large deal. This was not seen by the salesman’s manager but was visible to the analytics. The salesman continued with the company and was successful to meet his quota from the deals whose profiles are a match for him.

In a second example of a NYSE transportation company, it used AI to forecast revenue throughout its major and minor business units. This forecast was then distributed to the business managers who adjusted the forecast based on their knowledge of the future. The combination of AI and humans produced 97% to 99% average monthly forecast accuracy over 18 months.

As part of this forecast process, there was a separate statistical verification of the human adjustments to the forecast to assure the “reasonability” of the adjustment.

Specifically, the adjusted forecast was compared to another statistical forecast that correlated economic indicators to determine those that lead demand. The leading indicator forecast was then compared to the adjusted forecast. If the adjusted forecast was not within the statistical confidence interval of the lead indicator forecast, then the adjusted forecast was not considered “reasonable”.

When there was a discrepancy, the manager was asked to explain the nature for his adjustment. A salesman’s optimism of a deal closing or a manager’s pessimism of a downturn in a region would be rejected. Adjustments, outside the bounds of reasonability, needed a firm foundation based on definitive knowledge of the future.

The ability to deliver high accuracy forecasts over 18 months led to better planning and control of operations. Consistent performance assured the company’s higher valuation, as reflected in its PE multiple. The company received about an 18% higher multiple vs. a broad market basket of companies.

Leading finance functions make the future come true

Of course, Finance is not alone in making the future happen and rarely does the finance function execute many of the actions that make the desired changes happen. However, Finance is participating end-to-end, from problem identification through scenario modeling to following up on actions and optimizing performance.

As described above, predictions can be used to help sales operations put the right resources on the right accounts to maximize sales potential, or AI and leading economic indicators to deliver long range high forecast accuracy for better planning and control of operations.

This is what leading finance functions can do! They’ve stopped reacting to what has happened around them and instead become proactive making the changes happen. Finance can create a lot more value when decisions are based on data analytics, as opposed to, gut feeling. Are you ready to become a part of the leading pack!?


This article is a collaboration between the Analytic Intelligence Institute led by Robert J Zwerling and Jesper H Sorensen, and the Business Partnering Institute based in Denmark led by Michael Bülow and Anders Liu-Lindberg.

Upcoming Events

This concept of a “college” style summit has a coordinated syllabus that provides a roadmap for companies to develop a culture of data driven decisions through the use of analytics. The program methodically takes participants through structured tracks of lecture and hands-on workshop to deliver a step-by-step approach to implementing analytics. For more information and registration please link to: