A Cambridge perspective
on the digital world
 |
 |
July 2002
This month's View from the Fen is a personal view from Geoff Vincent, senior partner of Mediation Technology & CEO of mobile applications developer mTank.
Everyone who has not been on Mars recently (I presume this includes the mobile operators, though I have known people who had doubts about this) knows that SMS text messaging has shown phenomenal growth over the past three years. Those who study the figures will also know that text messaging represents a growing proportion of operators’ revenue and profit (up to 20% has been estimated). Also, text messaging is highly, highly profitable. A quick back-of-the-envelope calculation indicates that a five minute phone call contains roughly the same amount of raw data as about 1800 text messages. If phone calls were charged at the same rate as text messages - around 10p a go, if you're lucky, which translates to 62.5p/Kbyte - we would be paying about £180 for a five minute phone call. Nice work if you can get it. It means that text is a very nice little earner for the operators.
So given the amount of money operators are making out of SMS, and how little it is costing them to deliver, it would seem a safe bet that the operators would at least make sure that users receive a good service from text messaging. Wrong.
How do I know this? Because I have just spent three weeks (or was it three months… I’ve lost count) of 7-day weeks and 3am nights launching a World Cup SMS prediction application. For those not familiar with such things, this means an application running on a network server, which uses text messaging to communicate with users. In our case, players who had registered (by sending a simple text message to a shortcode number) received a daily gamecard on match days indicating the games to be played that day (eg England-Argentina, Sweden-Nigeria…). Players simply sent back a text message with their prediction of the results.
When the matches were played, the predictions were scored against the actual results. Players were awarded a score, and the highest ranking had a chance to win prizes (and also be insufferable to their friends for the next day because they got it right!).
The operators’ part in this was simple: deliver messages from the application to mobile users; deliver responses back from the application; and in some cases add the grand sum of 50p to the user’s phone bill for transmitting a premium message. Doesn’t sound hard; and there’s nothing here that networks in other countries (notably Scandinavia, and also Spain) haven’t been doing quite happily for some time.
What happened? Here is a selection:
Oh and by the way Network B, your clock is running about six minutes slow.
Names have been withheld - for now - to protect the guilty. However we should say that Network D - the only one that hasn't led to sleepless nights and hair tearing - has a name beginning with V. But even they will not provide a Service Level Agreement (SLA): a contract which guarantees (or even states) the level of service we can expect.
The networks, and everyone else in the industry, tell us repeatedly that applications are the key to the next generation of mobile revenues. But - please sir - tell us how we can possibly write applications, let alone sell them to the public, if we do not have an infrastructure that is reliable, and performs to quality standards that we know and understand? It's like building a hotel on sands that continually shift, and trying to offer holidays. "I'm sorry sir, the hotel lobby has entirely disappeared today, and there's no way to reach the upper floors… so you can't get to your room. No I can't give you an idea when (or whether!) the problem will be solved… try coming back next month. Oh you'll be back at work by then?" (Yes I know there are hotels like that - we can see them on TV every Wednesday night - but I don't think they stay in business for very long.)
I said we had withheld the names - for now. Since we are logging every message in and out of our system, we thought we would make use of this. We will be tracking the performance of each of the networks on a regular basis; and we thought we would publish the results. This time we will be naming names. Furthermore, we propose to make a regular award, on behalf of the mobile applications industry, to the worst performing network.
From the perspective of an application developer - and with the experience of trying to do this for real - I can say that the number of compelling applications we can develop depends crucially on the performance of the underlying infrastructure. If the infrastructure cannot reliably deliver messages within (say) a couple of hours, in all foreseeable circumstances, then the number of useful applications we can develop is a fairly resounding zero. If it can do this, well there are a handful of things we can do. (All this assumes, of course, that handsets are switched on and within range.)
If the infrastructure can reliably deliver messages within (say) five minutes, then an entirely new range of interesting and compelling applications becomes possible. If it can do it within a few seconds - say less than a minute - then the sky's the limit, and some of the things that were said about mobile applications revenue to investors and other interested parties might actually come true.
Meanwhile if the billing system does not perform accurately and reliably, and do exactly what it says on the tin, then we, and more particularly the operators, will be dealing with an extremely large number of angry and frustrated customers.
If our experience with the World Cup is anything to go by, the current infrastructure seems to struggle to meet even the first of these criteria. Also regrettably, the system is only as good as its weakest link. Like SMS messaging itself, SMS applications will only take off when all of the networks conform to a lowest common denominator standard. Until this happens, application developers, content owners and users simply won't bother.
What would it take to fix this? Mega-billions? Hardly. Compared with the billions being spent to install next generation networks, the estimated few millions it would take to upgrade the current SMS infrastructure to give consistent performance, and to bring in money now rather than in some indefinite future, would scarcely rate a decimal place on the balance sheet. Given the revenue and profits SMS is now generating, the payback period is likely to be measured in months, rather than the years or even decades that is usual for telecoms investment. All it needs is for the networks to listen to their customers, and give delivering a service that meets the basic provisions of the Sale of Goods Act - ie fitness for purpose - sufficient priority in their thinking.
If this is such an easy decision (which it is), why is it not being done already? When we ask the question, we usually hear two answers: (1) "All our attention is focused on next generation (3G)"; (2) "SMS is old technology, and won't last. We're thinking of the future!"
As one of those who was closely involved in setting up the European research project which originated what is now known as 3G, way back at the dawn of time in 1988, I am a strong supporter of next generation technology. It has enormous potential for generating new applications and new revenue, and in fact for revitalising the whole telecommunications industry - something we all desperately need, including the many thousands who are now losing jobs at telecommunications suppliers worldwide. But we will never get there unless we have a robust current infrastructure that allows the applications, the customers and the market to develop. This is exactly what happened in Japan, and it worked. i-mode (and before that PHS, if anyone remembers that) created the applications, the market and the demand, and prepared the ground for the successful launch of 3G in Japan, years ahead of anywhere else.
As for obsolescence, as application developers we can see a continuing use for SMS for years - and probably decades - ahead. It's cheap (well it should be), reliable (ditto), and even in the era of full-on 3G we can see many, many things for which text messaging will be the most effective solution. For one thing, text messaging will provide essential application-to application communication that the user never sees. It is increasingly clear that text messaging will provide the plumbing on which the entire mobile applications superstructure will be built.
MMS will build on and extend the foundation created by SMS. Many developers plan to use SMS to provide the essential billing function on which both MMS and 3G will rely. In any case, the payback on providing a quality SMS infrastructure will be obtained almost before the ink is dry on the cheque.
If we can't even build an effective and reliable SMS infrastructure, how can you persuade investors (let along users and developers) that the much more complex world of next generation will ever get off the ground?
So, to the bottom line for operators: Do you think you could get your act together and manage to deliver some messages from A to B just a little bit quicker than it would take to walk there, and without getting lost on the way? Then the rest of us might stand a chance of developing some applications that would actually give users what they want and make some money, not just for us but for you. The last time I looked, operators were still taking a rather hefty chunk out of every premium message charged to users.
Then we all might end up with applications that deliver on the potential for mobile that we all know is there: applications that deliver compelling benefits for users, new revenues for operators, and adequate returns to the people developing the applications. The mobile infrastructure COULD be significantly more reliable, and certainly more immediate, than the wired Internet. It would not take a great deal to achieve this, and to make this reliabilty well known to users. It could be done in months rather than years, with relatively modest investment; but on the evidence to date, we aren't quite there yet.
If we (or rather users) do have to pay through the nose for having a few bits delivered from our servers all the way to their mobiles… do you think you might be able to do the job and actually deliver them?
