Media Servers Should Skip the Middle Man

Opinion: The challenge for mobile carriers is anticipating customers' needs, and media servers from companies like NMS may be the answer.

The similarity between mobile phones and advertising, not often noted, is fascinating.

I was chatting with an old Saatchi vice president the other day (now retired) and he put it like this: "Advertising is an odd business, where the experts—the agencies—often prevent disruption. If you want to introduce new advertising ideas, you have to go directly to the customers of those agencies, and get them to innovate."

Its much the same, these days, in the phone business.

If your customers dont know what they need, you pretty much have to sell them what they want.

And that may just be the time to smuggle a Trojan Horse into the mix by going to your customers customers, and getting them to ask for it.

/zimages/4/28571.gifApple unveils its iTunes phone. Click here to read more.

I dont normally get too excited about developer conferences, but the schedule for NMS Communications for this fall looks suspiciously like a breakthrough coming in mobile phone networks, particularly in Europe.

/zimages/4/28571.gifRead details here about how T-Mobile provided free Wi-Fi services after Hurricane Katrina.

NMS launched its "Vision" range of servers earlier this year. At the Munich developer convention next week, Im expecting some kind of news on where Vision is going.

But Im not expecting NMS to focus too much on where it hopes it may go. The vision of Vision goes beyond what its customers will be asking for.

The basic problem facing mobile carriers now is that they have an infrastructure which assumed that phone networks had to be proprietary technology.

Its starting to dawn on them that there is no escape from the Internet.

The old Irish story warns us: "You cant get there from here." Mobile carriers know where they want to be … but nobody knows the route. At least, nobody in the existing phone business knows.

I rather suspect that someone like NMS does.

What the carriers still seem to imagine they want is a way of distributing content to eager customers.

There are competitive reasons for thinking this. The carriers most successful rivals are distributing content, and are doing very well with it, especially in the business of retaining customers and reducing "churn" of subscribers.

The trouble is that what carriers actually need isnt what they want. What they need is a way of distributing data, profitably. Its all very well for them to tell analysts (as they do—endlessly) how much revenue they take in from ring tones, music downloads, even picture downloads, but what matters isnt the income, its the profit.

Average revenue per user is all very well, and ARPU has become a jargon word. But observers are starting to ask, "Whats your AMPU?" or average margin per user.

And, as Ive commented before, I dont know anybody who uses black ink for those numbers.

/zimages/4/28571.gifTo read about mobile phone use on airplanes, click here.

The average loss per download is appalling, ranging from the equivalent of 25 cents right up to a couple of dollars per download per user.

What NMS is trying to do is design hardware which meets the stated requirements of its customers, but which also anticipates the real needs they will experience two years down the line.

David Asher, who will be one of the NMS executives in Munich—and afterwards at similar developer conventions in Boston and Bangkok—will be talking IMS, the phone worlds version of the Internet. View the NMS events calendar here.

But in reality, Asher, and NMS generally, arent sold on the definition of IMS as being the one the industry thinks is coming.

"Were seeing pollution of the standards," was one of his comments when he stopped by in London recently.

"Were seeing people saying Weve got SIP or Weve got IMS. But when we look at the application they have in mind, its a stovepipe application, using their own definition, and not expected to work with rival systems."

Its no secret that NMS makes media servers—which it sells to all the big telco suppliers, from Alcatel to Nortel, from Avaya to Fujitsu—and that these servers are then configured using software, making them suited for a specific purpose.

In one sense, I can tell you exactly what NMS is doing. Its as radical a move as the original abstraction layers PC makers started building into disk subsystems and device drivers a decade or so ago, with a standard family of hardware that can be clustered, but onto which either proprietary, or standard services, can be built.

And the surprise will come when people start smelling the burning toast two or maybe fewer years from now, and start saying, "We have to replace all our proprietary stuff with industry standard equipment that does straight Internet Protocol bearer services!"

At that point, NMS will be able to say, "Thats what you have."

Its a high-risk strategy, because really, NMS isnt selling to the customer, but to the channel.

All the big switch makers are providing services which meet their own ideas of what the market wants.

What NMS really needs to do, of course, is to go to the operators directly, and say, "This is where you need to go!"

Until it finds a way of doing that, its operating with its hands tied behind its back.

Anybody who has struggled to get ad agencies to take the Internet seriously over the last ten years will recognize just how hard this is going to be.

But my prediction is that NMS will start talking to the mobile phone companies. Its the only way of saving them from themselves.

Contributing columnist Guy Kewney has been irritating the complacent in high tech since 1974. Previously with PC Mag UK and ZDNet UK, Guy helped found InfoWorld, Personal Computer World, MicroScope, PC Dealer, AFAICS Research and NewsWireless. And he only commits one blog—forgiveable, surely? He can be reached at

/zimages/4/28571.gifCheck out eWEEK.coms for the latest news, reviews and analysis on mobile and wireless computing.