GSMA Mobile Broadband announcement - a billion dollar ho-hum

Lots of fanfare this morning about the announcement of GSMA's big "Mobile Broadband" branding exercise, whereby notebook vendors will put a sticker on embedded-3G notebooks, reminiscent of Centrino or WiFi.

Some good stuff in the announcement, notably a $1bn marketing budget which is obviously likely to help drive the uptake of both embedded and external solutions. But also lots of self-serving propaganda as well (as exemplified by the quote from Gemalto, maker of SIM cards).

But the focus on WiFi hotspots in the PR points out just how narrow a niche the initiative is addressing. Personally, I use HSPA a lot as an alternative to hotspots - but that's because I'm the type of user who really needs access in a lot of different places, and I cringe at the stupid prices charged by the likes of Swisscom in hotels and airports. My next laptop might well have an embedded module - as long as the price is right, and it's not locked to a specific MNO's data plan.

But I'm in a minority - as, I expect, are most readers of this blog. The vast majority of notebook buyers - consumer or corporate - won't be buying an embedded-3G notebook any time soon. Many don't need mobility anyway, and others will be happy with dongles and tethered 3G phones.

I'm buried in other stuff today, but I'll get back later in the week to why the curve of 3G "attachment" in notebooks will be much shallower than that experienced by WiFi a few years ago.

And as for the semi-serious mention of refrigerators with embedded modules in the PR.... I remember the "screen fridge" in all those house-of-the-future demos back in the late 1990s. It says a lot about the balance of reality vs. wishful thinking in the whole announcement.

Multiple devices per account on WiMAX + also wholesale options

I saw an interesting announcement this morning from Bridgewater Systems, about a solution to enable WiMAX operators to control the number of active connections per user/device. While it's being pitched as a sort of fraud-management tool, it highlights something pretty important:

WiMAX services will enable multiple devices per subscription

This isn't news as such - many market participants like Sprint have been saying this was likely for some time. But it's something that the existing cellular providers and architecture are pretty bad at. Yes, you can get family plans and corporate group accounts. Yes, discounts are available where you get both phone and 3G dongle from the same provider, linked at "customer account" level. 

But with very few exceptions, operators don't offer the option for a single subscription (ie "phone number") to support multiple devices. A few operators provide 2+ SIMs, but there's typically a clunky procedure for the user to denote which device is "active" at a given time.

Increasingly, I find myself using two or three screens simultaneously - making a phone call while using my laptop, using handheld device for email while doing something else on my PC, and so on. There's no reason why I should need multiple mobile accounts, any more than I do running two PCs over a single ADSL subscription. Why shouldn't I have two, three or more devices sharing my mobile broadband, as long as my aggregate traffic is less than the monthly cap?

WiMAX has an inherent advantage here in not needing SIM cards, in my view.

Interestingly, Bridgewater also has a wholesale solution for WiMAX providers - another angle in which I'm expecting some differentiation from cellular networks. My view is that there will be a broader selection of partner/wholesale business models in WiMAX, compared with the fairly monochrome choice of MVNOs on cellular. I'm expecting to see regional wholesale networks (eg for local councils), or ones based on specific QoS metrics (eg low-latency gaming MVNO vs secure enterprise MVNO vs. streaming-optimised MVNO). 

To some extent, WiMAX operators will be driven to these niches because they've pretty much relinquished the PC-connectivity market to HSPA/EVDO dongles in the near term. But it also reflects the likely open-minded nature of new WiMAX entrants to try innovative Telco 2.0 business models.

3G Netbook + broadband bundles.... niche?

It's not great surprise to see a number of MNO announcements about low-cost, small "netbook" PCs with embedded 3G - Vodafone+Dell, and T-Mobile+ Asus/Acer [link in German] in particular. Interestingly, Telecom Italia Mobile has gone the MID route instead.

Regular readers will know that I'm skeptical that what I'm calling "Mobile Broadband Computing" products, sold by carriers with embedded 3G, will amount to much. I think there's a lot of wishful thinking going on, which isn't going to fit with the ways in which customers buy and use computing devices.

One thing I haven't been able to track down yet is whether the mobile broadband contracts for these devices will be have standard flatrate dataplan T's and C's, which prohibit, block (or even charge extra) for applications like VoIP and IM. If they do, the proposition will likely fail even sooner than I anticipated. It would be a very serious mistake indeed.

I can just imagine the marketing pitch - "Get our free laptop with embedded HSPA, and ditch your ADSL service! It's like fixed broadband but better in every way! Only, er, you can't use Skype. Or MSN. Or that IM feature built into FaceBook, although as it's a web service we can't see you using it, anyway".

One thing that MNOs will likely find it tricky to grasp is that in the computing domain, you can always give extra applications to customers - perhaps an SMS client, even IMS RCS - but you can't stop them from using their choice of other software. (Unless you're an enterprise IT administrator... and they are doubly certain in not wanting to have PC applications determined by their carrier).

In any case, I'm just not convinced that end users will want to buy their PCs from a mobile operator - and I suspect that those who are most tempted by "Free Notebook!" offers are probably the most likely to prefer prepaid mobile services rather than another 24-month contract. Or those whose credit cards or banks aren't too keen about providing payment plans for PCs.

I'm definitely of the opinion that some embedded-3G offers are really just about massaging up carriers' apparent data ARPU when in fact all they're doing is simply retail finance masquerading as subsidy. I wonder when we'll see cars with embedded HSPA modules for telematics being sold by MNOs too "Our ARPU is now €450 a month".....

One more passing thought on this - given the numerous horror stories about time/hassle in replacing broken or faulty handsets, do you really fancy leaving your embedded-3G PC, with all your data, at the local phone shop if you have a problem with it 6 months down the road? Or perhaps buying separate PC + dongle is the way to go after all?

Could Google save us from the tyranny of the SIM card?

From the department of "not likely to happen any time soon, but interesting nonetheless" comes this report of a Google patent that suggests that future phones could institute an auction amongst available wireless providers, each time a connection is needed.

If there's one thing worse than being a dumb pipe, it's being one that has to compete continually on price with an array of other dumb pipes.....

(This reminds me a bit of the original view of web services, where applications would in theory pick providers of XML components from a directory in the cloud, in real time. Although given that mobile commonly has 5-10 second call setup times, there's probably a bit more scope for this to work for phone calls, than if it was embedded in a enterprise app needing millisecond response).

Convergence gateways - some quick thoughts

I had a meeting this morning at NetEvents with a company I hadn't really spoken to for a very long time - GenBand. They acquired NextPoint last week - itself a combination of the former ReefPoint and Nextone, and one that I frequently encounterer when talking about femto and dual-mode WiFi architectures.

Basically, the company plays in the "convergence gateway" market - which as a telecoms industry sector is an ugly mix of security gateways, media gateways, SBCs, GGSNs, DPI boxes, femto & WiFi gateways and assorted other application-oriented boxes.

Essentially, the sector involves something sitting between the access part of the network (fixed/mobile) and the core applications (legacy CS switches, IMS, NGN, Internet, VoIP etc). This "something" or "somethings" perform a selection of security, control, aggregration, transformational and monitoring functions.

Many companies operate in this general space - Sonus Networks, Acme Packet, Mavenir, Cisco, Stoke and so on. Many of the larger telecom companies like Ericsson and Motorola buy modules on an OEM basis and are this also involved as aggregators.

There is the usual debate about best-of-breed vs. integrated boxes, but I tend to agree with the view that certain things will make sense to blend into unified platforms. The GenBand / NextPoint merger looks quite a smart move against this.

I'm also wondering whether this layer of the network might be the right point to open up network APIs & expose capabilities for various Telco 2.0 business models.

