Supply-chain planning is software that forecasts demand weeks or months ahead and turns that forecast into a plan: what to make, which parts to buy, and how much stock to hold in each warehouse. The supply chain is the whole path that goods take, from the parts and the factory to the warehouse, the truck and the shop. Supply-chain planning (SCP) software is among the largest markets in business software, and much of it is built by engineers in India.
The software does three jobs, in order. First, it makes a demand forecast: how many of each product will sell, in each region, in each week. Second, it makes a supply plan: how many units each factory must make, and when the parts must be ordered so that they arrive in time. Third, it makes a stock plan: how many units to keep in each warehouse, so that shops can be refilled quickly.
Take, for example, a TV maker with a factory near Noida. In June, the software forecasts that the company will sell 3 lakh TVs in October, the month of Diwali, against 1.2 lakh in an ordinary month. The TV panels are imported, and they take ten weeks to arrive. So the software tells the purchase team to order the panels in early July. The factory cannot make 3 lakh TVs in one month. So the software plans for it to make 1.7 lakh TVs every month from July, and to build up the extra stock before October. It then plans how many TVs to send to warehouses in Delhi, Mumbai, Kolkata, Chennai, Bengaluru and Ahmedabad, in proportion to each region's forecast.
The plan does not stay in the planning software. The purchase orders and production orders are created in the company's enterprise resource planning (ERP) system, the system in which a company keeps its orders, stock and accounts. The stock targets go to each warehouse management system (WMS). The truck movements go to the transportation management system (TMS).
Demand never arrives exactly as forecast. Say a rival cuts its prices in September, and sales in the south fall behind the plan while sales in the west run ahead. The planning software then moves part of the stock from Chennai to Mumbai. Three of its parts handle changes like this. Concurrent planning updates every plan when one of them changes. Demand forecasting and replenishment refills each place before it runs out. AI planning agents carry out many of the changes themselves.
How is it different from merchandise planning? Merchandise planning decides which products a retailer will sell in each store, and how many. Supply-chain planning decides how much to make, buy and move so that those products are there in time.