Insight. The one true currency of a finance professional. If we don’t have any insights to share with our business stakeholders, then we’re not helping improve decision-making. A lot has happened on the insight front in recent years. We’ve talked about going from insights about past events to insights in real time. Few finance functions have landed there yet, but they all have the same aspiration. However, we must go beyond even real time and talk about “future time,” or in other words, predictive and prescriptive analytics. First though, let’s clearly define what “insight” is.
Insight is when you share something with your stakeholders that 1) they didn’t know and 2) will help them make better decisions.
This is a very broad definition, but insights can come in many shapes and forms, and they can be financial and nonfinancial. If the information you share can satisfy the two criteria, then you have “insight.” Now let’s dig into what’s happening in the insight space in the disrupted finance value chain.
From past to present to future
The main topic in the past few years has been to increase speed to insight. In the past, it could easily take two to three weeks after month-end before finance had anything to say about the numbers. Even worse, there was only a monthly cycle: no weekly or daily cycles. That meant finance could only contribute a very limited amount to improved decision-making.
So, we went on a journey to increase speed to insight by shortening the month-end cycle, creating weekly and daily cycles for more operational KPIs. Month-end is now almost automated and can be done in a matter of days. Weekly sales numbers are available at the end of the week. Plant performance is looked at daily.
However, real time is no longer enough either. Why? Because it still says something about past and present performance. We know from economic theory that the past and the present are not necessarily good predictors of future performance. Something else is needed!
That’s why we’ll now start to talk about “future time” insights. This includes predictive analytics: What would happen if we did XYZ? Even better is prescriptive analytics: What should I do to improve performance? A machine learning, or even true artificial intelligence, overlay on big data would make this possible. Let’s look at the standard workweek and see how this could function.
It’s very likely that the report you received on your desk Monday morning already had elements of predictive and prescriptive analytics included. However, now the real work starts. Now you must present to and discuss the insights with your business stakeholders in the weekly management meeting.
You will discuss past performance and what led to this performance in an effort to qualify the recommendations already put forward by the analytics engine. You won’t conclude anything in the meeting except agree that you and a select group should spend Tuesday working out what to do. On Tuesday, you work with the insights provided and run a series of what-if scenarios as well as brainstorm on other potential ideas. In addition, you discuss if the competencies you currently have in the company can deliver on the various solution options. Depending on how complex the problem is, this exercise could extend into Wednesday until you reach a recommendation (which we’ll talk more about next week).
Jumping a bit ahead, as a team you’ve decided on what actions to take, and you now rely on your analytics engine again to get the results of your efforts. What worked and what didn’t? More importantly, why did or did it not work? Those are all insights that’ll be used in the next weekly cycle.
Two parts to insights
As you can see, there are two parts to insight. One relates more to analytics or analysis wherein you derive the insights. This can be more or less advanced as we discussed in this and last week’s articles. The second part is when you go and interact with your stakeholders to present the insights and discuss their views on the issue. It’s through this more social process that the insights are given real meaning and context. This is the process where decisions must be improved further. Otherwise, we might as well leave running the company to the algorithms and robots.
Here’s what I want you to take away. Insights should move from past to present to future, and you must lead the process of qualifying them in a discussion with your stakeholders. These must go hand in hand, because if all you do is send the insights via e-mail or rely on your stakeholders to gather them themselves from the dashboards, then you might as well not show up to work!
Are you ready to go to future time, or do you want to deal with the past and present for a while longer? The sooner you make the shift, the sooner you can start improving decisions even further.