Vodafone Portugal, prepay data SIMs & WiFi...

I'm in southern Portugal for the NetEvents conference. I had a few minutes to kill waiting for my transfer from Faro airport when I arrived, so I decided to see if I could get a prepay data SIM to run in an unlocked phone.

The good news: Just €5 bought me a SIM with a €7.50 credit. No registration, nothing. And for €7.44 the nice lady in the shop set me up with a month's worth of HSDPA (I think with 100MB cap)

The great news: That sure beats £3 or £2 per MB or whatever the ridiculous data roaming charge is on my 3 or T-Mobile or O2 SIMs I have with me.

The bad news: This gives me a separate access point (a Voda Live! one) which doesn't allow me either to send outbound POP3 email, or use SSL to connect to sites like Yahoo, Gmail or British Airways

The slightly better news: There's another APN called Internet Vodafone which allows me full web access. Plus Nokia Maps seems to work, but Google Maps doesn't.

The ridiculous news: I check into my hotel & decide to see what's up with WiFi to connect my laptop. The hotspot is run by, guess who... Vodafone Portugal. Prices for Voda customers are €5 for one hour, €20 for a day, and €30 for three days.

In other words, a day's WiFi costs 4x a month's HSDPA. From the same operator.

So despite my usual dislike of downloading apps to smartphones, I decide to join the tethering fraternity, and install JoikuSpot on my E71, which turns it into a little 3G-backhauled hotspot.

Joy.

I have both phone & WiFi access online for a low price, having navigated through several orders' of magnitude of attempted rip-off. I haven't signed any piece of paper with terms & conditions, I paid cash, I'm totally anonymous (until writing this....), I can send email via my PC/Web interface.

Not only that, but tethering my PC to HSDPA shows how fast it really is - and how much of its speed & latency is wasted by the browser's sluggishness on the Nokia.

Plenty going on with femtocells

After a few quiet weeks, it's all happening in femto-ville this week:

Top of the list is the NEC/Ubiquisys rollout for Softbank. The Japanese operator has long been rumoured as the most likely near-term commercial deployment candidate, and is interestingly basing its architecture around IMS rather than the new 3GPP femto standard. Like DoCoMo, (which is planning its own "HomeArea" femto service with Mitsubishi) it is slightly hamstrung by near-term regulatory limits on installing cell-sites, which mandate an engineer be in attendance. However, I suspect that Softbank may be a bit more aggressive than its larger & more conservative peer in getting that law changed, or (perhaps) scrutinising it for loopholes.

One thing I'm less convinced by is Motorola's femto/picture-frame combination. Which looks remarkably like the original Motorola femto/picture frame prototype from 3GSM a couple of years ago that never went anywhere. (Unfortunately I can't find the name or a picture of the thing....). Over at ThinkFemtocell, there's a list of other possible widgets that could incorporate femtos , but my view is that all of these things should actually be static 3G endpoints which connect to a central femto elsewhere in the home. In other words, "phones" specifically designed to work with a femto, a bit like the various WiFi home gadgets like Internet radios and the cute Nabaztag wifi rabbit. I discuss the concept of femto-specific devices in this report.

One of my analyst peers has got quite a lot of press this week with the notion that LTE could be underpinned by a femto layer. I agree entirely - back in May I suggested that "femtos will only become truly massmarket when we get to LTE deployment" and my current forecasts only really forecast a major uptick in shipment numbers above 10m per year, from 2013 onwards. There's a good article from PicoChip on LTE femtos here as well. 

My only slight concern on this is that the vast majority of offloadable mobile data traffic comes from devices that have WiFi in them as well - especially PCs, iPhones and high-end Symbian & Windows smartphones. Collectively, I reckon it's 99%+ on many networks at present - a figure that might fall (and obviously there's voice traffic to offload as well) , but the "inverted pyramid" effect will stay in place for many years.

Lastly, I should mention I'm moderating a panel on Femtos vs WiFi at the NetEvents Press & Analyst summit in Portugal on Thursday - speakers include representatives from Motorola, Airvana, Ubiquisys, Trapeze and GenBand (which just acquired NextPoint)

Huawei WiMAX dongle certified

One of the biggest boosts for HSDPA adoption came when Huawei introduced its E220 USB dongle modem in mid-2006. It enabled operators to have a lower cost device than previous PC cards, in a form that retailers and users intuitively understood. Variants have now shipped 10m+ units, and have massively expanded the market for mobile broadband into the consumer segment.

So the certification last week of a 15Mbit/s Mobile WiMAX variant of the iconic Huawei device is an important milestone. It's intended for the 2.5-2.6GHz band, so must be a contender for a US launch as well as future European & Asian rollouts.

It's not the first USB WiMAX modem - Airspan and Posdata have them as well, and there's even a Xohm-branded one from Chinese rival ZTE. Samsung's got one coming as well.

But given their track record of ramping up volumes, plus their enviable record at signing up distribution channels, the advent of the Huawei device signals to me that WiMAX is due to have one of its periodic upswings in confidence over the next few months.

500MB mobile data a month without rich media

I just looked at the logs for my main mobile data/Internet smartphone. In the last 30 days I've used half a GB - 480MB download, 40MB upload. Probably 70% of that has been over 3G, with the rest on my home WiFi.

And that's without doing anything too fancy - no YouTube, no streaming audio, no VoIPo3G, no P2P. Just normal web browsing (a mix of mobile and "real web" sites), email, RSS, plus a fair amount of Google Maps.

Yes, I'm quite a heavy user - I'll update my RSS feeds while I walk to the tube station & read them on the train, and I'll click through to content-rich web pages that are 500k-1MB each to fully load. And I've downloaded a few PDFs of reports and other documents as well. But I probably have the device in my hand as much as a typical BlackBerry or iPhone owner that I see around London.

And this is with fairly lacklustre 3G/HSDPA performance. I don't think it gets above 1Mbit/s at all, and I'd say its considerably slower most of the time. And on a device which has a UI I'd give 6 or 7 marks out of 10.

My cellphone data usage has gone up roughly 50-fold with the advent of flatrate data, HSDPA and decent browsers.

With an iPhone 3G, or an S-E Xperia or a similar device, with a real 3Mbit/s+ peak speed, I could certainly imagine doing 2GB a month, maybe more if I decided to watch video podcasts on the tube rather than text RSS, or streamed Internet radio. And if I used the phone as a tether for my laptop rather than having a separate dongle, you could probably add another 1-2GB.

OK, I'm probably a top-1% user at the moment. But I'm definitely wondering whether we're going to start seeing some real congestion issues on urban 3G networks sooner than we expect if sales/use of upper-tier devices really kicks in.

How many devices will Apple want to sell you?

The other day I did a post on MIDs, which as with a number of my posts gets syndicated on SeekingAlpha. Leaving aside the usual rants by Apple-ists thinking I'm somehow anti-iPhone (I'm not - I even bought one for my father), the comments have raised an interesting point in my mind.

Some people think that the iPhone is an example of a MID - and that it will absorb the future capabilities of a whole range of other devices.

But other people think that the iPod Touch is an example of a MID - and that it (or perhaps a slightly larger version) will absorb the capabilities.......

So here's a question - will Apple try and structure its product line so people end up buying two devices from the company, or just one? (plus ideally a PC as well). I don't mean the serious fans who'll buy everything - I mean normal customers.

Will the iPhone be the uber-smartphone, sensible to hold to your ear to make calls, but with a good web experience as well.... while the Touch becomes a larger, different beast - perhaps a PSP-replacement, or Netbook-alternative, with ergonomics optimised for these purposes? (and with a bigger battery and screen).

