The Novatel MiFi - possibilities for new mobile broadband business models

OK, I realise that I've been a bit grumpy and critical of some things recently. But before everyone assumes I'm getting more cantankerous as I get older, I'm actually going to say something nice for once:

I really like the little Novatel MiFi unit I've been playing with the last couple of days. (Peter Judge does too). For the uninitiated, it's a cross between a 3G dongle modem, a WiFi access point and a Linux-powered device.

Basically it's a portable, battery-powered personal "hotspot" that up to 5 people/devices can log onto, multiplexing their data traffic through a single SIM and data connection. Various larger, chunkier 3G routers have been around for quite a while, mostly for residential use, but this one is small, cute, and doesn't need to be plugged into a power socket. You could also compare it to using a 3G handset as a modem/tether.

At one level, it's nice just as a standalone product, especially if you have lots of devices, or work in a small team. It means you don't have to install connection-manager software onto your computer - or worry about conflicts with an existing clients. It just uses the normal WiFi connection, and appears in the list of other WiFi AP's your PC can hook up to. It's also good to have your mobile broadband being powered by a separate battery, rather than mentally juggling your own power budget between screen and modem.

Part of the cleverness, though, is the fact that it's programmable - if sold via an operator (or another channel), it can have various additional capabilities. It doesn't have an open SDK or set of public APIs, but in theory it could be usable for all sorts of cool things - coupled with its memory-card slot for example, an operator could choose to push video or other content down to it during quiet network periods. I'm probably a long way ahead of the curve here, but I guess in the long term something like this could probably be used in conjunction with a content-delivery network or web cacheing architecture.

More prosaically, I can operators trying to charge extra for a MiFi-enabled mobile broadband plan than for a simpler dongle-based one.

On the downside, if I'd bought it via retail I'd have to find a good data-only SIM and tariff. Not too tricky in the UK, but harder in places like the US. You'd have to go via the (quite simple) admin screen on the device to set up the APN and other config settings. It's not obvious to me how it handles SMS-based alerts from the network ("You're nearing your monthly bandwidth cap") although that's a minor niggle.

The one problem I can foresee though is around network data capacity - in areas with congestion, do operators really want to risk too many five-in-one connections? Will it further increase data consumption, without adding revenues in similar proportion.

I'd quite like to see a dual- or triple-SIM version, allowing the user to pick & choose which network to use. The OS could even perform a least-cost routing function, as it has GPS in it. "Hmm, I'm in Spain, I'll use the Yoigo SIM instead of roaming on Vodafone".

One other thing: it looks really good. With a couple of LEDs blinking on the outside, I've had a few people come and say it looks interesting.....