As with most innovations in business, the emergence of e-procurement has generated a flood of hype. The usual suspects - software suppliers, management consultants and pundits have queued up to extol its virtues.

The list of benefits is certainly mouthwatering: lower transaction costs, greater price transparency and better buying information - a genuine step, it seems, towards the fabled "perfect market" of economics textbooks.

But how does the reality compare to the hype? Does e-procurement live up to its promise? Or will it, as the deluge of waffle recedes, prove disappointing?

Suppliers are having a particularly tough time. In the rush to trade online, many are finding that the costs risk outweighing the benefits. The tight margins of commodity and indirect materials are also a hindrance. "With a 2 per cent margin, it can be difficult to justify the expense," says Joe Baah, European sales director of indirect marketplace MRO.com.

The problem is that there is no simple way to trade online. Scale is the issue: the cheap-and-cheerful credit card e-commerce solutions that are appropriate for small retail outlets, or niche manufacturers, won't work for large companies with thousands of products and price and discount structures built up over years of established trading.

Citing evidence from the US, where more suppliers have taken the plunge than in the UK, Baah claims "it can take $300,000 to set up shop, and $1 million if they want to carry out meaningful transactions".

Yet the alternative - membership of an electronic marketplace - is little better. The problem, explains Paul Rose, head of business service at e-commerce specialist Basaar, is that these marketplaces are constructed around electronic catalogues of products, and to ensure consistency and ease of use, suppliers' catalogues must be converted for online use. "Depending on the state of a cat alogue, industry experience suggests that it will cost between L2 and L6 per line item to make it ready for e-commerce," says Rose.

That, in itself, is not a problem, but often customers don't only want their suppliers to trade electronically - they also want them to trade in the marketplace of the customer Is choosing. This has at least two significant benefits for the customer: a slice of the transaction fee - if, as is increasingly the case, the customer has a financial stake in the marketplace - and a greater leverage over suppliers by generating more competition.

Suppliers then find themselves having to maintain several catalogues, all in slightly different formats and with slightly different ways of connecting to the core marketplace technology. This takes suppliers further and further away from their core activity of manufacturing or sourcing goods and into maintaining databases and software connections.

Recognising this problem, third-party specialists are springing up to manage the process on suppliers' behalf. One of these, London-based Cataloga, sees the catalogue problem as "a significant barrier to the success of e-procurement." Until it is resolved, the company says, "the difficulties involved in suppliers having to provide multi-format catalogues will reduce the adoption, penetration and effectiveness of e-procurement".

Even so, outsourcing the problem to a third party doesn't eliminate it, although it can make the cost of living with it lower - a specialist should be able to manage catalogues more efficiently than a supplier. But the service must be paid for, and this cost either comes out of the supplier's profits or is built into the price paid by the customer.

Clarifying benefits

The benefits of e-procurement may be unclear from a supplier's point of view, but they are equally blurred from the customer's perspective. Take the claimed reduction in transaction costs. The logic is straightforward - compared with the cost of manually processing purchase orders, electronic trading is bound to be more efficient. But by how much? And where does this efficiency actually come from? Oracle, for example, claims that its implementation of e-procurement at Xerox resulted in a drop in transaction cost per order processed from L95 to Ll6. But how realistic is the L95? And how realistic is the reduction? Also, are such reductions, on their own, adequate reason for adopting e-procurement?

Numbers can prove anything, particularly when the sale of a costly e-commerce system is at stake. "There's a huge amount of 'vapourware' and 'slideware' out there," says Harry Cassar, strategic e-procurement manager at BP, which, under the impetus of its chief executive, Sir John Browne, has probably pursued e-procurement more thoroughly than any other UK-based business.

At first glance, figures as high as L95 seem ludicrous. But considering the time it takes for several people to raise, process, review, place and pay for an order, L95 seems less unreasonable. The more that automation can eliminate the people content of the process, the lower the cost will be.

But it is difficult to pin a definitive number on an activity when some costs are fixed and some are variable. If purchasers have to place or approve one less order, it makes no difference to the monthly salary they receive. And the same can be said for computer costs, communication costs and others - the closer one looks, the more fixed the make-up of a transaction cost seems to be.

Which isn't to say that the promised reduction is totally spurious - it's simply hard to quantify precisely. The perception is growing that the real benefits of e-procurement will come not from lubricating the purchasing process electronically, but from redefining it completely.

At BP, for example, it's accepted that very little of the estimated L125 million that the company expects to save through e-procurement by the end of next year will come from reduced transaction costs. Instead, says Cassar, the major benefits will come from using e-procurement to adopt such practices as enforced compliance with pre-negotiated purchasing contracts and reverse auctions, as well as from the greater price competition that electronic marketplaces can provide.

BP has now conducted about 3o reverse auctions, says Cassar, ranging from "the spectacularly successful to those that didn't generate a single dollar of savings, but still provided a useful learning exercise".

Serious reductions in transaction costs will have to wait for greater standardisation of e-procurement systems and processes, adds Cassar. "It's one reason why we've pursued the exchange route [to e-procurement]," he says, referring to the TradeRanger electronic marketplace, developed jointly by BP and 13 other oil, gas and chemical companies. "It's a clearer `win-win' situation for buyers and sellers." Particularly, he says, when suppliers have more fully integrated their systems with the exchange.

This will take some time, not only within TradeRanger, but across electronic marketplaces generally. Almost universally, digital links to core financial functions, such as general ledger and payments systems, are lacking. As a result, companies are buying and selling online but are settling invoices with devices such as purchasing cards, and this is still standard practice in many exchanges.

Under construction

But finance is not alone in this: beyond the direct transaction between buyer and seller, nearly every digital link in corporate back office systems is "under construction". Although enterprise resource planning vendors have been slow to wake up to this need, most are now tackling the issue.

This is a shame, because it is here that the greatest benefits of e-procurement will probably arise. Forget shaving a few percentage points off price levels or reducing administration costs - seamless integration of buyers' and sellers' key systems across the whole purchase spend, rather than commodity or indirect items, is more interesting.

"There's an awful lot of potential locked up in extended value chains that today's disjointed e-commerce systems can't get at," says Ian Walker, an associate director of supply chain strategy consulting with Computer Sciences Corporation, one of the world's largest consulting and IT services firms. "People have been talking about the extended enterprise for a decade, but it's only now can we begin to see it taking shape," he says. The ideal scenario is when a customer places an order with a first-tier supplier and matching requirements seamlessly flow down subsequent tiers.

It's a wonderful vision, especially when combined with genuinely lower transaction and e-catalogue costs, whenever these arrive. But how likely is this? The next article (on page 11) explores this issue.