Forecast Accuracy

[Aju Appukuttan]

Dear ED,

 

         How can we calculate the Forecast Accuracy in sales

 

 

Aju

[truetrue]

Based on your forecast sale and actual sale. If you have forecasted
for 100 and your actual sale is 50. Then your accuracy would only be
50%. Monitoring your Accuracy can improve stock movement and inventory
level. Prevent OOS situation leading to lost sale.

Have a nice day.

[S M Patnaik]

I had tried to address this a couple of days back.
Forecast accuracy (FA) is calculated in a band. It could be +/- 20% or
+/- 25% or any other band that the organisation thinks proper. If +/-
20% is the band then against a forecast of 100 units of an SKU, sales
more than 80 and less than 120 is considered to be an accurate
forecast. if the sales is less than 80 or more than 120, then the
forecast is not accurate. Now if there are 10 SKUs and 6 out of them
have sales figures within +/- 20% band then FA is 60%. One can
calculate FA by following steps mentioned below.
- In an excel work book put the sales offices on the column heading
and SKUs in the row heading.
- Keep three rows for each SKU
- For each SKU enter the forecast number for each sales office and do
the same for actual sales.
- Then calculate the sales to forecast % in the third row of each SKU.
- Copy the forecast % numbers of each SKU and sales office to another
sheet.
- Calculate number of incidences where the forecast can be accurate or
inaccurate. This is typically = No of SKUs X No of sales offices.
- Calculate number of forecasts which fall between the accepable band
of +/- 25% or whatever you have decided.
- Divide the no of accurate forecasts by no of incidences. You have
the forecast accuracy.

[S M Patnaik]

Hi Aju,
Forecast accuracy (FA) can be measured at different levels of
aggregation. You would have number of SKUs, brands, sales offices,
distribution points, slaes managers, geographical zones and many such
variables. FA can be measured for all such combinations. Such
combinations would determine the level of aggregation. Typically FA is
calculated based on an predetermined band of +/- 20% or +/- 25%. For a
band of +/- 20% if the sales is 80 or 120 against a forecast of 100,
then FA is 100%. One can calculate FA ( I will explain how) at
different levels
- On a country level if the sales forecast for the month was 100 and
the actual sales is 105, then FA is 100% - This is the highest level
of aggregation.
- One can work out FA for brands at a national level. This is also the
highest level aggegation.
- One can do the same on a value basis for North, South, East, West
Zones. This could be a secod level of aggregation.
- One can do it by sales office level, distribution point level and so
on. These are lower level of aggregation.
- The lowest level of aggreagtion could be SKU wise sales office wise
FA calculation.
Pl note that FA becomes highest at highest level of aggregation. It
tends to become lower and lower as one moves down in the level of
aggregation.

Method of calculation of FA
- If you are calculating SKU wise sales office wise FA, then in a work
sheet put all the sales office in the column heading and SKUs in the
rows. For each SKU allocate three rows. One for forecast, one for
actual sales and the third one for sales as % of forecast.
-Enter all the data.
- Copy the third row data to another sheet.
- No of incidences where one can forecast accurately = No of data
cells = No of SKUs X No of sales offices. If there are 10 SKUs and 20
sales offices, then number of incidences = 10 X 20 = 200
- Put a formula to cound number of sales having value higher than
79.999% and lower than 120.0001%. (Assuming that a band of +/- 20% is
taken as accurate forecast).
- Divide that number by the no of incidences.
- You have the FA.

In case you need furhter clarification, you may send me a mail and I
will explain further on the subject and methodology.
Regards
Sukhendu Patnaik - sukhendu...@gmail.com

On Feb 11, 1:05 pm, Aju Appukuttan <aju.appukut...@gmail.com> wrote:

[S M Patnaik]

[Aju Appukuttan]

 

Thanx for u r support
 

[ISMAIL ABBASS]

DEAR ALL
am really interested with this study or experience
so would you please any one of you both send the excel sheet as i really cant understand all the written above
lol