I'm pretty sure that Nokia has long welcomed the trend of people to buy and carry multiple phones - why sell people one product when you could sell them two or three? It will be interesting to see how Apple organises its portfolio around this opportunity.

Is Nokia missing another screen resolution transition?

In 2004-5, Nokia lost significant handset market share. This was attributable in part to its late adoption of clamshell-format handsets.

But at the time, a secondary factor I identified was a slow adoption of the (at the time) leading-edge QVGA 320x240 pixel screens. Nokias high-end phones (9xxx Communicators excepted) were hamstrung by the early versions of Symbian OS, which mandated a 208x176 pixel screen.

This meant a visible difference in screen quality between high-end Nokias and competing devices from SonyEricsson, Samsung, HTC and others. Screen quality is one of the most noticeable indicators of "quality" to consumers after the external aesthetics of the phone.

Nokia tried to move its range to a "double resolution" but non-standard 352x416 pixel display on products like the E60, but has subsequently reverted to QVGA for most of its devices. Even its new N96 flagship is still sporting a 320x240 screen.

Meanwhile, there is a slow but inexorable move towards much larger formats - capable of rendering PC-grade web pages. Some phones in Japan have had VGA (640x480) screens for two years already. A number of high-end Windows Mobile phones from HTC and others have also sported VGA displays for a while.

And more recently, we've had the iPhone (480x320) - and as well as the multitouch capabilities, this generates a significant "wow" factor from users. And the other day I had a quick play with the new SonyEricsson Xperia X1, which has a screen which is physically smaller than the iPhone's, but crams in 800x480 pixels, showing incredibly fine detail. HTC has just announced the "HD" version of its Touch handset with a similar screen. And unsurprisingly, various phones in Japan have been there for a while already.

Now to be fair, the Nokia N800/810 tablet and E90 communicator both already have similar resolutions.... but in terms of consumer-type phones intended for web & media usage.... we're still waiting.

Mobile broadband and notebooks - a challenge to operators' auditors?

Imagine two mobile broadband offers:

£15 a month for 3GB data + free dongle
£35 a month for 3GB data + free laptop with embedded 3G module

The question is: when allocating revenues for these in the operator's financial reports, how much should get allocated to non-SMS data revenues?

And should it be different?

Does some amount of revenue get allocated to non-service equipment sales? (assuming that a margin is made on the wholesale price of the PC after the subsidy has been repaid)

Hypothetically, if I was an operator worried about the current precarious financial markets, I'd want to keep up important headline metrics like ARPU. And if was an auditor or stock analyst I'd want to look very closely at customer acquisition costs for *data* services vs. CACs for traditional phones.

My gut feel is that it's appropriate to allocate the cost of the module or dongle (both well-known, as that's what the IPR royalties relate to) as part of the "service" revenue stream. But it's not good accounting practice to add in the price of the rest of the PC. By analogy, for a fixed/cable operator offering triple-play together with a finance package on a plasma TV, you'd perhaps include the set-top box as part of the service revenue, but not the screen.

It's possible that this is a non-issue and that all operators will follow more conservative accounting approaches with mobile broadband. But I'd be surprised if at least some didn't try to fudge the figures a little.

I'm speaking at the Inst. of Industry Analyst Relations about blogs & social media

I'm speaking at the Institute of Industry Analyst Relations meeting in London tonight.

[Note - for readers outside the analyst "village", AR is a parallel discipline to Investor Relations and Press Relations within many technology companies, intended to deal with people like me. While there's many different sub-types of industry analyst, we all tend to be outspoken, hopefully influential - and occasionally prima-donnas with an over-inflated sense of self importance as well. So basically, many product vendors & service providers use AR professionals to act as a valuable interface point to the rest of the company.... and also as far as possible keep tabs on what we're saying, to whom we're saying it, and what company representatives are saying to us. The IIAR is essentially a trade association of AR practitioners].

Along with a couple of analyst peers, I'll be discussing the role of "social media" in the context of my work.

For me, that's mostly this blog, online b2b social-network 1.0 things like Forum Oxford and my relatively negative views of Twitter, Facebook etc as business tools. I'd also characterise my fairly prolific conference appearances as real-life social medium too.

In particular I will be talking about the implications for AR folk and how it impacts the way they should measure and engage with analysts. It's an area that polarises opinion quite a bit. Some AR professionals think it's relevant and improves relationships, others are more old-school and perceive tools like blogs and social networks as reducing control and visibility. It's a fine balance and one to which there is no single 'right answer'.

I know that as a regular and sometimes contentious analyst-blogger I risk being seen as a 'loose cannon' sometimes. Some seem to pigeonhole me as a blogger who analyses, rather than an analyst who blogs (hint: count the adverts on my site). I get few comments on the blog direct from AR people (unless maybe they're anonymised), although it stimulates a lot of offline email and phone calls. Sometimes I also get contacts from dedicated Blogger Relations teams too, which can be confusing but also gives me a second channel into key companies.

Ideally, I'd write a bit less on the blog and a bit more in paid research and consulting documents. This in turn would be picked up by AR and a virtuous, but mostly invisible, circle of 'influence' would ensue. But ultimately as an independent analyst, I don't have the corporate sales or promotional might of rival analysts from Gartner or Forrester. So to be as 'influential' as my peers, (and from an AR perspective to be seen as such), I need to use less conventional channels to increase my footprint throughout all parts of the mobile ecosystem and its customer base.

I know it's an uneasy balance at times. I'll often comment rapidly on announcements and tie them into ongoing research themes. My broad remit (and the nature of the medium) means that I can't realistically bounce every idea off executives to triple-check - I occasionally make mistakes. But my view is that this is very similar to the role I used to play as an equity analyst - I'd see the newswires announcements at 7am, and have to have some form of initial comment out on the email to investors at 8.15am, backed up, refined (and occasionally demolished) by later, in-depth reports. That said, I'm sensitive to the realities of AR and try not to tread on toes unnecessarily. And unlike some journalist-type bloggers, I know what's totally "non-disclosable" even if I haven't signed an NDA.

I also need to balance the analytic aspects of this blog, and my AR relationships, with the commercial realities of my business. This blog is one of my prime ways to reach existing and new clients - it generates research sales and consulting leads. (hint: Want strategic advice, workshops and forecasts on wireless operator business models, femtocells, mobile broadband or handsets? Congratulations, you're in the right place. Click HERE). I suspect this differs quite a bit from my larger-firm rivals, and also explains why I feel the need to update so regularly.

Overall, it should be interesting to hear from both my colleagues, and understand the AR professionals' view of what I do. Who knows, maybe they'll even convince me to start Twittering....

Financial market turmoil - impact on telecoms

James Enck has an absolutely heroic post on this topic.

As a former investment analyst myself, I've been watching this week's chaos on Wall St and in the Square Mile with a sort of detached horror.

And I've been watching tech & wireless stocks escape with (comparatively) little damage. Who would have thought, 7 years ago, that it would be banks' and insurers' turns in the firing line?

