The commercial Internet brought with it a cascading series of business models, each different from the one before and each claiming to the "Next Big Thing." First there was the broadcast model, until it became clear that nobody could actually find anything on the Internet so there was no real audience to broadcast to. Then there was the search engine/portal model until it became clear that banner ads couldn't pay for themselves. Then there was the business-to-business model, which was supposed to keep us going until the broadcast and portal models matured, but business-to-business wasn't big enough to carry the rest of the Internet on its back.

Then there was the peer-to-peer model, which was supposed to cut back-end and bandwidth costs for providers but the record companies and movie studios wouldn't allow it to succeed. And somewhere in there was the Application Service Provider (ASP) model, which was intended to put an Internet face on clunky mainframe systems, turning us all into airline reservation agents and Wal-Mart inventory clerks. The jury is still out on that one, which is the subject of this column.

What set the ASP model apart from the others is that real work is being done. That "S" stands for service and an ASP only qualifies if the product is something you can pay for and receive. The idea is that there is some service like accounting or sales support or managing a big mailing list that you could do with paper and pencil but the ASP does it for you and all you need is a PC on the Internet and a little money. It is a simple concept, but like the other business models, ASPs are a hard sell.

What makes ASPs a hard sell isn't that they aren't needed or can't be financially self-sustaining. ASPs fight a much different battle on several fronts. First, ASPs have to fight the PC, itself. Whatever the task — for example, I use the PayCycle ASP to help pay my cleaning lady — there is always a PC application to do kinda sorta the same thing. Maybe that application can calculate employment taxes, but can it pay them? Paycycle can. The bane of the consumer ASP's existence is the standalone PC app, and a shareware PC app is even worse.

Among ASPs aimed at small to mid-sized businesses, the equivalent of that deceptive PC application is an even more deceptive character — the owner's nephew. "My nephew knows all about computers! Heck, he can knock out that (inventory system, point-of-sale system, antigravity system, whatever) in no time at all!"

Except he can't do it. If you are the owner of a small to mid-sized business and you have a nerdy nephew, understand right here that your nephew has precisely a TWO PERCENT chance of delivering a usable application that will survive twelve months on the job. That's 50-to-1 against. Yet thousands of business owners every year bet those odds. And they lose.

Among ASPs in the retail sales industry, the bogeyman that needs fighting over and over is Wal-Mart, the world's largest retailer. "I want a system, just like they have at Wal-Mart," says the crafty seller of melons and caftans. Alas, Wal-Mart isn't really a retailer at all as much as it is an IT organization. Wal-Mart spends billions of dollars per year to computerize every aspect of its business that can be computerized. With $100+ billion in sales, Wal-Mart has no choice, but the good shopkeeper who always kept his inventory pretty much in his head does have a choice. And more often than not he'd be wiser to stick with his head and forget the computer.

One thing the world of high tech really loves is to trash years of hard work by making a simple technology change. Much of what we do with PC's in small businesses today was done a decade ago in larger businesses using minicomputers. And it was often done better. But who would even consider buying a minicomputer today? This is the age of the PC, right? And the Beowulf cluster. That can be too bad for the minicomputer guys who spent decades tuning their algorithms only to be put out of work by a spreadsheet.

My candidate for ASP poster child is StoreReport.com, from Ridgeland, Mississippi. StoreReport is an ASP for owners and operators of convenience stores like the one down on the corner, though at least a few shops running StoreReport are in Siberia. Yes, Siberia. StoreReport is run by Bill Scott, who saw the PC revolution and then the Internet coming and as a result used IBM Websphere software to turn into web applications all 1200 AS/400 minicomputer applications he had written for the convenience store industry. I really want to write here that StoreReport does everything but refill the Slurpee machine (or perhaps the vodka machine in those Russian stores), but what it actually does is keep track of all the numbers — how much of what was sold to whom and have we been paid yet? StoreReport does this all for around $200 per month per store while reducing labor costs by around 60 percent. Heck of a deal.

"Our customers run convenience store operations," says Bill, "hundreds of remote sites, where workers may not even know what the owner looks like, and see their supervisors maybe once a month. Is it any wonder that it takes 3-10 days for an owner to find out what's going on at one of his stores? The biggest operations collect daily transactions and then send them to the office via modems. There, they are viewed and sent to the main computer for processing. If there is something wrong with them, it may be days before someone can communicate back to the store to correct the mess.

"I have learned a valuable lesson over the past two decades. An office worker will let you do anything for them that you are willing to do, and they will continue to let you do it until they have transferred their entire job over to you. Given sufficient time, I can gradually take over a company's administrative functions until all they have to do is greet customers who frequent their stores. And why shouldn't they let me do it? I can train one person to do the same job for 6-8 different companies. Payroll is just one example of the mundane tasks that administrative workers perform. Giving up control to a trusted business partner makes a lot more sense than turning your business over to a staff of employees who may not have your best interest at heart. When you consider that the greatest security issue is 'sabotage by employees', why not eliminate it entirely by eliminating employees? Computers replace people."

Against this rational if minicomputer-centric thinking, we have the alternative of spending $30,000 per store putting in Windows 2000 systems that talk .NET, spewing XML statements back and forth. In this retail context, XML replaces an earlier standard called Electronic Data Interchange (EDI). EDI failed because it was too big and slow, requiring too much bandwidth. XML requires four times the bandwidth of EDI. What do you think is going to happen?