Pages

Pages

Thursday, July 31, 2008

Reader survey: thoughts on Twitter?

I'm pretty skeptical of the value of Twitter.

Nobody has been able to construct an obvious reason why I should donate my time to get my head around it, play with it, update it, follow other people and so forth. When I've expressed this view at events and other online forums, the usual response has been "Oh, just start using it, and you'll understand the value over time".

I'm sorry, but that doesn't work for me. I don't like playing around with new social media tools "just for the fun of it", and my primary client-base isn't in the social media industry.

(Edit - the other argument I've heard is "You'll be able to have lots more conversations with lots more people". Which also doesn't work, as I already have more options for conversations than I have time. I unfortunately have to turn away briefings, conference speaking invitations, delay responding to emails and voicemails and deprioritise other interactions. I'd need to know that Twitter wouldn't just add to the problem, and that I'd be conversing with the "right" people more - ie the ones with unique and relevant knowledge, money to spend, or a good sense of humour).

I need a clear & immediate "win" if I'm going to take time away from other work or personal activities. There's a million bits of cool software, social media, or networking technology & events that I could invest time in, so what's special about Twitter? Would say, 4 hours "invested" in Twitter yield more revenue & value than 4 hours spent phoning or emailing some old contacts or clients, whom I haven't had a chance to catch up with in ages? Or 4 hours updating my website? Or 4 hours at something like Mobile Monday?

I'll contrast that with this blog, which I write because it yields obvious benefits in terms of visibility, interactivity with a knowledgeable audience, direct revenue and leads from new clients ("I read your blog regularly, can you also do XXX type of project or YYY type of event?"), and the ability to put down and "claim" ideas I have, when I don't have time to write up full reports.

On the other hand, Jonny Bentwood has highlighted just how many other analysts are Twittering away happily, so maybe it's time for a rethink.

(Socially, I'm totally disinterested in it - none of my non-industry friends have ever even mentioned it. Absolutely zero).

Now, I know that (obviously) quite a few people in the apps/Web part of my audience are avid Twitterers - but I have a variety of other avenues to reach them. Any comments on how you deal with analysts via Twitter would be interesting, nevertheless. But I'm more interested in people from the "techier" parts of this blog's readership - MNOs' radio and architecture and strategy teams, wireless infrastructure suppliers, device manufacturers, semiconductor vendors, regulators and so on. Do you already follow anyone "tweeting" on femtocells, or spectrum policy, or deep packet inspection?

Comments - either on this thread, or via email at FIRSTNAME.LASTNAME AT disruptive-analysis.com are very welcome. I'm willing to change my opinion on Twitter, but it needs to have a clear and demonstrable *business* return on my time. And the response "you won't understand until you've tried it" is unacceptable, as that generic argument could be applied to everything from religious evangelism to a new type of cheese.

Wednesday, July 30, 2008

Should device purchase subsidise connectivity, not vice versa?

Whether or not PCs come with 3G modems embedded (see the debate here) there seems to be a general belief that the way forward is to offer "free laptops" as long as someone signs up for a contract for either fixed or mobile broadband.

While I can see the appeal of this type of subsidy, I'm wondering if it is the wrong way around.

Given that millions of people seem pretty happy to stump up cash (or at least credit cards) for PCs at the moment, why try and dilute value in the industry by persuading them to get computers for nothing instead?

Why not try a different approach - have a $1000 (or $400) laptop, that comes with "connectivity for 2 years" included free in the price, or as a bolt-on upgrade option like extra memory? Lots of physical electronic products come bundled with some service element (eg a warranty) that eventually expires but could be renewed. Just like a warranty, you could have terms and conditions, excusions and policies.

The problem here for the MNOs is that they would no longer "own the customer". But so what? Customers are certainly not owned by the warranty company working with the dealer of their new cars either... but warranty companies are (generally) highly profitable.

This approach - "embedded connectivity" - and the reverse subsidy model of hardware purchase including "free service" isn't new in technology markets either. Telematics products like car tracking systems often come with a year's service included in the upfront price. Software may include a year's maintenance or upgrades.

This also gets around the thorny issue of monthly billing, as well as limiting the credit risk the mobile industry might expose itself to if it started to subsidise a good proportion of the world's roughly $100bn worth of annual laptop sales.

Tuesday, July 29, 2008

Why the obsession with monthly billing for mobile services?

I've just been reading this post from SMS Text News about remote-wipe capabilities for handsets. This has been something that's been discussed for years, especially in the context of enterprise-issued devices like BlackBerries or other smartphones.

But while I agree in principle with the idea, I don't agree with the notion that this would best be provided as a service for (say) two pounds per user per month.

