After an IT outage, warehouse staff need to establish which orders are ready to leave. Some orders may have been entered by hand during the interruption. Staff must check them, find the right stock and confirm transport. If the day’s last collection has already left, they may need another lorry or a later delivery date.
For a business selling food and other everyday goods, a late delivery can have different consequences from one product to the next. Some customers can accept a later arrival. Other goods may no longer be saleable, or staff may lack the records needed to show which production batch they came from. These differences should guide spending on recovery.
Find out what stops an order leaving
Follow an order from the customer’s request to the lorry leaving the site. Ask staff what they need to accept the order, reserve stock, pick the goods and book transport. Include their ability to sign in, scan items and receive information from customers and suppliers. Test that whole sequence. A working warehouse application is little help if customer orders still cannot reach it.
Operations managers should identify when a delay becomes harder to recover. How late can staff finish loading before a driver must leave? Can the next production run wait? How long can a product remain outside its required storage conditions? Separate orders that can move to tomorrow from promises that tomorrow’s delivery cannot fulfil. Check the customer agreement before assuming that a missed delivery incurs a charge.
Measure recovery by the work staff can complete
Set a recovery target that states what people must be able to do, not just when an application must restart. For example, staff should be able to receive an order, check its stock and release the correct goods for delivery. Also decide how much recent information the business can afford to lose. If the latest usable backup is several hours old, someone will need to recover the orders received during those hours before work can proceed safely.
Measure how long it takes to send the first correct order, then how long it takes to handle a normal shift’s orders. The waiting orders need their own plan. If all available staff and equipment are busy with new orders, the queue will not shrink. Managers may need extra shifts, later delivery dates or a decision about which customers to serve first.
Check what people can do without the system
A paper procedure is useful only if the shift can carry it out accurately and fast enough. Try it with the staff and equipment that would actually be available. Check that staff record which batch each order contains and do not send the wrong goods. If they can handle only part of the usual orders, agree which customers or products come first and who can approve a different choice.
For perishable goods, quality and safety staff must help decide what can leave the site. A warehouse IT failure does not necessarily stop refrigeration. Check which temperature readings remain available, whether staff can trust them and what other evidence they need before releasing affected stock. Do not assume either that all stock is safe or that all of it must be discarded.
Count the losses without counting the same cost twice
Ask finance to separate the cost of repairing systems from extra wages, replacement transport, charges owed under customer agreements and stock that must be discarded. For sales lost permanently, estimate the money the sale would have left after the costs of supplying it. A late order that the customer still accepts is not automatically a lost sale. Check that the stock cost is not included both in discarded goods and in the lost sale calculation. Keep possible future customer losses separate where the evidence is uncertain.
Work out costs for different points in the shift rather than assuming every hour without IT costs the same. In the morning example, completing the order checks before the lorry leaves may avoid replacement transport. Finishing afterwards may mean another booking and a late delivery. Use the exercise results and actual customer agreements to explain the difference.
Buy an arrangement that gets deliveries moving
Compare a second order connection, another warehouse or additional backups using the same delivery problem. For each alternative, check whether it and the current arrangement depend on the same network, administrator accounts or supplier. A single failure could disable both sides of the pair. For ransomware, which can lock or damage business data, the plan must include separating affected systems and checking that restored information is safe to use. Automatically copying data to a second system can copy the damage too.
Include setup, maintenance and practice costs in the price. Exercise the switch to the alternative and the return to normal working, including checks for orders entered twice. Faster repair of one system is worth less if staff still cannot complete another step needed to send the goods.
Before approving spending, ask the operations manager to show which deliveries the proposal would save and how many orders staff could handle. Use a realistic shift exercise as the evidence. A promise to restart a server within a few hours does not answer that question.