His comments about the gap between network operators' funding vs. necessary near-term capex are particularly interesting to me. Some additional thoughts:

- Now probably isn't a good time to be buying spectrum, especially if it's with the aim to build out entirely new networks. Maybe O2 & T-Mobile have, ironically, done both BT & the UK Treasury a favour, delaying the UK 2.6GHz auction through their legal action? Otherwise that would have been ongoing right now.
- I reckon that some particularly dongle-rich 3G networks are going to start feeling capacity squeezes over the next 12 months - a combination of the need to fire up extra 5MHz channels on existing sites, add new sites, and/or upgrade backhaul capacities. I hope all that's in the current spending plans....
- From a risk management point of view, all the signs are pointing more towards HSPA+ / HSPA Evolved as a cheap & safe interim step before LTE. If I was the CFO, I'd be asking exactly what LTE would add to the bottom line, and what the risks are that positive ROI might be delayed for a couple of years, because of "unexpected issues" like poor availability of devices and applications. I don't think the situation is going to do WiMAX or FTTH many favours in the near term either.
- I've talked about this before, but do mobile operators really want to start subsidising laptops at the current time? Given that a lot of the "Free Laptop!!! £30 a month subscription" deals look suspiciously like consumer credit arrangements, I hope you've checked your underwriters' fine print. And vice versa.
- I'm mindful of a new item I saw recently, saying that chocolate sales go up during recessions, as it becomes an affordable treat, chosen in place of more expensive luxuries like restaurants. There's a lesson there for certain "feel good" mobile services, surely.

Is the MID a market?

MIDs (Mobile Internet Devices) are being pushed by some notebook OEMs and silicon suppliers as the next big convergent handheld category. I've seen some presentations pitching the notion that they'll subsume smartphones, encroach on netbooks and UMPCs, and obsolete standalone GPS devices and media players.

There's no standard definition of a MID but the archetypes tend to be the Nokia N810, a long thin prototype device Intel's been showing, and according to some viewpoints, the iPhone. Most people think MIDs will have a 4-7 inch screen, a grade-A web experience, some form(s) of wireless connectivity, and plenty of memory and battery life. It'll probably have a lightweight OS of some sort (Linux, Android, Apple, some cut-down version of Windows etc), and is likely to have various consumer multimedia capabilities.

Many are quite likely to come bundled with some form of mobile broadband - either a familiar 'vanilla' operator Internet access, or perhaps a more under-the-hood wholesaled connection like an Amazon Kindle on steroids. Although the concept started as being quite WiMax-centric it has now evolved to 3G as well.

So far, so good. Basically a 2008-era PDA paradigm with connectivity built in and all the toys like GPS, right?

Maybe, but I'm less than convinced by the general-purpose MID concept. Anyone with a 6 inch device will have a normal phone as well (quite possibly a small smartphone itself), and their own or a shared family PC too.

Yes I'm sure there will be a few road warriors who want a day-trippable device rather than a laptop. Yes some teenagers with backpacks could have one instead of a PSP. If it's cool enough, the gadget fiends will be queueing outside the shops. But 100's of millions of units? Hmmm.

However.I can certainly buy into the idea that things like PSPs, TomToms, N810s, even iPod Touches could use some form of mobile connectivity. Not all of them, not all the time, maybe just for specific apps rather than general Internet services. Each category will adopt mobile at a different pace, and it will probably vary by country too.

Consistently, divergence has been more important than convergence for the last 10 years. For every two things that have combined, three new ones have emerged in their place - a consequence of Moore's Law and economies of scale.

So I see a small core market of converged general-purpose MIDs, really like mini web-optimised computers, bought as a 3rd or 4th device by urban techno-leaders. But that niche will be surrounded by a cloud of app-specific products, some wireless-enabled pseudo-MIDs, some WiFi only or remaining unconnected. Most of these will support decent browsers and extra secondary software clients too, but won't be optimised for them. Developers will face even more fragmentation than is the case for smartphones, but as the value for most of these products will be hardcoded 'out of the box' it won't matter too much.

For what it's worth, I'm still viewing the iPhone as a smartphone rather than a MID, as it can work as a primary phone and you don't look stupid holding it to your ear. (Its size is on the lower limit of MIDs, though, so I won't argue if you define it differently as part of the category - it's only semantics). Some of the larger WinMob devices, however, clearly fall into MID-land, which means WinMob7 will need to seriously step up the pace if Microsoft wants a role outside enterprise-type mobile products.

Overall - I'm cautiously optimistic about MIDs, but I see it as a fragmented movement towards enabling "mobile broadband computing" in a variety of different devices, rather than the genesis of a new generic computing segment.

Only about 2% of this blog's visitors using mobile devices

I've had Google Analytics installed on my blog for just over a month now, and it's showing up some interesting trends. Some are to be expected - peak usage is on Mon/Tues, most accesses come from US & UK, with a lot from India as well.

Now, obviously this is only the direct web hits I'm measuring - the bulk of readers use some form of RSS or other feed reader. It also misses out sites like Wireless Watch, and SeekingAlpha, that syndicate my posts, as well as assorted other 'unauthorised' aggregation and screen-scraper sites.

But unless someone tells me otherwise, I would have thought that people who access the blog directly on the Web are pretty representative of people who see a link on Google & think it looks interesting. Or who generally just prefer using the browser to read blogs.

So, given that I'm in the mobile industry and so are most of the people reading this, I find it pretty surprising that only about 2% of visits are from browsers on mobile phones. 0.6% from iPhones, 0.3% from Symbian, a handful of Windows users, what looks like about 0.3% of Opera Mini, and a proportion of the 0.4% of non-Firefox Linux users. iPod Touch users were another 0.2%.

And some of the mobile users were in fact me, checking the site looks OK.

Now, I'd hope and imagine that >2% of RSS users are mobile, but overall this really points to the fact that fairly normal, non-demanding web usage is still predominantly PC-based, especially in the B2B domain. It's certainly not a sufficient level for me to seriously devote time to optimising a mobile-specific version of the blog, or even worry whether the occasional graphics I add to the text will render OK on small screens.

A Theory of Convergence?

I've been watching and analysing various types of telecoms and computing "convergence" for many years.

One thing jumps out at me: not all expected types of convergence actually happen - and even more happen *much* more slower than anticipated. But some occur almost overnight.

I'm starting to think about patterns here - what converges successfully? And what remains "unconvergeable"? And how predictable are these?

(Mobile Phone + Digital Camera) is one of the fastest, most successful examples of convergence.
But (Mobile Phone + TV) is one of the slowest - and is arguably an outright failure.

In the enterprise market, IP-PBXs have slowly reached the mainstream, blending voice and data networks. But it's taken 12 years or so, and the convergence process still isn't complete.

Ideally, I'd like to develop a "Theory of Convergence". At the moment, I regularly try to predict what will/won't blend, but it's on a case-by-case basis. So for example, recently I've been assessing whether the PC and phone might converge - and based on numerous specific reasons like battery size, ergonomics and distribution/business model, it seems highly unlikely.

Some of the elements of such a theory would be:

• technical convergence of *most* components of the convergent products, not just the central computing or communications element
• ability to converge distribution channels (this delayed enterprise VoIP for 5+ years)
• ability to converge useability and user-experience
• ability to converge business models / purchase process
  

So for example, converging handsets with small digital cameras has relied not just on the phone's processor, but also convergence of memory on flash RAM, small/lighter lenses for basic digicams, the reduction of the importance of printed photos meaning that camera business model is now "convergeable", and a mostly low-touch distribution model for cameras (ie they can be sold on a standalone basis by unskilled staff in retail or online stores). And basic digicams have small, sub-1000MAH batteries.

Conversely, these factors don't apply to high-end cameras which are essentially "unconvergeable"

As well as batteries, to my mind, PC and phone also have other "unconvergeable" aspects around distribution and business model. In particular, PCs aren't sold with associated "services" and, particularly, are not locked or customised for a given "service provider". Distribution is also a challenge, as are various aspects of user-experience and ergonomics.

PC/dongle (ie PC/module) is a slightly different convergence situation (dongles don't have batteries), but also falls down in the business model / distribution stakes. For example, dongles have legacy SIM cards, and attempts to integrate them with notebooks' connection managers are clunky.

It doesn't mean these types of convergence won't happen at all - but just that the Venn diagram won't overlap that much.

