Speaking Engagements & Private Workshops - Get Dean Bubley to present or chair your event

Need an experienced, provocative & influential telecoms keynote speaker, moderator/chair or workshop facilitator?
To see recent presentations, and discuss Dean Bubley's appearance at a specific event, click here

Wednesday, November 29, 2006

IMS handset clients and JSR281... resurgence of mobile Java?

I'd more or less given up on mobile Java being relevant for anything other than occasional gaming downloads. It just doesn't seem to have made the jump to anything remotely "mission critical" on handsets. Numerous initiatives like SavaJe's seem to have disappeared without trace.

But suddenly, a few separate discussions have referenced the very unglamorous acronym JSR281, which is the proposed Java extension that would give a Java app access to IMS capabilities on a device. This could be a good way to circumvent the woeful lack of standards around mobile handset IMS client frameworks. So, potentially, an innovative developer could write a cool messaging application or interactive game in Java, exploiting JSR281 APIs, without having to expensively port it to the various bits of proprietary on-handset IMS software middleware from NMS, Ericsson, Ecrio, Movial, Comneon, Qualphone & all the others.

It's unclear exactly when JSR281-enabled phones will ship, but my bet is Q3-2008 for massmarket devices, perhaps with a few high-end ones a bit earlier. That probably means that decent 3rd-party applications to exploit Java-IMS combinations on phones will start to appear in early 2009.

At last! The new mobile killer app is....... fixed-line VoIP

I'm at an IMS conference in Geneva at the moment, and it's helped me crystallise various separate bits-and-pieces I've observed into a coherent theme.

The next "big thing" for mobile operators is..... offering own-brand VoIP services over fixed broadband.

The logic goes something like this: all our best customers are getting ADSL or cable at home. This is inevitable; broadband using fixed-3G is a tiny niche. At some point, they will start to use their fixed broadband for VoIP - maybe bundled from the ISP or telco if they're aggressive about transitioning from PSTN, or else Skype or other Internet-based VoIP. This is, again, inevitable. Given that they're getting VoIP anyway, why not an operator-branded and -billed service?

There are variations on this theme as well - maybe offering VoIP-enabled USB sticks so travellers can use them in Internet cafes, for example, hooking into their usual operator services like SMS via a PC.

And, of course, offering fixed VoIP gives mobile operators a reasonable application to use as economic justification for an IMS deployment. Given that there won't be mobile IMS-capable handsets available in volume for several years, it makes sense to target other IMS-capable endpoints like PCs. And fixed broadband gets around the thorny issue of indoor 3G coverage that would otherwise limit the usefulness of real-time or multimedia mobile IMS applications.

Certainly, there's considerable vendor push already. I saw a Siemens presentation yesterday that specifically focused on this issue, and Nokia has this morning announced something similar in its Communication Suite offering.

So for example, it seems probable that both O2 and Vodafone will leverage their broadband partners/acquisitions to offer VoIP. The Skype/3 deal is obviously a different approach to the same opportunity.

Also, long term, this sets the scene for true mobile VoIP, which will happen eventually. It should be easier (and more valuable) to link existing fixed VoIP customers with wireless VoIP at that point. Gaining a foothold is essential, as otherwise the whole VoIP world will be even more dominated by the fixed/broadband/Internet/enterprise providers.

Friday, November 24, 2006

Great article on UK low-power GSM... apart from the missing bits

There's an articulate and interesting article on picocells and the UK's new low-power GSM licences over at The Register.

Unfortunately, the author appears to have overlooked the announcement of Teleware's launch of Private Mobile Networks (PMN) a month ago....

... and also the minor fact that it's not just Opal/CPW that has an MVNO available to roam onto, so of course does BT.

I'd also disagree strongly that "O2 has a headstart" as the type of usage case posited - home femtocell/gateway at <£100 + Be's DSL + roaming onto macro network - is unlikely to be a feasible commercial offering before 2008 at the earliest. At the moment, there's still no real-world major tests of standalone femtocells, and I'm not aware of any having been announced that are integrated with a DSL/WiFi gateway.

Thursday, November 23, 2006

GSMA survey & MMS.... maybe not as bad as my initial reaction suggested

I have to say I've become deeply wary of missives from the GSMA recently - there's certainly been plenty of blinkered & unrealistic 2003-vintage "mobile rules!" hype of late. So the latest one about "Strong global demand for MMS and mobile email" raised my eyebrows as well, especially as it's based on a consumer survey, described with minimal detail in the press release.

Now, leaving aside the fact that the GSMA clearly refers to mobile email through gritted teeth, as it's usually delivered via a BlackBerry or other non-operator server, the results look more or less fair enough. Obviously SMS is at #1 in terms of preference, but I'll admit that I'm a bit surprised that MMS rates so highly, especially above web browsing/search. The survey will make unhappy reading for mobile TV advocates too, and notable by their absence are any mention of music downloads, push-to-talk, community-type services, ringtones or presence. The fact that nobody is remotely interested in video calls should surprise nobody.

Given my general skepticism about consumer surveys, I asked for a bit more detail on the methodology. In terms of demographic breakdown it seems reasonable, although the fact that the survey is web-based will tend to be a bit self-selecting, even in markets where Internet penetration is high. I have a suspicion that people who can be bothered to fill in web survey forms are probably, on average, geekier/techier than the population as a whole.

What is clearly nonsense, however, is some of the phrasing in the release. The idea that MMS is "indispensable" for more than 40% of consumers is farcical. 800 million people couldn't live without photo messaging? I challenge any of the readers of this post to name just three MMS-addicts from their personal acquaintances, friends, or family.

I guess the line "40% of users who can be bothered to fill in a web survey form find MMS vaguely entertaining when they're drunk & taking stupid pictures of their mates in the pub" wouldn't have quite the same ring to it....

