Supply Chain Management Blogs by Members
Learn about SAP SCM software from firsthand experiences of community members. Share your own post and join the conversation about supply chain management.
Showing results for 
Search instead for 
Did you mean: 
In today’s fast moving world of growing customer expectations, shrinking lead times, reduced profit margins and inventories, Demand Planners have a big responsibility to act as the first gatekeeper to be able to foresee changes in Demand Signals and accurately make predictions for future. The margin for error is continuously decreasing, as any errors cause a major rippling effect on downstream processes. For example, a change of forecast bias by 5% may induce the following risks;

  • Reduced customer service levels

  • Increased lead times

  • Improper mix of inventory levels leading to possibility of stock outs or over stocked DCs

  • Added costs due to firefighting production builds and expediting fulfilment (air freight)

To cope up with this fast-paced supply chains, new metric must be adopted by demand planners all over the world. One of those is Forecast Value Add. I am going to take you thru details of Forecast Value add and explain how we can use this to improve the forecasting process. I would also be explaining examples using SAP Integrated Business Planning.

What is Forecast Value Add?

Forecast Value Add, more commonly referred to as FVA, is metric to measure the performance at any 'level' of the forecasting process. The 'levels' can be process step, participant or planning characteristic levels of the process. FVA helps in analyzing if value is being added, or worse being destroyed, at individual stages of the process.

To illustrate this further, let’s take a pretty standard example of a forecasting process with steps involved in the monthly review process as follows;

Fig1: Example Forecasting Process

With the exception of the initial Data Preparation stage, the forecast is changed at every stage of the process. So, the forecast that is created in Step 2 (Statistical Forecast) may be different from Final forecast agreed in Consensus Review.

In the traditional world of measuring forecast performance, the accuracy is usually only calculated on a single key figure - i.e. the final forecast. The intermediate steps, e.g. the statistical forecast, sales reviews etc. are usually ignored from any formal KPI measurement. However, FVA emphasises that we should measure accuracy at different levels and compare against each other.

Naïve Forecast

Before I explain the FVA concept, I will first introduce you to the concept of a Naïve Forecast. Naïve Forecasting is simply a method to use the last period's actual (or a simple moving 3 month average) as this month’s forecast, without any adjustments or statistical algorithms. However, a Naïve Forecast is not an effective solution for the majority of businesses as this would lead to lot of buffer stock driven by the volatility of sales. But, a Naïve Forecast can form a base to compare or measure your other forecasts against, with the assumption being that the forecast created with advanced statistical models, complemented with human intelligence should be better than a basic Naïve Forecast.

Comparing with a Naïve Forecast allows businesses to understand whether efforts spent on creating an advanced forecast is adding value and if so how much value is being added.

Fig2: Comparing different forecasts

As you can see above, I am comparing forecast accuracy for Naïve Forecast, Statistical Forecast, Sales Forecast and Final Consensus Demand.

So, what does above chart illustrates –

  1. For Euro Trade Wholesalers (Germany) – Naïve Forecast is worse than the Statistical Forecast, but better than both the Sales and Final Consensus Demand forecasts. We are better of using the Statistical Forecast for this Customer Group and should exclude this Customer from further manual interaction.

  2. For Ruhr DIY, we can see the Statistical, Sales & Consensus Demand forecasts are all worse than the basic naïve forecast, so the valuable time and effort spent on the process has actually removed value from a basic naïve value.

As we can see from the examples above, this helps in analysis of the products and ability to focus on products at different levels of process to understand where value is being added. Also, since the process would include less number of items at each step, the review becomes more manageable, rather than going through all products during the review.


You have seen the FVA done at Customer Group level in Figure 2 above - similarly FVA can be done at any level. We can take say Product Group level as per Figure 3 below and identify if value is being added or destroyed for products within that Product Group.

Fig3: Comparing different forecasts by Product Group

Similarly, the FVA can be done at combination of levels as well - Customer-Product Brand level to understand whether there is any value being added / destroyed for the customer for a specific product group. Using the SAP IBP analytics app, we can have a drill down feature which allows us to drill down to individual products within the product group and then we can explore what this FVA means at product level.

Fig4: Comparing different forecasts by Product

We could also do FVA analysis on Forecast Bias as per Figure 5 below;

Fig5.jpegFig5: FVA can also be used to look at Bias


I have shown how we can create FVA at different product/customer dimensions, but some organizations may argue that this may or may not be correct way to measure their individual organizations business. So, FVA can also be customised to record the accuracy based on your organizations principles. Examples;

  1. We may calculate accuracy or bias on lagged forecast. Say we compare last month signed forecast with lag 1, lag 2 forecast or use lag 2 forecast for different stages of process

  2. We may use rolling 3 months average data for calculating the FVA accuracy and FVA bias comparing to Naïve forecast

  3. We may also financialise the bias and compare how much difference it makes in terms of comparing FVA

  4. Segmenting the products into ABC, XYZ to create a FVA with different KPI targets for each segment

Fig6: FVA with Segmentation


Some of the benefits of using FVA analysis;

  • Easy to understand concept and highlights where the focus should be on trying to improve it

  • Enables measuring value added in terms of avoidable error and cost and then the ability to slice and dice by either Product, Customer or Location dimensions (hence focusing efforts where there is the greatest scope of improvement)

  • Simple metric which makes individual step, participant and level accountable for adding value to the process


At Olivehorse we have our own fully built and integrated SAP IBP system with role based dashboards and KPIs. If you want to see how FVA and associated dashboards can help your organisation, or you want to see FVA dashboards in action, contact me for our free IBP taster session.

I do hope this blog was helpful, please do share your views.
Labels in this area