The MS plan for smartphones: Get Nokia!

The real weather was cold in Cannes last week, but Microsoft couldn't have asked for a better climate in which to roll out its smartphone/PDA roadmap to the assembled scribes of GSM World Congress 2002. Arch-rival Symbian CEO Colly Myers had fallen on his sword/hung it up amicably after a valiant and successful battle* on the eve of the show, resulting in the "postponement" of the new version announcement press conference and a severe exec-deficit.

Microsoft on the other hand had a plausible story, is never short of execs, spieling, for the use of, and had Nokia firmly in its cross-hairs. The Register, having on day one perhaps ill-advisedly described one of the first visible products of Microsoft's new phone reference design strategy as a small brick, found Microsoft Mobile Devices VP Ben Waldman become unexpectedly available for our 9am slot on day two. In the good old days you could be hundreds of miles away before it hit, but things happen so fast on the web that you can hear marketing discussing whether the story you've just written is more down on Microsoft or Nokia while you're choking on your pre-interview coffee.

"Angmar will see you now, Mr Lettice."

But as it turns out, Waldman's a pretty good sport who's prepared to tangle over things away from the overheads and the sound-bites; he has a pitch to make, obviously, but there's enough to that pitch to convince you that he has a good shot at success.

The sound-bites first though, as delivered at the launch event the previous day. By producing mobile phone reference designs manufacturers can run with, Microsoft and its friends are bringing "decades of data experience" to the business. In mobile phone design previously, "software was almost an afterthought... we see software as becoming a key enabler." Microsoft will bring the rich experience of the PC to the mobile phone. You shouldn't have to be a "second class citizen" on the Internet just because you're a mobile user.

There's more, but you get the picture - the reference designs will be the key enablers that establish a platform anybody can build, Microsoft software expertise will build wondrous, rich experiences atop that platform, and GPRS/3G will mean you get the whole Internet from them, not just an edited experience. As a reality check we asked ARM VP Steve Evans how much longer he reckoned bandwidth constraints would apply, he said five years at least. And GPRS barely worked at all in Cannes during show hours - funny that.

One of the first things I put to Waldman, who was ably assisted by product manager Ed Suwanjindar when he could get a word in, was that the plan was to use a kind of retread of the PC OEM relationship by way of the Xbox to create a coalition of little guys (PC companies counting as 'little' when it comes to mobile phones), using a common platform to outflank the existing phone players. It seemed to me that Microsoft had now tacitly accepted that it'd never sign up the Motorolas, Sony-Ericssons, Siemens and Nokias of this world. He demurred, merely accepting that Nokia would probably never sign up, but that the others were still possibilities. More on that anon.

Nor did he accept that the Xbox model and the reference design one were essentially different aspects of a corporate blueprint that had been deemed a good thing. Xbox is by its nature a fixed design specified by Microsoft and built under contract, whereas on the one hand the MS-TI and MS-Intel reference designs are so specific that Microsoft can say to people "here's where you can buy the bits," while on the other they can use "as much or as little of the product as they want." There are things like buttons they can't remove, because the software wouldn't work if they did, but according to Waldman it's pretty flexible. According to me that's having your cake and eating it, but waht do I know?

The differences between the TI and Intel co-operations, incidentally, are that the TI reference is largely aimed at smartphones while the Intel is smartphones and PDAs, and that the TI design was ready to roll by its announcement last week, whereas the Intel will be there later this year. There's obviously been a substantial period of co-operation between MS and TI prior to the unveiling, while Waldman isn't specifically telling how long or how short the Intel deal's been running.

How about the accusation that the small band of partners Microsoft has on board so far aren't experienced in building phones? He points out that Compal, the Taiwanese manufacturer Microsoft announced at the show, builds for the big names and produced around five million last year. Less than Nokia, granted, but good business and experience nonetheless. Sendo, the outfit set up by a clutch of Philips defectors, has mainland China production. It also has a deeply screwy idea of what you use approximately $75,000 worth of yacht rental (an informed estimate) for. A huge party? Apparently not, but it was not quite as quiet as IBM's $75,000 worth (which if I know my IBM cost more in the region of $150,000), and anyway I didn't know all this when I was talking to Waldman.

I argue with him about the 'Microsoft will never build hardware' one. I can go into a store, buy something with Compaq written on it, it's probably been built in Taiwan under contract. So Compaq doesn't build hardware either, right? He declines the philosophical challenge and says the word "Microsoft" won't appear on hardware. What's this then, I say, pointing to the reference design phone on the table, which says Microsoft on it. It's got to have something on it to show the manufacturers where their badge would go, right?

OK, draw. But we're getting into interesting territory here, because there's a particular reason in this particular field why not making hardware has particular significance. That's right, the N-word.

Nokia does software, and Nokia builds hardware, the best-selling hardware. The Symbian partners of course are all shareholders in Symbian, and Symbian designs are available equally to them all, but the way Waldman tells it, Nokia is calling the shots. And maybe there's just enough truth in that for it to stick; Nokia is certainly far more successful than the other partners, who are more than likely to resent how effectively the company has eaten their lunches. I argue that it's not Nokia's fault that Motorola cancelled Odin, nor Nokia's fault that Ericsson built worse phones than Nokia did. But you can see why some people might go for betrayal rather than their own failings as an explanation. The 1918 analogy does not, fortunately, occur to me at the time.

Ed gets a word in. "They [i.e., Nokia's partners in Symbian] are the ones who've approached us. Lately people tend to call more often with more interest." But no names, no promises of a big coup announcement later this year. Waldman (who is clearly doing this deliberately - he wants it to be known he's gunning for Nokia) says Nokia has more or less taken over Symbian, and you can see why "Motorola, Sony-Ericsson and Matsushita would be upset." Nokia has "strung them along for years, then stabbed them in the back." They must be asking themselves, "at the end of the day, does Nokia have my best interests at heart?"

It is wondrously, outrageously, partial stuff, but it could play. These are indeed deadly rivals in there, and attempting to drive a wedge in could pay dividends, particularly as two of them (Motorola and Ericsson) have been doing a deal of strategy revising over the past year or two, and might ultimately be persuaded it was all a dumb idea anyway. And tagging Nokia as the problem is far smarter than going for Symbian as a whole, because that might unite them.

We cover design, where he argues that one chip with two different processors is a good idea perplexingly calling it integration, and I am unconvinced (this is something Symbian has historically poked fun at). We cover pricing, where he cunningly dances clear of my attempt to weasel the licence fee tab out of him (mail me, somebody, preferrably with the agreement attached), and points to a "certain set of companies who charge exhorbitant amounts for services." I can't get him on the cost Win2k Server, that's not his department, and he won't accept my contention that screwing the early adopters for a huge wedge then dropping the price to confetti 12 months on is just the way of the world, and something Nokia is pretty damn good at.

But his trophy of the show appears to be the certain company that charges $6.5 million for 10,000 users, which is Nokia, and which is for MMS Server. Because Nokia is a software company too, and maybe that's another reason why it's in the crosshairs. But as I say he's good fun, and prepared to argue his corner. And although I disagree with him a lot, I'd buy him dinner for a good row any time. Don't you just hate it when you like them? ®