After over a century of growth and stability the New Zealand postal service has been changed forever by forces well beyond its control – privatisation, deregulation and the internet. The effect has been a significant and rapid decline in the volume of mail and a steady increase in the volume of parcels albeit in a very competitive market. NZ Post needed a computerised model of their mail collection, sorting, transport and delivery systems – a model which would allow the organisation to model the financial impacts of multiple business options and make the right decisions for future business growth and profitability.
They came to Orbit Systems to develop the model.
CASE STUDY
New Zealand Post provides the Government-owned postal service for New Zealand. It processes nearly 1 billion items of mail every year and delivers mail to around 1.84 million delivery points. It employs 2500 postmen and postwomen. Its on-time delivery of mail is one of the best in the world at 95.5%. The NZ Post transport network has 158 trucks that travel over 14 million kilometers a year.
NZ Post needed to redesign its traditional business model which had originally been developed as a postal service which was labour-intensive and designed to provide, under its deed of understanding with the Government, levels of service which were no longer relevant to the market. While NZ Post still dominated the mail processing and delivery market, this was rapidly diminishing year on year as a result of the internet and email communications replacing post. Parcel post however was increasing with the increasing number of long-distance transactions. It requires different modes of collection, sorting, transport and delivery systems.
Orbit Systems was asked to develop a model of NZ Post’s mail collection, sorting, transport and delivery systems which would allow NZ Post to model the financial impact of various scenarios.
Orbit first had to identify what type of model would be the most effective for NZ Post. A model cannot include every aspect of a complex system so it must be less complex than the reality on which it is based.
In this case the number of sorting centres at various locations, types of mail, transport distances, volumes and many other variables all contributed to make this a particularly large problem.
Orbit tackled the challenge in a number of ways. First we limited model options to exclude clearly unlikely solutions. For example by limiting the number of sorting options for each centre to two or three we considerably reduced the complexity.
Then we analysed the data to see where existing knowledge could assist us. By analysing the data we could see what type of mail had and had not been picked up. If a particular type of mail had not been picked up at a particular centre then no variable needed to be generated for that type of mail for the pick-up period.
By following this and other methodologies we were able to significantly compress the size of the problem to one which could be solved and modelled.
Although the financial modelling done by NZ Post was commercially sensitive we estimate the company is saving millions of dollars from a relatively low investment. Following modelling the company made a number of changes including the closure of some sorting centres, investment in sorting machines and the introduction of postal codes for all addresses.