Blue-sky stuff: wireless power

Spotted a fascinating article in this week's 50th Anniversary edition of New Scientist magazine (essential reading for anyone interested in what's what in the wider science & technology arena btw).

A group of scientists at MIT are apparently working on something called "evanescent coupling", which, if I understand it correctly, involves switching magnetic and electric fields in a set of two matched and "resonant" copper rings. Powering up one of these induces current in the other - useful if the first is attached to mains electricity, for example. Now, although I have a physics degree, I have to confess that electromagnetism was one of my least-favourite parts of the subject, so I'll have to take a lot of this on faith.

The practical upshot is that it might be possible to "remotely" charge a wireless device at a distance of up to 5 metres. Could be rather useful for the mobile industry, methinks.....

Of course, actually putting this into commercial use is an awful long way off - the MIT guys are still at the stage of computer simulation of all of this, and are apparently now trying to build a prototype.

Sidenote: I spoke to a company called SplashPower with a charging "pad" which could transfer power to phones & other devices at a very short range, like an electric toothbrush, about 3 years ago. It used a sort of induction loop thing attached to the battery. Still seems to be around but I haven't seen much in the way of real-world products.

Wednesday, November 22, 2006

So... any more UMA handsets in the pipeline?

At various conferences this year, people have been anticipating as many as 20-25 WiFi UMA-capable phones on the market by mid-2007.

Well, given that the current crop are all pretty long in the tooth - Nokia announced the 6136 at 3GSM in February, as did Moto with the A910. The Samsung T-709/P200 was also being talked about in the same timeframe, although it was only formally announced in July. The LG-CL400 was announced back in Sep 2005, but as far as I know hasn't shipped commercially as part of any of the current UMA services. There was also a Chi Mei ODM one, that seems to have disappeared too. (plus the Bluetooth V560 & RAZR)

But other than those.... no announcements, no launches, no sneak photos circulating on the web. Unless there's a huge imminent launch of several new ones, given the typical 4-8 month test & release cycle that seems to be normal with FMC devices, I can't see too many more being ready to ship by Q2 2007.

Mind you, there's a BT analyst event about wireless tomorrow, so who knows, perhaps I've spoken too soon....

Monday, November 20, 2006

Reaction to 3's flatrate move

Interesting to see the variety of reactions to 3's announcement last week. Several of the comments on this blog have essentially said "well, they've messed it up in the past, why should this time be any different?" - which is fair enough, although clearly it may provoke a reaction from the other operators before any signs of lack of execution become clear.

Keith at TeleBusillis has a similarly cynical take, invoking both Monty Python and The Lord of the Rings as amusing analogies.

Then there's the "acquisition" angle. But even if Voda buys 3 tomorrow, it seems probable that the usual lengthy acquisition & regulatory process would mean that X-Series would launch in a blaze of publicity anyway. The mobile broadband Internet cat is definitely out of the bag, in my view.

On the other hand, there's been quite a positive response from the business press, such as the Sunday Times' opinion headline of "3 finds the right way to do mobile internet"

And I have to get a riposte in to Keith's Tolkeinesque characterisation of "The Internet Barbarians" slavering like an army of Uruk's at the gates of telecom.... maybe we should turn it round and consider Frodo 3 and trust sidekick Web'n'Walk breaking into the walled garden of Mordor and confronting Arun Sauron?

Femtocells - Ericsson and O2

I see that Ericsson has jumped into the fray with a femto product. I haven't yet been offered a full briefing on this, but will most more as I find out.

This line in the release is intriguing: "it offers full integration with the radio macro layer. The same radio frequencies can be used both by the Home 3G Access Point frequency and the outdoor radio macro network allowing maximum usage of the scarce and expensive radio spectrum". I guess this means that some of the RF planning is simplified - although to what density of femto's is unclear, or whether the operator's existing macro network needs to be from Ericsson. Given that many of the femto's will also live "inside" a firewall embedded in a DSL/cable gateway, I'm also unclear as to how the management of this might work - unless it's the same operator's customised gateway box rather than a retail Linksys/D-Link/Brand-X one.

It also appears that O2 made encouraging noises about femto's last week at a press event. A couple of articles cover this - Peter Judge at Techworld mentioning an Ericsson device in his piece, but with a confusingly contradictory mention of an ip.access box in this one.

Separately, I notice that Ericsson announced an update to its IP-PBX range, including a "corporate telephony client" for smartphones. Unfortunately, it appears to only support devices from the company's part-owned SonyEricsson range. This is a huge mistake, and one which indicates how far out of touch Ericsson is on the realities of fixed-mobile convergence in the enterprise. Device choice is extremely unlikely to be made with PBX interoperability as the sole criterion, especially from such a small range of suitable products. It is notable that its IP-PBX peer Siemens now has generic support for Windows Mobile devices, while clearly Cisco and Avaya support Symbian S60, Windows and increasingly even Java/BREW.

Thursday, November 16, 2006

3's conversion to the Internet faith.... No IMS... but also no wVoIP...

Filling in a few more details about the Hutchison 3 mobile broadband announcement....

- The company's CFO made a very public & emphatic U-turn, using terms like "tearing down the walls", and "in principle, what's free on the Internet ought to be free on your mobile". He decried operators who priced their data services according to a resource-scarcity model, and deemed the idea of tariffing connectivity so as to dissuade people from using open-Internet services as "unacceptable". And this is from the guy in charge of the financials, not marketing...
- contrary to various reports, the Skype service is NOT wVoIP - it's a circuit-switched call from the phone into a Skype VoIP gateway in 3's network. I suspect this is because of both radio resource inefficiency of VoIPo3G, and the limitations on indoor coverage. The Skype client on the phone uses IP-based presence, which works over 3G or 2G.
- although there are various proxy servers for things like the IM applications and (I think) Orb, none of the things announced today use an IMS in 3's network. No CSCFs or all the rest of it - most is just a big fat pipe straight to the real Internet. There's some policy management stuff in there to maintain network integrity & throttle back particularly heavy media streams (and, I suspect, watch out for things for filesharing), but it sounds to me that 3 has launched a whole bunch of new appealing things without any of the complexity of IMS.
- Chatting over lunch, it seems that many of the people at 3 share my view that 900MHz refarming & Elisa's trial in Finland is indeed of critical importance.