Why? Because I perceive this as a sort of insurance policy. And insurance to me is not something that is "worth" a monthly fee, because on a monthly basis nothing is done by the provider. I have an annual car insurance policy, and annual home insurance policy, and (because I travel frequently) an annual travel insurance policy. Each year the premium and the fine-print conditions are renegotiated, and I expect to be able to shop around between competing providers.

... and if and when I actually need to make a claim, I accept that there may be an extra "per event" charge (the excess) involved. And in the case of car insurance, a whole range of extra factors are taken into account to determine my risk level (garage, where I live, job, traffic offences etc) and price the premium accordingly. And for each year that I don't claim, I get a discount on the next year's premium (which is transferrable to a competing provider too).

Now, I understand why mobile operators are focused on monthly billables. They are measured on monthly ARPU by their management and investors. And typical mobile contracts are of variable lengths, 12/18/24 months, so an annual insurance / remote-wipe policy doesn't fit so well.

But some of the work I've been doing recently for a billing & OSS client (Highdeal) has made me think about the lessons that could be learnt, and practices adopted, across industries. In this instance, I think the mobile industry could learn quite a bit about charging for insurance policies. Annual fees make sense, as does some form of recognition of relative risk for individual users. The mix of fixed and per-event fees makes the customer perceive value - even if they don't have their phone stolen and wiped.

It should also be possible to have 3rd-parties providing the insurance, to ensure competition in terms of both price and features. The operator could even act as "insurance broker" in this regard, representing many different device-management providers, and still take a commission and conduct the billing. This would also act as an advantage for those handsets sold through the operator's own channel, as they could be pre-configured correctly. But equally, it needs to be (easily) possible for me to select a totally independent provider (or self-manage the devices if I'm an enterprise) if that makes more sense.

There are plenty of other examples of mobile services for which the current bi-polar choice of "per event" or "per month" are both poor fits with user preferences. To be fair, we are moving to a few "per day" or "per week" options for things like data plans, and some promotional tariffs and linked prices. But there's still a heavy monthly billing-cycle cloud overhanging even innovative pricing methods.

I had an interesting recent example of this: I renewed my main personal mobile contract with O2 via Carphone Warehouse, which had actually been running for a year longer than the original 12-month term. I negotiated an upgraded handset, and a different tariff plan, including a new flatrate data allowance. But could it start immediately? No, not even for a longstanding customer. I had to wait until the next month for it to kick in, before which data was still charged at the prior ridiculous levels. So naturally when I got my shiny new phone, I switched off the 3G to stop myself accidentally racking up huge charges, and made sure not to use the new operator services on it until next month's cycle comes into force. Now, I'm pretty sure that most users' behaviour with a new phone's capabilities is set in concrete by their first week's experience with it. Counterproductive for operators wanting uptake of new services, eh?

So.... I think in general there needs to be much greater flexibility in how mobile services are priced and billed, rather than just using the current lazy catch-all monthly rating mechanisms. There needs to be flexibility on matters like annual billing. And some things need to take effect immediately, not sometime next month.

Latency in mobile - more needs to be done

When most people refer to latency in mobile networks, they are talking about the inherent lag in traffic going through the network, which includes aspects like buffering of traffic in the various nodes like base stations and controllers, as well as any processing in the network core and the handset itself.

There will always be a trade-off between latency and spectral efficiency, because the best compression algorithms have to wait for enough inbound data to arrive before processing it and crunching it down in size. Error correction also plays a part here.

All this is well-known, and the subject of focus in the industry. Recent generations of technology like HSPA have improved latency considerably.

However, there is a second class of latency which often gets overlooked. It is initial call-setup time, or initial data connection time. I'm old enough to remember when the PSTN network was digitalised - and how miraculous it seemed that a call recipient's phone started ringing as soon as you pressed the last digit on your own handset. Yet in mobile, we're still stuck with the type-in-numbers-then-hit-send mentality. We still have to wait for 5, 10, even 20 seconds before the call connects.

And it's the same for data too. What's called the "wake-from-idle" time on the current generation of HSDPA phones and modems is frequently dreadful (although this is also down to how the network is configured). The "idle mode" is used to conserve power, but it's a trade-off against good user experience. The first time I use the browser on a phone during a day (or worse, after I emerge from a tube station) it takes an age to connect, then display a downloaded page. Same thing after the network drops into idle mode during occasional use - it's a noticeable lag compared with a PC. Both connection time to WiFi/broadband, and response time when you go to a web page, are much faster in most circumstances.

