logisitik

Switch to desktop Register Login


Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 494

Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 494

Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 494

Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 494

Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 494

Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 225

Warning: Creating default object from empty value in /home4/cpc38/public_html/components/com_k2/models/item.php on line 494

Forecast time horizons

Rate this item
(1 Vote)

Here the objective is to establish the demand forecast for a company. But there are different time horizons or levels of forecasting:

  • Level 1 - The Strategic Plan: used to forecast on a 3-5 year time horizon the company strategy such as diversification, new products development, equipment investment or divestment, capital expansion…
  • Level 2 – Sales & Operation Plan: used to forecast over the next 6-18 months the customer demands and build the required capacity and material supply plan in accordance. Here the main decisions that could be taken are hiring employees, subcontracting, buying a new machine,…
  • Level 3 - Master Production Schedule: used to forecast up to the next 3-6 months demand to build the manufacturing schedule and material supply schedule as well as the shopfloor capacity loads. Here the forecast must be accurate enough to schedule the manufacturing operations properly and making adjustments if required.


The time horizons depends of the company industry sector and could be adjusted depending of the product life cycle. Those time horizons are absolutely not the same for a PC manufacturer or a nuclear power energy manufacturer.

There are many forecasting models that we further detail in the next articles, but the key criteria to choose a model are the following ones:

  • Historical data: Are relevant historical data available for a given product or product family? Was there a significant change in the market? Customers? Macro-economic changes? Economic crises or grow?
  • Required forecast accuracy: it is difficult to be accurate for one single item but quite easy for a family of 500 items, so issue is to find the appropriate level to get a suitable accuracy.
  • Accepted cost to build the forecasts: How long could I work on? How many people involved? Do I need to purchase market studies or economic reviews?


Bear in mind that by nature a forecast is uncertain. But this uncertainty - this is too often forgotten – can be estimated so must be taken into account!
There are many statistical models that can be used to calculate it as we will further details in next articles.

Moreover the more a forecast is aggregated and the more the time horizon is long, the more the forecast is accurate.

This means that either we should forecast on the long go with aggregated product families either we forecast per product with a very short time horizon. Otherwise even with the best statistic models, forecasting over 3-5 years on single product item will never work.


Last modified on Tuesday, 28 August 2012 08:26
Login to post comments

© 2009-2012 Logisitik - All Rights Reserved

Top Desktop version