Overall, I saw the whole event as an example of "beautiful heresy". All the usual so-called operator nightmares - Skype, flatrate data, applications not services, free IM (potentially replacing text), placeshifted media - are being embraced.

Yes, we still don't know the price, but to be honest, it doesn't matter. I think there are plenty of people who will instinctively want a mobile ISP service which enables them to use their handsets the way they use their PCs & broadband modems. Having a select range of tuned / tweaked / customised / optimised Internet brands like Skype and Sling is the icing on the cake. And the fact that they are not subject to extra per-use charges is the cherry on top.

I know I'm sometimes very cynical about mobile operators, but I have to say I've been genuinely astounded by all this. I've been jokingly using the term "legacy mobile-only operators" for some time in comparison with more Internet-savvy FMC carriers. But this announcement makes it clear that not only are some mobile operators going to try & keep up... but that many others are now beyond "legacy" and heading towards "history".

3 finally "gets it" and launches flat-rate ISP-style mobile broadband (and Skype)

Blogger binned my first post, so this is a very quick "hooray!" to 3 for launching a broadband flatrate ISP-style mobile offering, complete with optimised versions of Skype, Sling, Yahoo & MSN.

Given that I heard the 3 CTO stand up a few years back and denounce anyone who "wanted Internet access on their mobile phone", as a way of defending its prior ridiculous walled-garden-only policy, I think this is a great move.

"This charging structure overturns the traditional telephony model of charging per minute, per message, per click, per event and per megabyte"

"Customers in the future will be attracted by greater and greater choice, and higher and higher usage levels, for fair, attractive and transparent access fees."

"Sending and receiving text instant messages with an X-Series mobile will be free"

"Skype to Skype calls on a 3 mobile will be free"

Basically, this looks like it is starting where T-Mobile's pretty decent Web'n'Walk leaves off. The customised Skype & Sling functionality looks particularly smart. 3 has recognised that there is a sizeable market of ADSL/cable-educated people who want a good quality pipe, with a couple of optimisations to take account of handset form factor. Not trying to do something idiotic like creating an own-brand IM is a sign of real-world Internet-savvy maturity that many operators would do well to embrace.

The split-tariff flat rate (depending on whether people use video or not) seems like a sensible compromise between user expectations of a broadband service, married to the practical considerations of resource-constrained radio spectrum capacity and backhaul from cell sites.

Sure, there's still a bunch of questions - can you download other 3rd-party Internet clients for VoIP or messaging or mapping? Is the 3 Skype actually VoIPo3G, or just a normal circuit call into a gateway? What happens with "rich" web browsing like YouTube, MySpace and various Flash/AJAX type things? Roaming? Is any/all of this IMS-based, or if not what else is sitting in the 3 network?

But overall - this looks like a proper Internet-on-mobile proposition. At first glance, and obviously depending on the pricing & the fine details, it gets top marks.

I'm off to the launch event, which involves an expedition to Hutchison's painfully-inaccessible offices in the wilds of Battersea.

Tuesday, November 14, 2006

Quadplay? I feel lucky I can get single-play.....

I'm about to move house, so I thought I'd be able to take full advantage of all the wonderful new triple- and quad-play services we keep hearing about. I thought I'd be able to rationalise my current combination of BT landline, O2 mobile, Pipex broadband - and maybe get digital TV thrown into the mix.

Well, that was the theory, anyway.

It turns out that my new place (in central London, not the middle of nowhere) needs to have its BT line re-connected - which I've been told could take more than 2.5 weeks, plus then another 5 days for broadband. Given I'm travelling in the middle of all that, I'd essentially be looking at mid-December to be up-and-running with phone & Internet. Nope, I don't think so. Goodbye BT.

As far as I can see, none of the LLU providers can actually reconnect a dormant line - BT has to do this themselves. No sign of O2 being able to offer me Be services either - and Be's website tells me it needs a BT line. Ditto Pipex.

But... NTL has a cable running down the street. Great - so after a confusing look on their website, I call the sales number. Only a minimal game of telephone hockey ensues, and I get through to someone helpful. Yes, I can get 10MB broadband, at the princely price of £35 a month, although apparently they can install it within 5 days. But... and this is ultra-weird for a cable company... they can't provide me with either TV or telephony. Never mind Naked DSL, I'm being offered Naked Cable Modem service. Something to do with the infrastructure in my part of town, apparently. Very odd. And they don't do an integrated home gateway if I want WiFi as well, so it's back to a random standalone router bought through retail (so I assume NTL/Virgin isn't planning a dual-mode solution any time soon....)

So... my next idea is to get an NTL cable modem & then run VoIP over it. Great idea, but then I realise I can't port my current BT number to most UK VoIP suppliers as they're not Ofcom "PATS" providers. There are a couple of firms I'd never heard of that can do this, so I guess I'll have to check them out.

As far as I can see, there aren't any UK fixed-3G modem services available, and they'd be too slow/expensive anyway. Urban WiMAX doesn't offer service either, and it's aimed at the much higher-priced symmetric business user market anyway.

Now, maybe I'm the exception here... but if this is in any way typical, I'm not holding my breath about bundling for very long.

Vodafone results - reading between the lines.....

