News

More "optimisation" at 3GSM - this time by Microsoft. It could mean new toys.

During the mobile show, 3GSM in Cannes last week, the exhaustion of the busy exhibition wasn't in any way relieved by the dreariness of those professing to offer phone networks some sort of optimisation of their over-costly computer systems. But Microsoft's version was a bit different from many.

What it was offering was CSF: Connected Services Framework. It's a terribly obvious idea: all services offered by a network should be connectable. So (as Andrew Steven, director of solution sales, EMEA, for Microsoft) put it:

"Take a ring tone download service; a nice simple operator scenario. Suppose your operator says: 'I want to drive additional revenue; I want to offer that user the chance to purchase the full music track of the ringtone.' You can download that 2-meg file pretty quickly over 3G; they'll charge $3 for that. And if you like that, they'll maybe offer you the album for $12, and send that via DSL. By the time you get home, it will be waiting for you'."

Which is obvious, but unfortunately, it's not the way the system works today. Instead, the software that runs the ring-tone sales is provided by supplier A (possibly, a ring-tone agency) the single track by someone else (like iTunes, perhaps) and the album by Amazon.

What do you reckon the chance is that all three providers use the same computer, the same software, and the same billing system?

It would be nice if they did. It would mean, for example, that network operators could try out products, even if it wasn't certain that they'd sell. The process of adding a new product - say, a rock video clip - would be quick and easy. That being the case, the sales department might say: "Give it a go - it won't cost us much, won't take long, and it might work!" where, today, they'd say: "Don't think we can justify the development expense. By the time we get it working properly, people may not like rock videos... and look at the up-front cost!"

But is it possible to have "the same computer, the same software, and the same billing system?"

Nearly. Microsoft has been in this particular business for a while. It originally launched its software package, Biztalk, to reconcile different people's "enterprise resource planning" (ERP) software - but it's been catching on inside large corporations to reconcile their own systems.

As with most modern software systems, the key is an "abstraction layer" - a sort of imaginary ERP computer which you talk to in its own language. If everybody uses the same system, then it can handle all the things that are involved. That's the basic functions of finding products - if you define the products in the ERP terms, they'll all fit in the same catalogue - or charging for them - you define the prices in fields that the Biztalk system offers - and so on.

In other words, you may all have different computers, and most of your software may be different - but when it comes to products sold over the network, you all have the same system - Biztalk - running on top of your other hardware.

Terribly simple in concept; not terribly simple to do. But the key is that it works, and this has been enough to make Biztalk one of Microsoft's bigger software successes in the last couple of years.

That has led to the announcement of the Biztalk-based CSF at this show; complete with three genuine customers.

It may take a while to permeate the industry, but if it succeeds, we (customers of the mobile networks) will benefit a bit, and the networks will benefit a lot.

However, Microsoft's rivals aren't likely to see it quite like that. Microsoft code doesn't run on an awful lot of systems, and people who run those are excluded from the party. Can they stop it?

I'm no expert on this sort of software, and the people I know who are, tend to have loyalties to one system or another. What I do know is that Biztalk is XML based, and that most people are adopting this as a data-transfer and system communications model; and therefore, even if Biztalk can only function on the Microsoft .Net framework at the moment, there's probably a reasonable chance that someone will develop something that will bridge the gap.

For a different flavour of comment on this, you might like to see  this column, too. It will be archived on this site here