The Study And Practice On Demand Forecast Of Lucent Qingdao

In order to stay the hot competition, the enterprises are forced tohead towards core competence oriented strategy rather than completely self-dependency.This brings opportunities to the enterprises and challenges as well. The change ofrelationship between the buyer and the supplier forces the enterprises to put greateremphasis on demand planning management. And the enterprises have to face thechallenge of optimizing the supply chain output with well leveraging of demandplanning management. However, many enterprises are still in an embarrassed situationof high forecast error. Taking the worldwide industry——telecommunicationsindustry——as example, the study that this thesis does regarding demand forecast as asystem will hopefully benefit the demand forecast practice in other enterprises withinthe telecommunications industry or in other industries as well.It brings up a new methods combination——CPA (Critical ProductAttribute)——of managing demand forecast for products with heavy customization,dramatic market demand fluctuation and long purchasing lead-time of raw material. TheCPA method gets market demand forecast using Time Series, generates product demandforecast using Multiple Line Regression and establishes the translation method frommarket demand to product demand via correlation analysis. CPA method not onlyapplies traditional statistics but also enrolls the product design information and productlife cycle information into the forecasting process that enables a more realistic forecastresult and eases the problem tracking if any. The CPA method helps the China SystemIntegration Center of Lucent Technology to reduce its forecast error from 150% to 52%in one and a half years.Comparing with the old approaches, Multi-Aspect Forecast Accuracy (MAFA) isable to measure forecast accuracy from multiple angles and at multiple layers. Whenapplied together with Critical Product Attribute, MAFA could provide the managementwith information that could be used to adjust business strategy and the operational levelwith the information used to make decision on forecast adjustment on each forecastitem.