DO SOMETHING RADICAL: kill off your purchase order system (POS)! Many FDs I know and respect are wedded to a centralised POS. But they’re fundamentally flawed.

At an off-site meeting I had recently, I noticed the FD was trawling through his emails during the break, quietly cursing the latest inter department battle over POs. the board had just approved a Ł30m marketing burst. It was all properly minuted, given the extraordinary expenditure. So did that decision really need to be followed up by a cascade of Pos? Was the FD’s time well spent reconciling them to the original plan? If you are a proud owner of a POS, it probably works something like this. The company needs to buy something, so junior goes to boss who’ll approve the purchase in documentary form. The purchase order will be sent off, the goods or service will arrive with, you hope, some sort of “goods received notification” (GRN). Then an invoice will arrive. You can see why they’re popular. The spend has been approved in advance by someone with sufficient authority; and a PO can be reconciled with the GRN and the invoice. Once the purchase ledger clerk has matched up all three documents, the supplier can be paid. You only purchase goods and services you require and you’ll only pay the agreed price for them. So where does it go wrong?

Well, the last time I looked, the UK was a service-based economy and the provision of services doesn’t work like that. Let’s take, for example, the supply of marketing and advertising services. You may agree a price of Ł100,000 for marketing material as part of an advertising campaign. But when the invoice comes in from the agency, you’ve actually been billed for Ł103,000. Why? Well there could be a dozen reasons – reworking, cost of runons if you’ve upped the original order, peripheral costs such as couriers and other incidentals – which are perfectly legitimate and perfectly impossible to predict. So what does the purchase ledger clerk do then? He or she sends the invoice back to the marketing depart-ment for post transaction approval. Messy. Think you can fix the expected-but-unknown items by sticking contingency on top? Not a great idea. Your supplier sees the value of the total budget on the PO and ensures the invoice hits the amount to the penny. Who can blame them? Even if the invoice and the PO match, the receiver of the service is unlikely to have entered a GRN – so the accounts staff have to send the invoice back to the department where it originated to get confirmation that they did get the services they’re being invoiced for.

My solution? Devolution. I know one company that had a POS in chaos. The back-log was driving the staff in the accounts department and elsewhere in the company mad. You can imagine what it was doing for the suppliers who saw it as a conspiracy not to pay. After establishing that no one in the company used the central POS, it was abolished in favour of department heads approving invoices (in this case using document scanning and workflow tools). And if you think that the weakness in that system is that executives will abuse their signing-off authority – well, in my experience they don’t. Most people in business take their duties seriously. We hear a lot about trusting and empowering people – and abolishing central POs does that. (And if you think you need to keep POs because they stop fraud – well, maybe. But it doesn’t take a genius to circumvent any of the POSs the company used the central POS, it was abolished in favour of department heads approving invoices (in this case using document scanning and workflow tools). And if you think that the weakness in that system is that executives will abuse their signing-off authority – well, in my experience they don’t. Most people in business take their duties seriously. We hear a lot about trusting and empowering people – and abolishing central POs does that. (And if you think you need to keep POs because they stop fraud – well, maybe. But it doesn’t take a genius to circumvent any of the POSs I’ve seen. In fact the accounts staff in this company thought the POS was there for the benefit of the purchasing departments; and the purchasing departments thought it was an essential task for the smooth running of finance. Either way, no one was taking it seriously which means they’d created loopholes any decent fraudster ought to have been able to exploit.)

Those who approve invoices have budgets – and they know how much the bills are likely to be because they ask for quotes which are kept and cross-referenced to the invoice within the department. Any theoretical loosening of cost control is more than compensated for by ending the eternal paper chase, freeing up staff in both the purchasing department and finance to do productive work. OK, so I can hear you asking one more question: what about accruals? When you’re drawing up the month end, what better way is there to ascertain the level of total creditors than adding up the goods and services received but not invoiced? Hmm. It’s great in theory, but it does require brow beating business managers into providing GRNs at month end – it’s the receiving of the goods or services that creates the accrual. And recording GRNs seems to be the one task that’s never done, even in the companies that enforce POs with vigour. I accept that there are manufacturers and retails out there for who a stock PO system works like a dream. But for the rest of us? POs should be consigned to history.