A Web front-end is nothing but "lipstick on the face of a pig". The pig is the back-end - the order fulfilment system upon which your business stands or falls, IT directors at the recent Computer Weekly 500 club meeting were warned.
Apart from companies whose product is pure information, everyone else deals in the tangible world of material goods - and these cannot be downloaded via telephone wires.
"E-business puts the spotlight on order execution," warned supply chain management consultant, Simon Bragg of ARC Advisory Group. Too much spotlight can, he pointed out, lead to "flawed customer service."
"A vehicle can cross the UK in eight hours, so why should deliveries take a week?" he asked.
Flawless customer service requires an equal spotlight on the back-end - the fulfilment. "E-fulfilment gets complicated, there's a lot of components," warned Bragg.
These include the whole gamut of activities from tracking and tracing to carrier selection, manifesting, warehousing and getting proof of delivery - a lot of steps involving a lot of people, just to move A to B.
"Perfect order fulfilment means that it's on time, undamaged, in the right quantities, and is billed correctly and with no unauthorised substitutions, and with the right documentation included," said Bragg.
Easy to define, but tough to do. Yet e-business will eventually stand or fall on how well it can be done.
Bragg outlined three ways to do e-fulfilment, of which you will probably use a combination. You can build your own, outsource, or buy fulfilment in a spot market.
To do it yourself you will need quite a hefty outlay, especially on software, with items such as online order management and online configuration systems on your shopping list. Prices easily run into the hundreds of thousands of pounds.
"E-fulfilment systems are immature and expensive," warned Bragg. "There is a very fragmented market and no-one has the complete solution - for example, some companies have a good front-end and bit of transportation management but no warehouse management. There's no easy solution."
Most will have to integrate best-of-breed components.
There is also the issue that e-fulfilment is not the same thing as ordinary fulfilment. With the goal of e-business being able to sell profitably to a market of one-offs, the emphasis is not on bulk orders but a high proportion of small orders, said Bragg.
This means that traditional processes in the fulfilment chain will have to change - such as wave picking in batches in warehouses, where people are used to handling, picking and storing pallets, rather than one-offs. Because of the nature of e-business you will need very high inventory accuracy.
Policies on substitutions may also have to change in an e-business environment. A customer buying groceries over the Internet may not be happy if they get a different brand of baked beans in the delivery box, even if, were they in the supermarket themselves, they would quite happily buy a different brand if the preferred one was out of stock.
It is telling, Bragg points out, that two of the supermarket chains with a high profile in home-shopping on the Web are taking different paths to fulfilment - one is building dedicated e-warehouses and the other is picking from the supermarket shelves. It will be interesting to see, he says, which model proves more successful.
One thing is for sure, says Bragg, e-fulfilment is more complex, and therefore you will need higher managerial talent in the warehouse.
But if e-fulfilment is complicated and untried, why not solve the problem by outsourcing it?
In the USA some companies are already offering to design your Web site, handle all the transactions and payments and do your marketing, warehousing and transportation using established nationwide delivery companies. This won't come free - expect to pay 35% of the value of the goods to the outsourcer. This compares with a standard 45% if sold through retailers.
The third option is the growing emergence of Internet exchanges. There are already about l5 public exchanges for transport in the US, but few are doing any real volume of business, and some have already failed.
At moment, the main emphasis of the exchanges is on price negotiation, using comparison engines to seek out best prices for buyers from the suppliers on the exchange. However, their main value may prove to be different, said Bragg. This is because, as so many of the Computer Weekly 500 Club IT directors pointed out, when it comes to buying transport, logistics is already pretty lean and mean. Most companies have gone to considerable trouble to set up very tight, very competitive contracts with the result that there is little fat remaining to sweat out of that particular pig.
Nevertheless, one possible fat roll is the issue of the backhaul - the empty truck's return journey. On a national - let alone a global - scale, the sheer amount of empty capacity in freight transport is a potential inefficiency. Could logistics exchanges sweat that inefficiency out by providing a means of identifying spare capacity and mopping it up? Demand aggregation would mean a genuine chance to take cost out of fulfilment.
The other area in which logistics exchanges might contribute value is in the services area, absorbing functions such as insurance and currency risk management or intermediate warehousing.
With all the variations available - and every one with limitations and disadvantages - the bottom line of e-fulfilment is that, as yet, there is no bottom line. Industry is still searching for one, and models are still being tried out, with quite a few more still awaiting their Eureka cue. As ever, the big question will be: is it better to get first-mover advantage, or abide by the adage that you can always spot the pioneers by the arrows in their backs?