I'm trying to dissect some of Vodafone's huge swathe of pronouncements today. Not only its interim financial results, but also an advertising deal with Yahoo! There's a webcast later, but some initial thoughts:

- notable omission in Arun Sarin's statement of much about mobile-centric content or data services, except for a vague line "applications using the benefits of mobile broadband following the introduction of HSDPA" and a reference to the Yahoo! deal. Instead, the focus is firmly on cost-saving, fixed-mobile substitution, and voice minutes
- Vodafone appears to be heading back to a position of POMS (Plain Old Mobile Services) supremo, with its rapidly growing emerging markets business.
- there a certain disdain of its own recently-announced fixed broadband products "We will continue to pursue a mobile centric approach, focusing on the core benefits to customers of mobility and personalisation, and will resell fixed line technologies only according to customer needs" . Question - would you really want to buy any service that is offered so grudgingly?

Now, lets look at data services. This line is very telling, and highlights some comments I made back in May at its last results "Data revenue increased by 27.2%, or 29.1% on an organic basis, with the primary driver being an additional 7.1 million 3G devices registered on the Group’s networks since 30 September 2005, bringing the total to 10.4 million devices, and in particular, the increase in devices in the business segment"

Lets dissect this further. For the six months to Sep 30, revenue from non-messaging data services (ie excluding SMS, MMS) was £650m, broken down to £190m in Germany, £134m in UK, £122m in Spain, £89m in Italy and £91m in "other Europe", less £23m of "eliminations", which I presume is on-net roaming fees, plus another £56m in emerging markets (less another £9m eliminations)

Now, let's look at these "devices in the business segment". There were 1.0m 3G Vodafone Connect Cards in use at the end of Sep, versus 0.66m in March and 0.81m in June. Let's say an average of 850k over the period. Unfortunately, the company doesn't appear to have given the number of Blackberry users this time around - presumably as it also now has alternative email platforms. However, it had 426k of these in March, so a reasonable estimate is for this to have increased to an average of perhaps 550k email users over the 6-month period.

Now, there's a lot of different tariffs on the data cards, so it's difficult to estimate an average. There are some low-end options, but also hefty slugs of roaming at the top end. Last time around I suggested £40 / month, but let's be more conservative and say £35 now. £35 x 6 months x 850k = c£180m . Now, although Voda is now pushing its email solutions down to SMEs and consumers, for the period under consideration it seems reasonable to assume the majority of users are still corporate-grade at perhaps an average of £25 per user per month. So, £25 x 6 x 550k = £83m. So, 3G cards + business email = £260m. Now, add in a bit of inbound data roaming, plus some 2G data per-MB services, and we get to perhaps £300m of the overall data services revenue being "pipes". Out of £650m , that's a very big slug indeed.

I wonder if this is why a search of the 48-page results document only yields one mention of the word "content", in the Forward-Looking Statements fine print ("the Group’s ability to develop competitive data content and services that will attract new customers and increase average usage"), and the only reference to music or TV is in the context of a new flat-rate data plan in Germany "Germany has had particular success from bundling data services with a new contract tariff which enourages data usage by offering free mobile TV, surfing the Vodafone live! portal and music downloads for a flat fee each month".


I just noticed something extra on data services. In June 2006, the company reclassified "messaging content revenues" - I guess mostly premium SMS used for ringtone downloads and similar -from "messaging" to "data". The financial small print says "The impact for the quarter ended 30 June 2006 is to increase the total Group data revenue as a percentage of service revenue by 1.0% to 4.9% and to reduce the total Group messaging revenue as a percentage of service revenue by 1.0% to 12.5%". Assuming the 1% is the same in Q2, this means that another £145m of the data total is premium SMS/MMS.

So, in short, my estimates:

Total non-messaging data = £650m
of which:
- Data cards = £180m
- Premium SMS/MMS content = £145m
- Email (mostly Blackberry) = £80m
- All other (music, TV, inbound data roaming, business services like telematics etc) = £245m

In other words, I estimate that all the Live! and consumer 3G services generate around £40m / month. Now clearly the 30m Live devices (of which 10m are 3G) will be among the major users of the premium SMS services, but the whole Live! infrastructure is clearly about more than just ringtones. I'd estimate that a Live! device probably needs extra subsidy (£10? £20? more?) to cover the cost of extra software, hardware & design/test. Add to that the costs of Live's delivery platforms, hyper-scale marketing and content acquisition, and it's financials can't be looking too pretty.

Monday, November 13, 2006

more significant than it appears: Nokia and Elisa demonstrate 3G at 900MHz

This morning I had a bit of a go at Nokia for putting out a release with a rather over-the-top headline about Mobile TV. Well, this afternoon's press missive from Nokia seems to go to the opposite extreme, understating what is quite possibly the most important thing to have happened in the 3G area for about 5 years.

"Nokia and Elisa complete the world’s first commercial WCDMA900 data call" . Yawn. Likely to get completely overlooked by most of the world's media, and probably half the specialist telecoms press as well.

Whereas in fact, this is possibly the first move of something that could actually make 3G a more mainstream platform, and solve one of the thorniest problems the technology possesses - indoor coverage.

The problem is this: at 2.1GHz, the main frequency band for WCDMA, penetration into buildings - as well as range more generally - is generally poor. It gets soaked up by walls, furniture and so on. Measured as a function of signal strength as a distance from a window, there's much more attenuation than you get with 2G. Yes, in dense urban areas it's improving, but it's very far from working ubiquitously. And even where you get coverage, because of the way that the CDMA RF works, you may find that a disproportionate amount of resource is taken by a few users on the fringes of signal availability, messing up the network economics & planning. If 2.5GHz gets used for WCDMA as well, it'll probably be even worse. And, ironically, most non-voice 3G apps are used, yes - you guessed it - indoors. Oops. Basically, it's a fundamental flaw in the overall 3G proposition, and it's down to the immutable laws of physics, and the practicalities of siting base stations.