HSPA Evolved (aka HSPA+) should help with the data wake-from-idle. But in the meantime, referring to HSPA as always-on is a misnomer. I also suspect that the whole idle/active state thing probably fits very poorly with always-on apps like many background services, presence, AJAX web pages and the like. As usual, there's been a disconnect between the network side of mobile innovation and application/user-behaviour aspects.

Yes, it's only a few seconds, but it's this type of thing that will lead to niggling customer dissatisfaction with mobile broadband compared to fixed connections. If you're used to instantaneous DSL or cable - or better still, ethernet in a school or workplace - then that extra bit of initial latency could have a disproportionate effect on overall quality of experience.

On the voice call set-up time, I'm not sure what the answer is, because I suspect that much of the problem is in the telephony application and signalling, rather than the network. It strikes me that this could well be a killer differentiator for VoIP in mobile - either operator-based or independent. It seems strange now, but I remember well just how different and better phone calls became with instant connection. Never mind price, or fancy voice mashups - maybe wVoIP companies should think about a really obvious and tangible benefit to end users - not wasting 10 seconds of your life every time you make a call.

"Instantaneous" has a lot of value. The mobile industry should focus on it more.

Wednesday, July 23, 2008

Mobile widgets - who wants them?

Maybe I'm missing the point, but I'm starting to think that the fascination that the mobile apps industry has with "widgets" (small Internet-connected applications) is misplaced. I'll happily confess that I haven't looked at the sector in huge depth yet, but on first sight there seem to be several weak logical connections in the whole premise. The argument seems to be go something like:

1) Phones are small with small screens
2) Therefore small applications are good, especially if they only take up a small amount of screen space
3) Phones don't have much memory, but are getting better browsers & faster/cheaper data connections
4) Therefore using some sort of central widget "engine" and UI framework makes it easier to have lots of widgets on one phone rather than lots of native applications
5) We want to sell lots of mobile advertising in future
6) Therefore having some sort of closed widget environment (but based on reasonably open standards) with built-in advertising functions gives us a great channel to the mobile user
7) Most "normal" people don't like installing native apps on phones. And developers don't like the fragmentation of platforms.
8) Therefore make a 'library' of mini-applications to enable people to search for and "discover" widgets and download them, working across lots of handsets.

But frankly, based on recent experiences, I can't see where the customer demand lies. I can see why the industry would like widgets to be adopted. But I fail to see why end users are going to be bothered.

In particular, there's a whole range of issues that are unsolved:

1) I still see no evidence that "normal" people want to "discover" stuff to download to a handset. Yes I know that umpty-million people are using iPhone app store, but they're not "normal" in a 3-billion-mobile-users sense of the word.
2) Always-on widgets are battery-killers. And given that different phones (and different networks) have different optimal methods for power-management, the whole abstraction / single-platform notion falls flat on its face. I could fry an egg on my Nokia E71 if I leave some widget stuff running in the background.
3) Based on my (admittedly limited) experience playing around with things like Widsets and Yahoo Go!, widgets are slow, clunky, and make a big song-and-dance (hey! rotating icons! wooo!) about doing basic stuff that could work perfectly well on a normal browser web page.
4) Browsers are getting *much* better. Many of the current widget tasks can be adequately performed with a decent set of bookmarks and RSS functionality, or using server-side aggregation of web services onto a single home screen.
5) Is there actually any evidence that widgets are popular on PCs? I can't remember seeing any newspaper articles about how cool they are. They're certainly not in what I'd call "the popular consciousness" in the same way that (say) Skype is, or even mobile broadband / 3G dongles. Vista's low uptake rate won't have helped either.
6) Wearing my "end user" hat, widgets appear to offer the same stuff that could be done in the browser, but with less flexibility (how do I delete that Yahoo! Entertainment default widget?) but with greater amounts of intrusive advertising. Hmm, let me think about this.
7) Widgets are at the mercy of the handset network connection manager, in dealing with intermittent coverage, use of 2G vs 3G vs WiFi, policy about data access while roaming and so on.
8) Typically there's lousy integration of widgets with the native functions of the phone - especially the dialler, the phonebook & SMS. This might get fixed with initiatives like OMTP's BONDI in the medium term.

As I said upfront, maybe I'm missing the point. I haven't really followed this area in great depth, but that also means I haven't absorbed the hype & groupthink. Or maybe this whole mobile widget thing is just another self-delusional ruse, dreamt up by the mobile apps industry to convince itself that massmarket end-users are actually interested in putting 101 bits of random software on the phones.

Tuesday, July 22, 2008

Femto provisioning - what else is needed?