I've still got quite a lot of thinking to do about creating a generic Theory of Convergence - but this theme of "convergeability" is one I'll be coming back to over the next few months.

Tethering...another approach to mobile broadband computing (MBC)

As if there weren't enough options for Mobile Broadband Computing (MBC) - dongles, data cards, embedded 3G, netbooks, MIDs.... - another concept is now starting to make itself felt.

It's "Tethering"

Which is what used to be called, more clunkily, "using your phone as a modem".

This isn't especially new - it's been possible since the days of GPRS at least. On the other hand, it's often been an enthusiasts-only sport, involving various configuration hassles (Bluetooth, non-standard cables and so forth), which while theoretically simple, have been enough to dissuade most potential users.

In the past, this practice has been viewed differently by different mobile operators. Some have actually encouraged it, because their “phone” data plans have been more expensive than their “PC modem” data plans. Others have limited the practice via terms and conditions, especially on uncapped “flatrate” data plans for devices like the Apple iPhone.

More recently, a variety of phones like the H3G Skypephone S2, and the Palm Treo Pro, have been designed specifically with tethering as a key usage case – they even enable the driver software to be loaded onto the PC automatically (rather than needing CD or separate download/install) and ease configuration through wizards and so forth. Some operators see them as a good way of enabling two devices via a single subscription – and help them upsell the user to higher bandwidth-cap tiers or faster speeds.

However, Disruptive Analysis believes that this approach (and other approaches to using ‘clusters’ of linked devices, for example in a “personal area network”) will remain in the minority. In particular, the use of a phone as a tether, attached to a PC on one side and a 3G connection on the other, is likely to exhibit a high degree of battery drain – although it possible that they could also draw energy from the notebook battery if a physical cable is used rather than WiFi or Bluetooth. However, where a cable is used, it makes it awkward to have a simultaneous phone call – and some implementations of 3G chipsets may not even permit two connections to run concurrently.

The price of modems/chipsets should come down to a point at which it’s just simpler to have multiple ones for multiple devices. Where it may become more important is in emerging markets, where a single 3G access could feed multiple users via WiFi.

Nevertheless, this approach gets around the issue of “lock-in” for embedded mobile broadband in notebooks – users tend to upgrade phones, and/or churn providers, once every 18 months or so, which makes it easier for the computing element to benefit from upgraded modems without disassembling the PC to the change the module. On the other hand, implementations of new wireless standards in phones often lag those in modems, especially in mainstream devices.

In a way, tethering could be a partial answer to the issue of session-based MBC without the need for a separate SIM card in a notebook or dongle.

Overall, the tethering approach using a handset is likely to appeal to some Mobile Broadband Computing (MBC) user segments and operators, but it seems unlikely to displace either external modems / dongles, or modules embedded in notebooks or MIDs. Disruptive Analysis is skeptical that tethering will be used by more than perhaps 10% of regular MBC users – although if useability becomes better, it may be used on an ad-hoc basis by a larger number of people with an occasional need for connecting a PC or other device.

How many people use PC in conjunction with their phones?

A few seemingly-unrelated things have occurred over the last few days, but have a common theme:

- I've uploaded some new pictures from my personal voice handset, to email to friends & put on the web
- I've been trying out Vodafone's FaceBook SMS plug-in
- I've had an update from NewBay about its operator-centric phone backup and social network platform
- I've topped up my 3 UK prepay SIM and paid for 2 months' flatrate mobile Internet access

What links them? They all involve using a PC together with a phone. Either locally connected via USB / Bluetooth, or by using a browser to access some sort of operator portal linked to your account.

It's made me wonder what proportion of users ever actually do something similar?

I remember several years ago talking to the guys at Palm, and they reckoned that most of their users synced via the handset cradle, but that that was unusual amongst even smartphone users at the time.

Today, I'd guess the situation is even more polarised with iPhone users linking to iTunes , but a long long tail of low-end handset users never even contemplating using both devices together. I suspect far more people "connect" by swapping out memory cards for sideloading, rather than installing PC-connection software or using operator portals.

Personally, I've stopped putting client software for phones on my PCs, as I switch phones fairly regularly and don't want loads of flaky OEM apps cluttering up my computer and maybe conflicting. I haven't even hooked up the new Palm Treo Pro which I'm playing with at the moment.

But on the other hand, I am using more web-type interfaces to get a window into my "mobile world". But I perceive that being inside the industry, I'm probably unusual in this - I'm not aware that friends or family do the same.

If anyone's got any hard stats on this phenomenon, I'd be fascinated, as done well, I think it adds a lot to both mobile & PC/Internet experience.

Remembering satellite services' role in mobile broadband

Reading this article has made me reflect how little focus gets put on satellite connectivity. Yes, we all benefit from GPS devices, and millions have Sky TV and its international equivalents.

But in terms of either two-way end user devices, or its use for cellsite backhaul, it's often overlooked. And yet services such as Inmarsat's BGAN satellite modems can often be cheaper than some of the more ridiculous cellular data roaming tariffs.

The latency argument in the article is interesting - many expect new radio technologies like LTE and WiMAX to make mobile web access (or VoIP) more useable by reducing the levels of delay. But for those devices connected to base stations beyond the economic reach of fibre or microwave, there's a big extra chunk of latency if there's a roundtrip to a satellite in geostationary orbit (not to mention any extra delay due to compression ). The new MEO network could be a significant factor in enabling real mobile Internet access from remote locations.

3G-embedded vs. 3G-embeddable notebooks

The world and his blog seems to have gone crazy over Dell's cute new Mini 9 netbook, which stacks up against the Asus eeePC and HP 2133 and a range of other upcoming mini-notebooks.

However, there is much disappointment in the US that the device won't necessarily come with 3G built-in, unlike Vodafone's loudly-trumpeted version for Europe.

I think this is something we'll see more of. I've been doing a lot of research lately on 3G-embedded notebooks and mobile broadband, and the simple fact of the matter is that (for now) it makes no economic sense to include a 3G module and antenna in a PC unless it's definitely going to be used for a mobile subscription. Otherwise it's just a costly lump of extra electronics - not ideal for consumer or retailer, and which would hurt the competitiveness of the PC in the marketplace.

At the top end of the market, 3G might start to become a "standard" feature, but if you're trying to sell a $300 low-end netbook with $50+ worth of unused tin inside, it won't make your gross margin look very pretty.

What we'll see is many notebooks coming with the possibility of 3G being embedded for specific markets and specific channels (ie mobile operators who want to sell/subsidise it), just the same as the laptop coming with the possibility of extra memory, or a red casing, or different preloaded OS and applications. (And no prizes for guessing that the slot might support a variety of other comms modules in future - perhaps WiMAX, EVDO, or one of Qualcomm's multi-standard Gobi's).

There might also be some clever ways of OEMs retrospectively paying (or getting a bounty) for the cost of a module if it's activated at some date after purchase, but I can't see that being too common a business model in the near term.

In other words, theoretically 3G-embeddable notebooks will be much more prevalent than actual 3G-embedded ones for the next few years at least.

Full web browsing on a phone? Not yet

The received wisdom in the mobile industry is that continued browser evolution is making handsets capable delivering a "full Internet experience", like that on a broadband-connected PC.

I think that's over-stretching the capabilities. Yes, the iPhone renders a lot of web pages beautifully, but it's still not supporting things like Flash. And that's the absolute pinnacle of web-on-mobile. When you move down to the next tier, you get options like Webkit-on-Series 60. Which is a lot better than a few years ago, and can do a decent job of many websites, but which gets seriously thrown off-course by some of the newer and flashier ones.