Hence all the myriad solutions for pico/femtocells for indoor 3G, a big push on conventional RF distribution technologies like DAS and repeaters, and the ongoing saga of dual-mode cellular/WiFi. While many of these also offer various other benefits and use cases, a key secondary rationale is just to improve 3G coverage inside buildings.

There is, however, a potential solution. In much of the world, 900MHz spectrum has long been used for the first batch of GSM operators, joined subsequently by PCS at 1.8GHz. If the 900MHz band can be "refarmed" for 3G, many of the indoor (and rural area) coverage problems could be fixed. The fact that Nokia is a big fan of this approach is unsurprising.

I am certain that we will see this example used in lobbying regulators and radiocomms authorities - the big stumbling block is convincing those in charge of spectrum to allow a modification of operators' licences from technology-specific (ie GSM-dedicated 900MHz) to one which also permits UMTS. The idea being that the operator keeps its spectrum but is allowed to use it differently.

Now, this raises some additional interesting questions - if it's not technology-specific spectrum, and the rules are going to be changed - should it be to just permit WCDMA900 (Nokia indicates that it's coexisting with Elisa's GSM), or should it be to a complete tech-neutral free-for-all? And if its just going to be kept by the existing licence owners, should they be charged more for the privilege of running a new (and potentially more lucrative) bearer in that range?

And, to be honest, if 900MHz is going to be up for grabs, I can see an awful lot of WiMAX and other spectrum-hungry folks pushing for full openness.

Either way, this sets an important precedent, and seems to suggest that all the complex dual-mode solutions with WiFi have a finite window of opportunity, before wide-area broadband wireless finally gets to compete head-to-head in the living room, office floor and basement.

"The future of TV will be personal" says Nokia. Well, maybe in some cases....

I'm waiting for for Nokia or another mobile phone vendor to come out in the next couple of years and say "Mobile convergence has won again - only 190m TV sets ship annually, but 300m+ TV-capable phones have been sold".

Last week we saw the precursor to that pronouncement with Nokia's announcement of a report it had commissioned with the London School of Economics, with the press release title of "The future of TV will be personal" . Now, at first glance this report looks to me that it's got a lot of good stuff in it, and, to be fair, PR-speak is always a little hyperbolic. And I am generally a believer in the concept (if not the business models) for Mobile TV. But I'm wondering if the industry is perhaps breathing its own exhaust a bit too much at the moment.

Although I don't profess to know much about the mainstream broadcast industry, it strikes me that where consumer are actually spending money at the moment is on larger TV devices, not smaller ones. Apparently, around 45m LCD and Plasma screens will sell this year, many of comparatively enormous size. Add this to the ongoing adoption of "home cinema" systems, and it seems that for many of the most affluent consumers, TV is actually getting "less personal".

And don't forget that product lifetimes here are much longer, so in fact "installed base" is a more important metric for advertisers and content companies than "devices shipped". (And, of course, another key metric for advertisers is the physical size of any given image). Now, I know that TV-watching patterns are under threat from various sources, including mobile applications and the PC-based Internet. But at the same time, the traditional broadcasters are fighting back with HDTV and other offerings. It is conspicuous that the LSE report fails to comment on some of the more visible corollary trends elsewhere in TV-land (the word "plasma" isn't in the document anywhere).

It will be interesting to sit back in 5 years time and look at the comparative size of the "TV ecosystems" - (equipment sales + advertising + pay TV revenues + licence fees etc) for large-screen home systems vs. mobile TV. While I am expecting to see ad revenues transition towards mobile (ie they'll definitely rise from near-zero roday....), I suspect that the big-screen plasma & surround sound devices will continue to command a much greater "Ooooh, I want one of those!" effect among product buyers, even at the $1000-2000 price points that seem typical.

What may emerge is that plasma/LCD screens end up being the platform of choice for "massmarket" content (ie big sports, movies, major news & documentary programming, soap operas etc), while some combination of the mobile and the PC become the preferred screens for the "long tail". There will be comparatively few "Two plasma/LCD screen households", and so personal devices will start to supplant smaller TVs in the kitchen, bedroom etc. (In fact, I heard a rumour that a significant % of viewing time in one of the better-known mobile TV trials was actually conducted from the throne in the loo....)

One last thing - it seems to me that one of the big plus points of plasma and LCD screens is their thinness. Sound familiar to anyone? Now, remind me again which mobile phone manufacturer hasn't exactly over-embraced the "thin" concept?

Friday, November 10, 2006

Quiz time

OK, you've probably already read about this Telco 2.0 survey on Martin or James' blogs ... but if not, have a go.

(And although it doesn't say so, it's quite clever & picks up where you left off if you half-complete it & then go back & finish it later....)

Thursday, November 09, 2006

Sharing broadband with other people for VoIP.... FON vs Free

I've just got back from the VON conference in Berlin - I'll be posting over the next few days about VoIP, SIP and other good stuff.

One thing did stick in my mind though - you know all the fuss about FON breaking some ISPs' terms of service because you're sharing your broadband access, by letting other wireless users do VoIP via your access point?

Well, apparently French operator Free (owned by Iliad) is going to be doing the same thing itself, with its own customers' APs and broadband. As a Free customer with a dual-mode handset, you should be able to use the WiFi capabilities of all the other Free customers' home gateways - potentially over a million hotspots in residential areas. The gateways have both public and private SSIDs.

The question is, if someone else is using my broadband, officially sanctioned by the ISP, how much are they paying me? And what happens if I live above a popular cafe or next door to a university - how many concurrent users can sit on my network, and does the Free box prioritise their realtime voice traffic over my email or web browsing?

