Demand Plan contents

Stability through global supply chains is unlikely in the foreseeable future, so structuring a Demand Plan appears to be difficult. However, the Plan is an important input to Sales & Operations Planning (S&OP) that provides a forecast of market demand, sales income, unit sales and gross profitability for the products on offer. The Demand Plan is provided with data and information from Demand Management elements:

• Demand Sensing: The past is not an indicator of the future, is a statement that has been extra true since COVID. Sensing identifies buying trends through analysing data generated within each sales channel. The most extensive analysis is across a range of external data, including: retail point of sale (POS); card/mobile phone sensors; unstructured statements posted on social media; web site ratings; reviews of products and services; warranty and product returns via eCommerce and physical sales outlets and weather forecasts.
• Demand Shaping: Incentive programs designed to increase customer and consumer demand for products, including: new product releases, upgrades to current products, product promotions, price reductions, commercial buyer (or trade) incentives and sales force incentives. If uncoordinated, these actions increase uncertainty through the supply chains, assisted by Demand Shifting. This distorts demand signals and increases costs in supply chains (especially air and express road freight) and includes: ‘end of month (and quarter) rush’ to get finished products out of factories to meet production targets; shipment of orders this month that should be delivered next month and ‘stuffing’ the sales channels with additional product that is sold on delayed payment terms.

Demand Planning contains the forecast, which is a time-phased estimate of demand or sales that covers a future period. Forecasts vary, depending on the needs of the user and influenced by:

• Why needed?: a retail grocery supermarket chain requires a plan for future assortment mix and shelf space by SKU. A CPG (consumer packaged goods) manufacturing business requires planning and scheduling for promotions, supply, manufacturing and distribution.
• Planning horizon: a grocery retailer could forecast sales for a maximum of 6 months. A CPG business could forecast sales up to 18 months ahead (time to install additional capacity)
• Frequency: a grocery retailer would most likely review the plan weekly. The manufacturer should have a monthly (or four weeks) Sales & Operations Planning (S&OP) process
• Unit of measure: a grocery retailer forecasts the product package. The manufacturer uses the outer shipper or tonnes/litres for forecasts of a ‘product family’ used in S&OP

Forecasts are wrong

The market demand for a product consists of variables that interact dependently, independently and interdependently, creating outcomes that emerge – they cannot be planned. So, forecasts of future sales will never be ‘correct’. But if the forecasts of sales will always be wrong, then presenting a single period forecast should not be done, especially when senior executives may assume that the number is ‘correct’.

Instead, the Demand Planner (or similar) provides a range that identifies an optimistic and pessimistic probability of the forecasts, together with an explanation. At this stage will enter the challenge of bias in forecasts. There is a threat in Marketing and Sales that those who want a certain outcome e.g. a sales mix that gains highest commission, will tend to be together, convincing others that what is wanted will occur and more likely to overlook possible failure. This is Confirmation bias, which can influence Demand Planners when they are too close to the action. 

Within the S&OP process, the forecasts is by ‘product family’. The Plan is then de-constructed to individual SKUs within the product group or product line, based on percentage of prior sales within the ‘product family’. These calculations are weighted by the Demand Sensing results, new product features and options and future promotions. The process may be called ‘proportional profile planning’ (PPP).

This structure of forecasts assists Procurement. The purchased items with long lead times can be ordered at the optimistic forecast level. This may result in periodic excess inventory of particular items, which is preferable to being out of stock. Items that are easy to obtain and on short lead times are ordered at the pessimistic forecast level. To provide cover for the potential of attaining the optimistic sales forecast, provisional orders are raised (with cancelation clauses), so the quantity required of an item can be increased at short notice.

Forecast accuracy

Forecasts using a range can improve forecast accuracy – especially if the ‘forecast error’ that occurs between forecast and actual sales is also measured. This is NOT a measure of performance, but a statistical term to identify the difference between planned and actual situations. 

The most common measure of ‘forecast error’ is the Mean Absolute Percentage Error (MAPE). This is the difference between the actual and forecast sales (change any minus signs to a plus) and divided by actual sales. Then sum the ‘errors’ and divide by the number of observations. 

Forecast accuracy will differ, depending on the product category identified in the extended ABC classification (to be discussed in a future blogpost). For products that have a high sales volume with a predictable demand, the MAPE could be between 10 and 15 percent. However, products with low sales volume with an inconsistent demand pattern can expect a MAPE of up to 30 percent or higher; so, using MAPE as a target without understanding the drivers can provide biased outcomes.

An additional measure is the Tracking Signal, which assists inventory control at the SKU level. It is calculated by dividing the cumulative variation between forecast and actual sales for the number of periods under review by the standard deviation of the variances. An acceptable tracking signal for products with a high sales volume and a predictable demand is up to 4. Products with low sales volume and an inconsistent demand pattern should be reviewed when the tracking signal exceeds 7.0.

Also, an important measure is Forecast Value Added (FVA), which considers forecast performance. The starting point is a naïve forecast; using last year’s actual sales for this year and comparing to the new actual. Additionally, a comparison can be made against each sequential step in the forecasting process. Surveys have indicated that the value from using sophisticated forecasting applications and processes is questionable.

The information required in S&OP from the Demand Plan is the most likely direction of product sales and their magnitude. The forecast data will never be accurate, but that is not essential. The need is to identify when the ‘turning points’ in the sales activity are most likely, so that tactical plans can be adjusted. That said, there must be confidence that the forecast data is reliable.

Please note: Learn About Logistics is taking a short break. The next blogpost will be on Monday September 8. Thank you for reading the blogposts and forwarding them to your colleagues.