My E71 consistently crashes when I try to use my new favourite travel-booking website, Kayak - there's no way I'd trust it all the way through to a transaction. It's got plenty of other annoying niggles as well, such as the switching between multiple windows. Plus it's still dependent on a whole range of network-side issues (and maybe handset software or memory problems too) like click-to-see time, that make web surfing a lot less pleasant and immediate than on a PC.

It'll be interesting to see what Google does with Chrome on mobile. And yes, there's all sorts of widgety goodness and Operatic performances that make the best of the small device.

But at the end of the day, they're all still "best efforts" approximations to the proper web, on a proper computer. I reckon Apple & a couple of others will start to trend asymptotically towards real "web" experience, but that will take a couple of years. And for that sort of performance to reach down to the midrange of handsets and beyond? I'll be surprised if we see a truly first-class browsing experience on a $50 handset (and its typical host network) before the middle of next decade.

Femtos and Asia - news & event

Thanks to Andy & co at ip.access's 3Ginthehome blog for alerting me to this femto development from Japan. If I'm reading it correctly, NTT DoCoMo has been using a Mitsubishi femto since 2007for what appears to be network fill-in purposes, but is now "planning to use Mitsubishi's femtocell small base stations for its "Home Area" service, which links mobile phone networks with the broadband environment.". As far as I know, up until now there hasn't been a residential femto service launched in Japan, not least because of some tricky regulatory issues concerning the need for official inspection at commissioning of new cell sites. I'll be keeping a close eye on this one.

Unfortunately, I won't be able to keep an eye on it at Avren's next Femtocells Asia conference, which will be in Kuala Lumpur on 22-24 Sep, as I'll actually be at another event aimed at press/analysts (where I'm moderating a panel on - you guessed it - femtos).

The KL event has been moved from its original location of Bangkok because of the ongoing political turmoil in Thailand, but it's looking like a good line-up of presentations - not least from DoCoMo's competitor SoftBank. I'm expecting a few other early Asian adopters for femtos as well - most likely in markets like Korea, Taiwan, Singapore and HK, but also possibly Malaysia & Thailand too. 

[Disclosure: Disruptive Analysis is an analyst partner for Avren Events - although that's a reflection of what I think have been some really good conferences they've put on. I'm hoping to be at their next one in Dallas in December.]

Where is the asymptote for smartphone penetration?

Despite the hoopla about the iPhone and the upcoming launch of Google Android-based devices, there seems to have been a sudden bout of queasiness among the smartphone market participants (and their investors). As a share of total phone shipments, there seems to have been a bit of a stall at around 11%.

Various commentators are scratching their heads, and blaming the economy and a range of other factors. In recent months I've noticed an almost religious fervour based on the notion that "soon, all phones will be smartphones"... and now we're seeing the first cracks emerging amongst the faithful.

Much of the recent enthusiasm has come from the US, where there has undoubtedly been a sudden leap in the sales and profile of high-end devices, driven by the iPhone and continued acceptance of BlackBerries among a wider group of users. All operators now have 3G networks up-and-running, there are some sensible data plans, and everyone has WiFi at home and wants to "take their Internet with them". In addition, the various carriers have suddenly started stumbling over each other in an effort to be seen as the most "open".

In many ways, certain user groups in the US seems to have leapfrogged the high-end featurephone phenomenon seen in Europe and elsewhere. There haven't been that many midrange phones with 3MP+ cameras, memory cards, full Bluetooth, decent MP3 capability (with side-loading) and so on. Device have suddenly jumped from being locked-down, carrier-centric products to new ecosystems like Apple's.

This seems to have led some American observers (and some European smartphone enthusiasts) to assume the same trend has been mirrored elsewhere, and that consistent meteoric sales growth and penetration of smartphone OS's was inevitable globally.

In my view, the truth is rather different when taken at a global level, and explains what will probably just be a temporary plateau in smartphone penetration.

What's happened is that the growth statistics in recent years have been driven not so much by people wanting smartphones, as being given them without choice or (often) awareness. I posted in length about this last year. In recent years, Nokia Symbian S60 and NTT DoCoMo shipments have hugely inflated overall smartphone sales numbers, and Nokia and NTT have selected smart OS's largely for their own benefit, rather than because of specific demand by customers for "smartness".

The US market has long been different. Most US consumers buy smartphones because they actually want smart devices, rather than because they want a high-end Nokia, or a contract with DoCoMo. Yes, there have been plenty of "real", conscious, buyers of smartphones in Europe and Asia, but they have been swamped by a flood of other customers who haven't really cared, but have taken what they've been given.

What we're seeing now is this blip in the statistics being worked through. We're no longer seeing huge growth in high-end Nokia shipments (as last week's warning highlighted), and the Japanese market is pretty saturated too. None of the other major manufacturers has really shifted wholesale to smart OS's as a default platform, to carry on the S60-led growth in "under the hood" smartness.

There have also been some other statistical shifts - huge growth in low-end devices in emerging markets, for example. This is what's boosting the overall market to 1.2bn-odd devices, but dropping the average sales price to well below the entry point for smart OS's. And although it's changing, most phones are still not sold with data plans being available (or certainly affordable), especially on the basic prepay tariffs used by 70% of the planet.

My take is that it's going to take several more years to get smartphone penetration to above 15% of global shipments, and that for the forseeable future (lets say 2013) we should probably treat 20% as the asymptote. In developed markets, that figure will be considerably higher - perhaps 50% or above in places like the US or Scandinavia, or maybe 30-40% in markets like the UK, especially if mobile Internet access takes off further. But the notion that someone in Mozambique or Bolivia or Laos,buying a $20 handsets and with a $2 monthly spend, will be buying a 2013-equivalent of an iPhone is unrealistic.

It's very easy to plot hockey-stick charts - but observers from North America shouldn't leap to the assumption that the majority of people follow their specific local habits. Globally, more people play hockey on grass than ice. And most people who buy phones still don't actually care about smarts.

Petition against deep packet inspection

I've got to congratulate the folk at TelecomTV for their principled stance over misuse of DPI. They're running a loud & well-articulated campaign called "Throttle the Package", arguing against some of the more draconian terms in some upcoming European telecoms legislation. 

It's essentially the EU version of the US Net Neutrality debate, with the debate heavily skewed by legacy content providers' lobbying might. I'd recommend that readers sign up at TelecomTV's site to tell the Commission what right-thinking people believe. 

I'm not against the protection of copyrighted materials by legitimate means, but much of what is being proposed is illegitimate, heavy-handed, malevolent and ultimately doomed to failure anyway. And more importantly, it risks some fundamental "collateral damage" to numerous innocent and innovative business models.

I've written at length before about the use and misuse of DPI, and all of its numerous problems and the likely response.

In a way, I take the same view of European DPI issues as I do about US net neutrality - any service provider stupid enough to try something really egregious - like Comcast in the US - will get a suitable competitive, legal and/or technical kicking. But frankly, while I'm confident that in the long term, reasonable compromises will be reached, this really isn't the time to be wasting money and effort fighting the filtering fools.

Vodafone Connect to Friends for FaceBook

Despite what one of the comments on a recent post of mine might suggest, I certainly don't view all MNOs in a negative light. I've been particularly impressed by Vodafone's deepening understanding of Internet services over the past two years - it appears to "get" the Web in a way that many of its peers do not.

The latest thing to catch my eye from Big Red is a FaceBook plug-in application that enables users (not just existing Voda subs, either) to send SMS to their FB friends in the UK. It's on a pilot rollout at present. 

It's really pretty clever, as it allows you to use the UK operators' PayForIt system to charge to your own non-Vodafone mobile account. So as I use O2 for my personal device, but Voda provides my Facebook interface, charges me money for SMS credits, but gets the cash by using O2 as a billing channel. Given the pricing points, both operators make money, O2 retains my billing relationship, but Voda "owns" my Facebook Mobile user experience. 