Interesting to see the Femto Forum look to standardise on a protocol called TR-069 for remote provisioning and management of femtos last week. This doesn't come as a huge surprise to me on a number of grounds, especially as the same approach is used for management of many existing broadband home gateways and modems.

A discussion I had in the past with a remote-management vendor highlighted the possibility of being able to "daisy chain" a number of devices using TR-069. Given that most femtos over the next few years will be standalone, and need to be plugged into the ethernet port of a separate router or gateway (possibly provided by a separate service provider), that is a very important consideration.

It also raises the possibility of outsourced management/provisioning. For example, I have a BT Home Hub gateway attached to my BT ADSL line - given that I'm unlikely to churn to a triple/quadplay provider just to get a femto, I could quite imagine BT offering wholesale "pass-through" device management services to Vodafone or O2 or another licenced-spectrum operator.

However, one aspect is left unclear - and is perhaps for "future development". I believe that it will be important to manage (and reconfigure / download in some cases) software on the handset to enable it to work better with the femto. This is especially the case if the operator wants to enable any new "connected home" services on the phone. While such changes could be done "over the air" or using approaches like OMA DM, it may also make sense to tie these in with the femto provisioning itself.

Monday, July 21, 2008

Embedded 3G laptop hype

OK, hands up which of you have got a laptop with an embedded 3G modem?

Right, now you can put your hands down again. Yes, both of you.

There are plenty of mobile technology areas that are getting hype at the moment, but the one which seems to represent the greatest triumph of marketing over reality is that of laptops with built-in 3G modules. It's probably the area where an onslaught of PR by Ericsson, the GSMA and assorted mobile operators has been most successful in skewing the industry's view of supposed inevitability. And of course, it's an amusing shot across the bows of the WiMAX fleet, which is only expecting a handful of Intel "Echo Peak" embedded PCs at the end of the year in the US only.

The laptop manufacturers have been partly complicit in this, but only because they're being bribed with the promise of subsidies, or possible cold hard cash in terms of "activation bounties" when customers sign up.

I've written extensively before about why the Embedded Emperor has no clothes. It's also something that regularly crops up in client meetings with operators and others. When I asked to explain why there's such as gap between promise and reality, I typically get as far as saying something like "....and seventeenthly...." before the penny drops.

One of the biggest issues is consumer ignorance and apathy. Compared with 3G dongles, embedded laptops don't even register on the radar screens of Mr Average Customer (especially in the UK, which is awash with dongle advertising). My perception is that the whole idea of "connectivity on a USB stick" is a nice easy concept for retailer and customer alike, especially given the ubiquity of USB memory these days.

A dongle is simply "broadband on a stick", with a cute & memorable name.

Just in case I was missing something though, I thought I'd do a quick check on the current retail marketplace.

First stop, Dell's UK website. Nothing obvious on the main consumer laptop page - no mention of mobile in the bullet points for the key Inspiron & other brands. No mobile filter in the "narrow your selection" filter. But right at the bottom of the page is a Vodafone "partner" link (as well as two Tiscali fixed-broadband ones). That goes through to a dedicated Voda page with 18 and 24 month contracts described. And a sidebar with "To see the full range of Dell products with built-in mobile broadband, click here". Which doesn't have a live link, but just a note to the webmaster saying "needs link". Clicking through to a couple of the PCs, I could add in the (Voda-only) 3.6Mbit/s HSDPA card for £89 (personal user) or £120 (business user, with some fiddly rebate offer). Alternatively, I could go to the Vodafone site & get the same dataplans, but get a dongle for free - including a 7.2Mbit/s stick if you take a 24-month contract.

So, £90 for embedded, or £zero for a faster external dongle. Same Voda data plans. Hmm, tricky decision, that.

Next stop, Carphone Warehouse. The "Home of Free Laptops" - subsidies all round! And about 6 or 7 different sorts of PC, on all sorts of fixed & mobile broadband contracts, including Orange, 3 and T-Mobile. Toshiba, Acer, Fujitsu-Siemens. All of them based on dongles.

And it's Dongle City on T-Mobile's UK website as well. Not an embedded PC to be seen - until I clicked down through 3 layers of the site and eventually found Internet-Ready Laptops on a sidebar. The page looks like it hasn't been updated for a year, citing 1.8MBit/s speed and roaming available in "18 countries in July 2007". The Toshiba PC on the page was introduced in Nov 2006.

Even on T-Mobile's German website I couldn't find a laptop - and they've probably been Europe's most enthusiastic operator about the whole concept. Although it has got a very cool dongle with HSUPA, HSDPA and DVB-T digital terrestrial TV as well. That's really innovative and certainly differentiated from the embedded 3G modules... guess I'll be getting to "eighteenthly" next time I'm discussing this in a meeting.....

