In the next few weeks (hint: CTIA is in three weeks) I'm expecting mobile operators to be offered a new tool, which will allow them to work out what on earth their mobile broadband customers are doing.

The mystery, however, is not "what is this tool, Guy?" - all will become clear quite quickly. What is mysterious is the answer to this question: "If mobile operators don't know what their customers are doing, how on earth do they know what to charge for the broadband?"

I put this question to an acknowledged expert, Simon Bransfield-Garth of CarrierIQ, who was typically forthright. "Their pricing plans make no sense to me. I think the carriers have little, or no idea of what people do with mobile broadband."

The mobile broadband revolution, in short, could be a bubble. If the carriers could provide ADSL-standard broadband for half the price of ADSL indefinitely, then the market would simply grow and grow. But it can't, and nobody in their right mind imagines they can; at some point, the bandwidth they offer will be swallowed up and they'll have to find more from somewhere. And that will cost real money.

Analysys Mason this week offered the opinion that the last six months have seen an "explosion" of mobile broadband, and quoted quite a lot more figures than we normally get to see.

The trouble is, there's no bottom line.

Take this cheerful-sounding statement from Matt Hatton's analysis of the mobile network operator's (MNO's) future:

"The rapid growth in mobile broadband and 3G data service adoption has far-reaching implications for MNOs' business models. Operators have tended to focus almost solely on providing narrowband voice and SMS, but the composition of their network traffic is changing."

Specifically, Hatton says, the volume of data is now exceeding the volume of voice. "T-Mobile reported in April 2008 that the volume of data traffic on its network in the UK had exceeded that of voice traffic for the first time in the first quarter of 2008.

"Mobile broadband pioneers, 3 UK and Vodafone, are likely to announce a similar trend this year. 3 UK reported a seven­fold increase in the volume of data traffic on its network in the six months to March 2008."

What this tells us is that they had more 3G capacity than phone customers could use, and are now soaking up spare 3G capacity by selling bits to laptop users. What it doesn't tell us, however, is whether these are profitable bits. In other words, when will the current growth rate reach the point where they have to start investing in new network capacity? And what will that cost? And will the broadband payments cover that?

"Contention will definitely become an issue," said Bransfield-Garth. "Already, phone users will be familiar with the 'network busy' sign when they try to place a call at rush hour. What we need to know is whether it really is better for the MNOs to keep one BBC iPlayer viewer going, and to prevent 40 phone calls going through."

Behind the scenes, some operators still seem to have their heads screwed on. Vodafone, for example, is a major player in the wireless broadband business in the UK, and I've had one of their dongles for almost a year. I had an unofficial chat with a senior engineer, and asked, straight out: "How long can you survive in this market, competing with the wired suppliers, if your users really grow at current rates?"

"Oh, I'm all right, Jack," was his reply.

Actually, the answer was phrased in gobbledegook, and I had to ask another consultant (he won't let me say who) for a translation. His summary: "They have just done a deal with BT on the 21CN rollout, whereby they sign up early, and get a good deal for at least three years, and will be able to keep up with whatever users do."

Put another way, the Vodafone position is that everybody may sink into a bog of broadband contention, but where everybody else is on foot, Voda is on horseback. They'll be the last to drown. But, says Hatton, drown they jolly well will:

"MNOs need to review their requirements for network capacity immediately in response to this rapid change in traffic composition," he wrote. "They will need to invest further in their RANs as well as the ongoing upgrade to HSPA. The demand for additional network capacity – from reallocation of GSM spectrum, the 2.6GHz expansion band and the digital dividend spectrum – will be substantial."

Analysys Mason believes that the amount of additional spectrum made available will be sufficient to support the predicted demand during the next 3–5 years. But that's not all: "MNOs will also need to increase their backhaul capacity, which is currently the most significant limiting factor in terms of the amount of available bandwidth."

Surely, the mobile operators must know what they are doing, though? You wouldn't get into a market for broadband, competing with ADSL carriers, if you couldn't make money on this, right?

"To be honest, I doubt if any of them have been able to look that far ahead," said Bransfield-Garth.

The problem in September last year, when the market was blown open by Hutchison 3G and its flat rate broadband deal, was that they had masses of unused 3G capacity, and voice users weren't taking it up. BT's head of mobile broadband at the time told me that "as little as 3 per cent of the capacity for mobile data, is actually used; there's not the slightest chance we could over-sell it".

Well, was he right? Many observers think not. They say that the operators simply had to do something, anything, to increase data traffic; and any sales at any margin would do. "When they say 'we can't run out of capacity' what they really mean is: 'Over the next two years, if we grow at twice the rate we think we can, we'll still have spare capacity'. But they honestly haven't looked any further than the next six months," said my anonymous source.

The availability of a bit of network software to analyse what sort of data traffic they are generating will be really handy, of course. It would show whether this traffic is premium data, or throwaway traffic. If it's highly-critical business, time-premium based, then putting up prices will just increase margins. But if users are only using mobile because it's cheaper than their ISP can do it over BT ADSL, then putting up the price will kill the business stone dead.

The smart money says it's not mission-critical stuff. As Hatton says, "As well as increasing capacity, MNOs will need to keep a limit on costs because mobile broadband traffic will be less profitable than traditional voice and SMS traffic." And judging by the detail in his report, this isn't guesswork.

Another estimate suggests that current data rate charges work out at as little as one-twentieth the price per bit of voice traffic.

So, are current mobile broadband users watching re-runs of Doctor Who? Or is it sweaty-faced financial directors scanning the red ink on Bloomberg in real time, and logging onto Euronews streaming? If it's the latter, the future for mobile data looks hot. If it's Doctor Who and Deal Or No Deal, then people will switch to Sky and Slingboxen just as soon as the price rises above Sky subscription rates, or as soon as the data caps drop below three DVDs a month.

And if people genuinely want HD video, but only if they can get it at flat rate TalkTalk prices, then no amount of fudging and data rate capping can save the mobile broadband business.