Tuesday, November 07, 2006

Vodafone new non-exec.... adding enterprise IT knowledge looks a good move

One of my major criticisms of most mobile operators is that they fail to understand - and focus on - their corporate customers. Most seem to be run by consumer-centric boards made up of directors' with backgrounds from Coca Cola or Procter & Gamble. Unsurprisingly, this results in a focus on entertainment, "content", and flashy consumer advertising and branding strategies. Usually, corporate-related issues like systems integration, migration, and interoperability with IT / LAN / WAN sides of businesses get overlooked.

So against this background, Vodafone's appointment of the ex-CIO of HSBC, one of the world's largest banks, has to be seen as a good sign.

VoIP + SMS = headaches

I'm recovering a post that I half-wrote a couple of weeks ago while at the VON Europe conference in Berlin, but didn't get around to finishing.

At the time, I was surrounded by a large number of free/disruptive/parasitic VoIP enthusiasts, as well as quite a lot of operators deploying carrier-grade PSTN replacement VoIP through IMS and other platforms. Also unsurprisingly, wireless and mobility featured quite highly - lots of talk of WiFi phones, dual-mode handsets and so on. A lot of this is good stuff. I met with various wVoIP firms like Fring and Truphone, as well as companies like Rebtel that essentially do circuit-switched callback using some form of either PC or handset interface.

Subsequent to this I've also had various other calls and briefings, particular on this "signalling over IP / voice over circuit switched into a VoIP gateway" theme, which seems to be flavour of the month. There was 3 with its Skype/iSkoot handset client last week, and I had a briefing yesterday with the blogosphere's apparent favourite, TalkPlus.

But, for the most part, there is something missing. The one thing that the existing cellular industry has created, that has taken on a life and importance of its own, and which can't be IP-substituted in as easy a fashion as circuit-switched voice.


While I disagree with many of the things that some of the more fervent mobile evangelists hold dear, I do agree that SMS has undeniable momentum, and is almost completely pervasive as a core element of mobile communications user experience. Currently, it's a market worth something between $50-75bn, depending on how you calculate it (eg including SMS-delivered content revenues net), with somewhere north of 1.2bn active users, some of whom send 100s of messages per day.

Now, I completely agree that various IM / IP-based pseudo-SMS services will emerge to cannibalise this market over time. The margins and costs of SMS are pretty ridiculous in most cases, and based mostly on "ignorance based pricing". It's a sitting duck, frankly. So I expect lot of operators will try to use IM as a form of "SMS plus" as a preemptive move, as Telefonica did recently, although I have my doubts that many of PC IM users will be hoodwinked into thinking this is anything new.

But this is a diversion for the current wVoIP and IP/cellular voice hybrid companies. For most users, SMS on mobile phones will continue to be extremely important for the forseeable future, especially as it is a lowest-common denominator service. The fact that all the IM competition will drive down SMS prices is, perversely, likely to entrench it still further.

So... if you are looking to hybridise VoIP and cellular, on single- or dual-mode phones, for consumers or enterprise, with wVoIP or circuit-to-VoIP-gateway, you need to have a good SMS story. And a good SMS user experience. It's one of the major flaws in many of the current SIP-based dual-mode WiFi offers, and perhaps UMA's only redeeming positive feature.

Some companies in the space do have an SMS story. Truphone's is impressive. Skype has an outbound SMS function, but it's not ideal, although perhaps it can get away without it as its primarily a PC platform anyway, not a start-from-scratch "mobile community" like the others. Apparently, TalkPlus will support in early 2007 - if so, I might be more impressed with its pitch than I was yesterday. (btw - TalkPlus linked via its Skype gateway was abysmal yesterday, after I'd been using SkypeOut all afternoon with fine quality)

Implementing SMS integrated with VoIP isn't easy, either. As well as the SMS Gateway or SMSC in the network, the on-device user experience has to be very good as well. Ideally, it will tie in with the "native" SMS client on the phone, using the same interface, the same predictive text, work nicely with phone book and SMS archiving and so on. Doing a kludged secondary SMS client in a Java client or Symbian app is going to be challenge.

One possible approach (a bit like Skype's in a way) is to have the VoIP number/name as a "pseudo-fixed" one, to which other people wouldn't expect to send an SMS. This is sort of OK if you get a "second line" on your handset from VoIP firm, or a geographic one as in O2's Genion or Vodafone Zuhause. But the type of "disposable" mobile numbers as pitched by companies like Tossable Digits or TalkPlus is useless without SMS support for most use cases - for example, TalkPlus' much-hyped dating example. Who organises a date these days purely on voice?

Bottom line - sorry wVoIP developers, but you're going to have to spend a lot more time & money than you anticipate on this legacy - but extremely popular - messaging technology. And don't suggest any half-arsed "oh, just use IM" alternatives without really thinking through the realities of on-device integration.

SonyEricsson and UIQ

Interesting. S-E has today announced that it's acquiring the UIQ interface business from Symbian. I guess in hindsight this really isn't any big surprise - apart from S-E, nobody else really makes UIQ phones in any meaningful volume any more (I think there's still a Moto phone in production, and maybe an ODM one). I can't imagine Symbian's other shareholders can have been too chuffed at effectively subsidising S-E's software development.

More interesting is what S-E will do with UIQ in future. It is unclear to me whether this signals a push to drive Symbian further down its product range - perhaps as a better way to do operator customisation & multitasking, or as a platform for IMS services. Given S-E's recent good form with handsets like its K800i, it must be in two minds - not messing about with an existing high-end featurephone platform that works fine, vs. having to continue to evolve this for the uncertain applications of the future.

If S-E does use UIQ in a greater range of devices, and it also carries on creating must-have phones, I reckon that Vodafone may already be rethinking yesterday's software platform announcement and contemplating adding a fourth OS to its short-list.....