Clearly, Voda is hoping that the convenience of sending SMS from my PC outweighs the extra cost versus sending them from my phone (I have unused hundreds of SMS in my monthly bundle). The fact that it works when I'm travelling abroad is a very clever option, too, as is the fact that it inserts my real phone number as a "Reply-To" contact.

And it gives Vodafone a fantastic database of people willing to try new Facebook / Social Network mobile comms applications - and presumably a good group to target FaceBook adverts at. 

Mind you, it glitched when I tried to install it first, and it doesn't seem to work on Google's new Chrome browser on my PC, but given it was only released last week, that's fair enough. It doesn't populate its phonebook from your friends' profiles (it doesn't have the permissions) so you have to enter them manually. And the fine-print terms and conditions suggest that "you should check these terms each time you use Connect to Friends" which clearly isn't the quickest user experience if you comply...   I'm also not allowed to link to it without Vodafone's consent, so you'll have to find it online yourselves.

But you know what? I don't mind that it's a bit flaky initially, or that some of the T's and C's are a bit overenthusiastic. It's a Web 2.0 beta on a PC - it's supposed to be flaky, and you're used to signing your rights away when you agree to use FaceBook in the first place. You can feedback comments on bugs and so forth. It's much better than trying to test it to death for a year, then release it & find something unexpected anyway. 

I've also seen an IMS-type FaceBook mashup app demo'd a couple of times, developed by the Fraunhofer FOKUS research centre in Germany. It also hooks in location for a friend-finder type app, an area which seems very buzzy at the moment. There's a demo on YouTube here.

This type of web-centric realism contrasts with the really sniffy looks I got from some telecom old-timers about network applications. When I suggested that the more "purist" IMS-RCS (Rich Client) would be an ideal Web 2.0 plug-in. I got the distinct impression that they thought that FaceBook should plug into IMS via a web service gateway, rather than the other way around. Well, maybe when there's an IMS community of 100m+ regular loyal users, it'll be the planet and FB will be the orbiting satellite, but at the moment the gravitational dynamics are the other way around.

Apparently, the application has been developed "in conjunction with the Value Creation Centre, a Vodafone joint innovation initiative with IBM, and the Vodafone Internet Services division at Vodafone UK" . That's pretty telling to me - if huge organisations like IBM and Voda can be this web2.0-centric, there's no excuses for any of the other old-world telcos to ignore new Mobile/Web realities either. 

Combined with other initiatives like the Dell netbook launch last week, it's also another pointer that the mobile industry is finally coming to terms with the PC remaining the central pillar of most developed-world Internet access for the foreseeable future.

There should be a SIM-free option for LTE

In terms of general market perception, LTE has been stealing much of WiMAX's thunder of late. Now obviously we've seen the ascendance of LTE coinciding with one of the periodic nadirs of enthusiasm around WiMAX, and we should start to see the balance redress somewhat when Xohm launches and other smaller deployments also start to come online of the next 4 months.

My current expectation is that by 2012/2013, we will see LTE user numbers for mobile broadband approach or even exceed those for Mobile WiMAX. And both sets of figures will be dwarfed by HSPA and HSPA+ figures.

But I can't help thinking that 3GPP risks leaving a certain amount of money (and usage) on the table for WiMAX. In particular, temporary, session-based access.

There are two ways to buy communications services - through an account (postpaid or prepaid) as seen with most of today's mobile and fixed services. Or on a one-off basis, as seen with WiFi hotspots, hotel broadband & TV, Internet cafes, pay-per-use web conferencing, even payphones of the past. In this circumstance, the customer uses a local service provider to fulfil a particular need.

Accounts make most sense when linked to a particular identity - especially a phone number. This means it is in the user's interest to establish a long-term relationship with a service provider.

Taken as a % of the overall telecoms market, the one-off payment proportion is tiny compared to subscriptions or SIM-based prepay mobile. But as a proportion of wireless/"on-the-move" data, it is rather higher.

The problem with offering one-off payments for mobile data in 3GPP technologies is the need for a SIM - or a roaming relationship with a home "account" supplier's SIM, who will probably charge a healthy (or often unhealthy) premium.

WiMAX (like WiFi, and indeed CDMA EVDO) does not require a physical token like a SIM card. You could flip open a laptop and (assuming it supports the right frequency bands), choose a temporary provider on an ad-hoc basis.

Besides the roaming approach, the only way I can think to do this with a SIM-less 3GPP device would be to have some sort of software-type SIM, or use a secondary SIM-based device such as a phone as an authentication token. Neither are particularly ideal solutions in my view.

Now, I don't think that this is the "killer application" for WiMAX which will defend it against the onslaught from HSPA and LTE. But it is a gap in the market for mobile connectivity services that is currently served poorly by the current cellular approach.

O2 offer of iPhone on prepay looks like a smart move

Regular readers will know that I'm a big fan of cellular business models in which upfront device purchase cost subsidises the service element.

I've also also talked about the challenges of getting prepaid subscribers to use data plans, something which has developed very quickly over the past 6 months.

And, lastly, I pondered about how the Apple's iPhone commission model would fit with prepay-centric markets.

Taking all this together, at first glance I'm pretty impressed with the way in which O2 UK is pitching its own new iPhone prepay offer. It's pricing 8/16GB variants at £350/400 inclusive of a year's data connectivity. (It's £10 a month)

It will be interesting to see how many people decide to use the prepay iPhone as a data-only mobile Internet tablet device & MP3 player, while keeping a separate phone for voice & SMS. If you treat the mobile data as being worth £10 a month, it's like paying £230/£280 for an iPod with full mobile web & email access. Compared with, say, a Nokia N810 or even some of the lower-end mini laptops (which probably would be used as a "second computer") that makes a lot of sense. 

In my view, it puts the iPhone centre stage in the new "MID" (Mobile Internet Device) category, albeit without some niceties like Flash support.  (Edit - Hmmm, I wonder if this is connected to the recent interesting in Flash Audio-based VoIP?)

(Yes, I know that many people will use it as a phone as well, but personally I like having a separate data device/contract that doesn't have an associated 'public' phone number I tell anyone, so I'm free to switch when something better comes along, with having to think about number porting)

The £10 a month for data in months 13 and onwards seems a bit pricy, compared to the £5 that 3 offers for a month's data on prepay SIMs in the UK (I have one), but I guess that predicted iPhone data usage is probably much 2x the average that 3 is expecting from its prepay data users. What's not unclear is whether the iPhone can be unlocked after Month 12, and a different operator's SIM installed.

Plenty of unanswered questions about the O2 NFC Oystercard trial

In general, my view of NFC is that it's interesting, but that it's far from certain to be a success. In some ways, it's a solution looking for a problem. It's not an obvious failure like some suggested future technologies, but neither is it a clear win. It will be more important than DVB-H or other mobile TV standards, but a bit less than WiFi and way lower down the pecking order than Bluetooth, memory cards, GPS and a range of other emerging handset features.

There has been much press ballyhoo about the results of an O2 / Nokia trial in London, which included subsuming the standalone "Oyster" contactless travel payment card used on buses and tubes, plus also the "PayWave" electronic-wallet function.

In some ways, I can buy the Oyster argument - in fact, it doesn't even need integration into the phone, a simple RFID sticker on the back or in the battery compartment could do the trick just as well. I see no advantage to the involvement of an operator in the value chain at all - apart from specifying NFC chips in future handsets. I don't buy the idea that Londoners would want out travel expenses appearing on the same bill as postpaid phone usage, and prepay users would almost certainly prefer two different "buckets" for transport and communications for ease of budgeting. Also, many postpaid mobile phone users are prepay Oyster users (me, for example).

