Forecast Value Added


00:08 Introduction
00:25 What is the idea behind Forecast Value Added?
02:37 Where did this idea originate from?
05:36 What is the problem with FVA?
08:30 Why has FVA gained in popularity?
10:37 Are there any areas to which FVA can bring improvements?
14:04 Why do people find it so difficult to put their trust in science?
16:45 Are people getting used to the idea that it is not just one person doing the forecast?
18:44 How can you differentiate between consultants that are providing added value and consultants that are just producing metrics for the fun of it?
22:42 What is the core message behind today’s episode?

Description

Much like the saying, “a problem shared is a problem halved”, Forecast Value Added is a management technique that simplifies a forecast by splitting it into manageable chunks. On this episode of LokadTV, we discuss just how well this works and why decomposing a forecast can actually lead to more difficult decisions.

Forecast Value Added is a process that likely emerged in the 90’s or 2000’s. It essentially amounts to improving quantitatively the forecast’s accuracy by identifying the various steps taken. For example, a forecasting team produces the baseline forecast, the marketing team then steps in to adjust this forecast based on their insights, then the sales team will probably add their own layer of corrections, then the production team, etc. The Forecast Value Added then seeks to establish metrics to see whether these contributions have degraded the forecasting accuracy or not.

This may all seem well and good on the surface and even appear logical, but it simply doesn’t work. For starters, no science backs up this theory. The idea that you can have a statistical forecast that bounces from section to section in a company is simply lunacy. For example, the various winners of the “M” forecasting competitions have certainly never used such a multi-stage forecast.

It’s a process that may be a crowd pleaser, as it allows many different people in the organization to get involved, but it is unfortunately very far from being the most efficient way to forecast. We could compare it to a master chess player who has to let a team hold a vote to decide what the next move should be.

To conclude, we talk more about the resurgent popularity that Forecast Value Added has been experiencing in the past couple of years (and why this faith is misguided), the importance of looking outward rather than inwards and just why manual interventions on forecasts bring more harm than good.