Mobile Web 2.0 - thoughts + book review. The emperor has a few clothes

A couple of things have recently made me consider the issues around porting Web 2.0 (social networking, user-generated content, AJAX & all that good stuff...) to the mobile domain.

Firstly, Ajit Jaokar from Open Garden and Forum Oxford sent me a copy of his book on the subject (co-written with Tony Fish). (some more specific comments are below).

Secondly, there has been an awful lot of noise from handset vendors, software firms, operators and others about things like user-generated content. The Symbian smartphone show, for example, was saturated in this.

I'm in two minds about all this, to be honest. I'm nowhere near as enthusiastic about Web 2.0 (fixed or mobile) as some of the louder evangelists. Yes, it's important, but no, I'm not convinced it's going to completely change the world. I actually think that it's a symptom of the Internet turning into a more usable but over-featured platform.

In some ways (warning, heretical comment ahead) Web 2.0 is like Windows or Microsoft Office on a PC - or even a PBX in an enterprise. There are hundreds of features & subsidiary applications, but you only ever use a dozen regularly, and some more occasionally. Sometimes you'll have an "a-ha!" moment where you find something new & useful, or you see a colleague doing something interesting. But everyone's got their own subset of preferred features.

The same will be true of Web 2.0 generally, and mobile Web 2.0 in particular. In the mobile domain, this will mean that some people will use some features some times. This will limit the spread of community, and will certainly mean that the mobile device will not become the central hub of most people's online lives.

So, for example, I write this blog on my desktop PC, or occasionally a laptop if I'm at a conference. But I would never use a mobile phone, as nothing I post here is that time critical that it can't wait until I'm in front of a proper keyboard & mouse, so I can add links, edit & re-paragraph and so on. Another thing - I can't see the point in tagging, and I don't have the inclination to find out. So, I'm picking and choosing which bits of the Web 2.0 universe that meet my immediate personal goals. The rest of it.... well..... life's too short.

Similarly, I'll take photos on my phone, but I can't see myself uploading them direct to any sort of Web 2.0 site. I'll wait, and transfer the uncompressed image files to my PC, where I can use them in multiple ways - upload, email, archive, print. Occasionally I'll send an MMS, but if I'm honest that's usually because I'm in the pub & taking silly photos.

So for me, the phone is about "capture" of data (eg pictures, maybe with some location context), plus time-sensitive applications like checking my business email when I'm out. I'll occasionally read blogs & web forum stuff on my mobile device, but I can't be bothered to "generate content". I could possibly envisage using IM on a phone, as long as I don't have to pay for it, and if it interoperates perfectly with Yahoo. I'm not an avid MySpace user (under-used profile I keep meaning to work on), nor Flickr, and I watch YouTube but don't contribute.

For other people, their priorities and preferences will vary. But nobody will use everything that Web 2.0 has to offer.

And for Mobile Web 2.0, there are other practical considerations which will mean it will be even more fragmentary. Everyone's handset has different software & capabilities. Coverage varies considerably - especially indoors, where 3G is mostly used, and where users will often tend to generate or view content. Data tariffs vary. Different operators will have different service preferences & partnerships. National differences are huge. Handsets' short life cycles will mean groups of friends have imbalanced capabilities between their phones. People will have capabilities split across multiple devices & operators. Frankly, although certain "islands" of mobile web 2.0 will exist, possibly at a nationwide-level, I just cannot see any applications gaining the same amount of viral-induced global acceptance we see on the PC-based web. For example, Keith at Telebusillis comments on high usage of MySpace among Helio subscribers - but I just cannot ever see more than a few % of the total global MySpace population - or whatever is temporarily cool in Oct 2009 - using it on their handsets.

Bottom line - Mobile Web 2.0 seems over-hyped before it's even been properly introduced, particularly around the notion of a world dominated by "user-generated content". On the other hand, the idea of "perpetual beta" makes a lot more sense in a mobile industry that has historically taken years to get substandard things to market. There will also be a few areas where "collective intelligence" (eg voting) is important - but they won't change the world.

OK, with that out of the way, more on Ajit's book. It's got some great stuff in it, and is a comprehensive treatment of a complex subject from the user's point of view. It introduces a lot of the key themes, albeit (unsurprisingly) with more of a tub-thumping utopian view than the one I espoused above. I like the ideas of things being more browser-driven, and as I said I like the "perpetual beta" concept applied to mobility.

For someone like me, who hasn't really "studied" PC-based Web 2.0, it's great to go through all the underlying principles, and pick out the ones you (a) agree with, and (b) use or exploit personally. That said, just reading about "social bookmarking" or "tagging" still doesn't make we want to spend time doing it, especially on a mobile device.

However, I think the book assumes a bit too much of the technology, and ignores some of the less-convenient factors around conventional mobile (and Internet) usage. It doesn't take account of the fact that on-device capabilities (memory, camera, processor) grow at a faster pace than connectivity, which will tend to prioritise offline over online applications. (This is a separate post I need to do at some point). It doesn't mention coverage issues, nor how prepay users can fit into this new world (absolutely critical in many markets). It's a bit too convergence-oriented, and doesn't take account of multiplicity - multiple devices, multiple numbers, multiple IDs etc. However, it has a lot of useful case studies & pointers to the "target applications" which may become more possible over time.

Overall, the book is a good starting point, and you can use it to generate your own divergent trains of thought. It catalysed my thought process on how mobile mashups could work in reality, for example - if you start with 2 GB of initial mapping data on a memory card, and just "top up" over the network where necessary, and add additional variables generated both locally and through the network.