But to be honest, nobody I know views the Oyster card as an encumbrance. It's simple, and it works - although most of us were extorted into using them in the first place by our former tyrannical mayor's policies on fare differentials. Integrating it with the phone gives one less thing to carry, yes - but also raises issues like whether the NFC would work when the phone was out of battery, or whether the thousands of people who lose phones on the Tube/Bus would then have the extra indignity of getting fined for riding without a ticket at their destinations. Roaming users would find it a pain, transferring balances to new phones/operators would be a pain, it would need to work seamlessly while phone was in mid-call, and so forth.

I see no upside, but it could probably work OK up to a point. Yes, the "survey" showed positive results but I've yet to see a clear breakdown of the triallists (are the representative of "average" Londoners or mobile users?), and the questions mentioned in the press release all seem a bit woolly. (I'd love to have a copy of the full questionnaire & results and see what didn't make it into the press release).

"87% said that availability of the service would be likely to influence their purchase of a new mobile phone" - I suspect that this was probably one of those questions that had options like "A critical influence", "A major influence", "A little influence" and "No influence". (And possibly "A negative influence"). And adding everything except "No influence" could yield 87%, but then with nothing to compare it to, it's meaningless - adding a flashlight or toaster function to a phone might yield the same result. It screams "PR soundbite" to me.

The mobile wallet function I am a lot more negative about. I've heard this mantra of "the mobile phone can subsume your wallet" more times than I can remember. I've been to numerous conferences & briefings, all of which are conspicuous by their failure to mention benefits to the end user. I can see why operators might want this (a slice of lots of payments), I can see why phone manufacturers might want it (more stuff to put into phones to keep up ASPs), and I can certainly just about guess why the army of consultants and integrators might be in favour.

Anecdotal feedback I've had from some PayWave merchants in London has been that it's been a dismal failure. Nobody uses it. It's not used for the card-based version that's being pitched, let alone the phone-based option. Transport for London has let BarclayCard/Visa have free sales pitches in various stations to tout the thing, and the people manning them have looked like the loneliest people I've ever seen.

Then there's a huge raft of issues about usability, security, theft, loss, financial regulations, ability to churn and so forth. It's a complete non-starter - even the Japanese use of Felica has been lacklustre. I don't buy the hype about this type of m-payment at all - although I have no problems with other forms like SMS-based for certain groups (eg people without bank accounts).

I think there is more chance of your wallet subsuming your phone, than your phone subsuming your wallet. NFC might eventually get some applications that are useful. Advertising hoardings and other information services - yes, although probably quite niche. Electronic keys (for doors etc) - hmm, maybe, but not anything really critical (imagine "divorcing" your operator when you churn, and having to change the locks....). Travelcard - possibly, as long as the operators can be distintermediated.

But payments and credit cards? I'm going to stick my neck out and say no.

NFC? Meh.

Cool and uncool in telecoms (lighthearted post...)

It is just possible that some readers of this blog are unaware of Top Gear, the car show which is without a shadow of a doubt is the world's best TV programme. Exported to countless other countries, it even has local versions being made in Germany, Australia and elsewhere, although these suffer without the wisdom of the demigod Jeremy Clarkson.

One of the most clever features on the show is the "Cool Wall", a section which plots cars on a scale of coolness. This isn't necessarily linked to performance, looks or other normal criteria. It is a separate axis which transcends these measurable characteristics and instead looks at the less tangible aspects of image and desirability. In particular, vehicles that are bought by people trying to be cool are always ironically deemed to be seriously uncool. By comparison, cars that are bought by people because they are sort-of interesting, unusual, functional or eccentric might be thought of as "sub-zero"

So for example, Audis are becoming seriously uncool, as they're being bought by people who used to want BMWs. Toyota Pious's [sic] are obviously in the same category for being sanctimonious. Conversely, Land Rovers are honest and unpretentious, and are therefore quite cool. Aston Martins are "sub-zero" unless you live in Chelsea in London, in which case you're just following the herd and you need to get a life. And so on. It's very clever, really.

So a hat-tip is definitely in order to James Enck, who has pointed me towards a telecoms industry extension of the concept.

I pretty much agree with the telecoms assessment, although I'd make a few amendments and additions.

(And before I get a zillion comments about analysts being uncool, please note that my tongue is firmly in my cheek with this. And iPhone fanboys - the lack of a sense of humour is by far the least cool thing you can imagine).

From the UK perspective, I'd probably boost 3 and Vodafone up a notch from "seriously uncool", as they're doing some fun stuff with Internet partners and so on. Orange probably deserves its own "so uncool it's almost cool again" category for its general old-school attitude to centralised control and things like IMS. The Apple iPhone (a very cool device on its own) is brought down firmly into "seriously uncool" as it's been bought & fawned over by so many geeky fanboys and wannabee-cool dullard hedge funders. Conversely, quite a lot of SonyEriccson phones are cool, which is probably why the company is on a warning from Sony over its recent lacklustre performance. All 3G dongles are cool as they aren't over-branded by the manufacturers.

The counter-intuitiveness and rampant subjectivity of the Cool Wall is its best feature.....

What else? Skype is now just merely "cool", as are most wireless VoIP companies, although geekiness is in danger of dragging them down. FaceBook is popular but useful, and doesn't have a "religious" aura, whereas Twitter is unbelievably try-hard and therefore uncool. UK regulator Ofcom is pretty cool, as is the FCC with its fairly principled stance on things like Net Neutrality.

I'd say that Qualcomm and Ericsson both qualify as "cool", as neither seems to mind an image of short-term conservatism as long as they reap long-term rewards. Nokia is a bit of a confusing mix - its hundreds of millions of self-effacing low-end phones like the 1100 are way cooler than its glossy-but-shouty E- and N-Series.

Any other suggestions?

Welcome back James Enck

The first blog that really made me sit up & take notice of how useful this whole blogging thing can be was James Enck's EuroTelcoBlog. At the time, he was doing equity research for Daiwa, covering European telecoms stocks, and writing pithy, controversial snippets that cut through a lot of the industry's B.S.

Unfortunately, he had to drop writing it a year or so back when a new employer's compliance department had less of a sense of humour than Daiwa's.

For reasons that James explains, that adventure has now ended. So he's now back in the blogosphere, and I for one would heartily recommend everyone here reads EuroTelcoBlog religiously.

Inflight cellular and inbound calls

In general, I feel that the world could happily live without inflight cellular, but I accept that with current fuel prices, airlines need every revenue stream possible if they're not to go the way of Zoom. And I absolutely no problem with the principle of availability of inflight WiFi for PCs, or SMS for phones.

But the big potential problem I see with inflight cellular is not outbound calls/SMS/VoIP - it's the inbound traffic & associated ringtones.

2-hour shorthaul flights are pretty noisy anyway, as they're in daytime, with announcements and catering trolleys and kids going on holiday. And I'd certainly imagine that few people would be so socially-unaware as to initiate noisy conversations during the hours of "darkness" on a longhaul flight. But the issue has to be sleeping in a cabin with 200 other people who could well be *receiving* SMS or calls from people (or computers) in their home timezone, unaware they're midflight.

If you work out the stats it starts to be quite scary - on a full 747 you can have about 100 people within earshot, say 10 rows around you - so I'd guess that during a 5-hour "lights off" part of an overnight flight, you could easily get 30+ inbound events with associated ringtones.

One option might be to make the inflight system accept outbound traffic only, during "dark" hours. Or just instigate a rule that all phones/PCs have to be used with headphones only, and that any ringtone offenders get moved to a spare "sin bin" seat next to the lavatories.