Bottom line on the book: Read it as a backgrounder, and as a starting point for what what might occur, if awkward realities didn't get in the way. Don't buy into the hype around Web 2.0 being the panacea for mobile communications for everyone, but pick & choose which bits you believe are feasible. Then look for niches where the problems don't matter, or invent solutions to fix the issues raised.

Monday, November 06, 2006

Vodafone's handset OS strategy

Vodafone announced this morning that it's cut a deal with Microsoft on handset software. Leaving aside the fact that the company still uses the appallingly anachronistic word "terminal" in its PR headline, this is actually a pretty interesting strategy.

The Windows Mobile deal is part of a Voda's three-platform appoach to future devices, also embracing Symbian/S60 and - interestingly and up-till-now hardly mentioned - Linux. This makes Vodafone look similar to NTT DoCoMo's approach to handsets, which is based on its own flavour of Symbian OS, plus Linux and recently also Windows Mobile.

The PR has an interesting paragraph:

"Over the next five years, Vodafone expects to focus on supporting three standard terminal platforms across its portfolio of mobile phones: Microsoft Windows Mobile, Symbian/S60 and Linux. The first device to use the software produced under the agreement is planned to be with Samsung and is expected to launch in the first half of 2007."

The first thing to note is the realistic 5-year view, which clearly shows awareness of the need for Moore's Law to catch up & bring up low-end devices towards the level needed to host a proper OS. The second thing is the qualifiers "expect" and "focus"... ie a get-out clause which enables Voda to add additional platforms as well if things change. The third thing is "which Linux"? (or is it actually several variants....)

And the fourth thing is to try & map how any of this fits to the major device vendors, and how this could fit around their own software roadmaps.

Nokia - OK, no surprises with S60 for high-end devices, but what happens to S40, its massmarket platform? Does it (a) disappear as S60 and Symbian moves downmarket? or (b) maybe it turns into a layer on top of a Linux OS like the 770 Internet tablet? or it it (c) Nokia ignores this & persuades Voda it needs special dispensation as it's market leader?

Moto - over a 5-year period, most of its high/mid devices should migrate to its inhouse Linux platform. May still have some low-end embedded ones, though, and if it continues to create "must have" phones like the RAZR, then again Voda might need to be flexible. Plus there's probably going to be a few MS-powered phones for enterprise, especially since the Symbol acquisition.

Samsung - seems to still be promiscuous, supporting Linux, Symbian and Microsoft. Still lots of inhouse embedded OS use, though....

SonyEricsson. Hmm, this is a tricky one given how popular some of its recent phones have been. It's going to be narked at having to change its embedded (and very good) featurephone platform, and also possibly ditch its UIQ smartphone platform. I wonder if S-E is also going to have a good reason to be on Voda's "exception" list.... although maybe it's got a Linux card up its sleeve....

HTC - no problem, clearly.

LG - again, like Samsung, pretty promiscous with OS's so should be able to fit into Voda's vision

RIM.... oh dear.

The other interesting question is what sits on the top of all these OS's. Will it be a proprietary Vodafone IMS+other stuff client? Will it be a third-party layer like Flash? Or will it vary between targeted customer segments?

And lastly.... I wonder if all this means Java's days are numbered?

Thursday, November 02, 2006

Operators, 3rd party managed networks and indoor coverage

I haven't really tracked the details of managed services in much depth to date. Clearly, there is a general trend towards mobile operators outsourcing chunks of their deployment & ongoing network/service operation to a spectrum of suppliers, many from the infrastructure industry, such as Ericsson, Nokia and Lucent.

A press release on a Nokia deal in Australia caught my eye this morning:

"Nokia will be responsible for managing Vodafone Australia's ongoing network operations covering HSDPA, 3G, GSM and core networks infrastructure including the detailed design, engineering, optimization operations as well as network management, monitoring, fieldwork and maintenance services for the networks"

Now, it's nothing to do with this specific deal or Nokia in particular, but it got me thinking. Where an operator outsources much of its radio and access network expertise, what happens in the future if its needs around indoor coverage, backhaul and so on, change radically? I've written before about the apparent disconnect between many carriers' radio engineering groups and their service/marketing functions.

Networks are not often designed & dimensioned with an eye towards new generations of devices, applications, use cases and so on. The fact that indoor-centric/high-bandwidth social-networking Web2.0 services (YouTube, maybe) may migrate to handsets is unlikely to be directly linked to decisions about antenna placement. Cell optimisation is a essentially a black art, involving much tuning & tweaking by specialists.

So... what happens when these specialists don't even work for the same company? And are governed by a rigid contract set up years previously? Does the "disconnect" get even worse?

"Oh, hi... we work for operator X's video application team. You guys manage our radio network, right?"

"Yeah, well, we've been getting some complaints from customers that our whizzy new OurTube service seems to really slow down, especially when they're using it at school & in apartment blocks downtown. It's disastrous, because we've got this TV tie-in thing where we should be getting a million users doing it simultaneously for a competition".

"Which specific buildings? Haven't got a clue I'm afraid. When can you fix it by?"

"Your managed services contract? How long? Then whose problem is it??!"

Wednesday, November 01, 2006

Hotel WiFi charges - momentum building....

Nice to see Silicon.com putting the boot into rapacious hotel chain operators (and their much-hated overpriced hotspot "partners"). OK, so actually the campaign started here but it's good to see some momentum building.

An open suggestion to Internet travel booking websites - add "free Internet access" as another check-box criterion along with "concierge" and "gym" and "swimming pool". Let customers filter out the offenders; maybe that will encourage the laggards to stop charging $10, 20, even $30 per day for WiFi.

I'd also encourage other readers to act on the principle of refusing to speak at events unless the organisers supply delegate WiFi (and, by extension, therefore kick the hotel/venue owners into shape on our behalf). I am now doing this, and if like me you find yourself inundated with requests to speak/chair, it's a nice simple filter to use to prioritise.