Thoughts on smartphone usage - Nokia E71

This is a personal post, based on my own recent experiences, as I've been playing around with a Nokia E71 for the last month or so. (Disclosure: the device was a freebie given away at a press conference, but the timing was spot-on as I needed a new data-type phone anyway).

As regular readers of this blog will know, I'm not hugely enthusiastic about smartphones as a personal user of mobile devices. I'm not especially interested in "discovering" things, I don't think that "content" is particularly important, and I'm certainly not bothered about wasting my life finding & downloading software to use on my phone. That said, from time to time I need to use various apps from a work perspective - trying out VoIP clients and a few other bits & pieces, for example. I'm also trying to force myself to use some of the more Web 2.0 things, but frankly it's still 95% professional interest to see if they work, rather than actually personally wanting them.

On the other hand, I do browse the web & email on mobile devices quite a lot, so having a convenient & fast mobile Internet device of some description is definitely a must, at least during "working hours". And I do mean browse, not "find" or do "local" stuff - I want to occasionally use Google, check up on a couple of web forums, download & read my normal RSS feeds, visit restaurant websites and so on. But when I'm "off duty" at evenings/weekends, having something with a decent camera, phonebook & SMS user experience is much more important than Internet access.

So for various reasons, I tend to prefer having 2 devices - one phone-type thing (I also have a new SonyEricsson C902) which is my main voice+personal device, and one web-type thing for times when I might actually need mobile Internet access & am not sitting with my laptop.

So - the E71. Physically, it looks & feels like a quality device - slim, quite weighty, metal rather than plastic, decent form factor with a QWERTY but not too brick-like. I'd say it's much cooler than most Blackberries & Windows devices and comments from other "business gadget" users tend to concur. Conversely, most "normobs" still look at it as a bland corporate device, at least in the UK, where any mobile with a QWERTY marks you out as a middle-aged boring worker-bee drone.

One side-effect of the metal casing is that it gets noticeably warm when using apps that use the radio a lot - which gives a good tactile warning of impending battery exhaustion.

Some thoughts on various of the features, apps (and the OS) that are native on the device or that I've subsequently installed:

- Browser (Webkit) is good, if a little slow at times (I think this is the software but could also be the network, especially the wake-up time from idle mode on HSDPA). Renders most normal web pages quite well, plus some Ajax-type things too. It's certainly useable for "the real web", although it's frustrating enough at times that I certainly wouldn't want it as my primary means to get online. At home, I'll still tend to fire up my laptop instead.
- Email client - not great. Doesn't handle HTML emails or attachments in a particularly intuitive way, and it's a bit annoying the way it displays headers & the like. That said, my experience is rather stifled by the difficulty I've had in setting up outbound email via 3's network - I can't get them to send via with a disruptive-analysis.com email address because of some POP3 config weirdness.
- Connection Manager - pretty awful. I can't seem to get it to have complex default rules - in particular I'd like to set it to "Use my home WiFi if it's available, otherwise use the 3 mobile APN". I suspect something like a DeviceScape or other 3rd-party CM client would fix this, but I haven't wanted to take the time to download & try and set it up yet.
- WiFi - This has improved immensely compared with my previous experiences of dual-mode phones. Very very quick to identify WiFi APs and authenticate, and doesn't seem to chew up battery anywhere near as much as it used to a year or so back. This goes to show what optimisations can do when handset & chipset companies put their minds to it. Let's hope the same happens with Femtocell-optimised phones, although I still seem to be the only person shouting about that particular issue!
- Widsets (Nokia's Widget platform). I'm using this as a sort of RSS reader at the moment. It's OK, but a bit frustrating. It's obviously intended as some sort of future advertising platform, but that gets in the way of the stuff I want to read. It's also very slow to download stuff, even over HSDPA (text appears in chunks slower than I can read them, so I find myself waiting), doesn't seem to include embedded links in the blogs I typically use, and turns the phone in to a portable toaster if I leave it running in the background. I have to say that I'm really not a great believer on widgets on either PC or phone, and this does nothing to dispell the view. If I wasn't such as lazy user of devices I'd go & "discover" a better way to get RSS feeds, but I can't be bothered & will just put up with the pain for now.
- Voice & SMS. Don't know, haven't used them at all
- Camera - pretty dreadful to be honest, despite being billed as a 3MP device. I suspect I may have a pre-release version of the software, because it does some really weird things to colours, especially in dim light. I know that Nokia has some excellent cameras on its N-series phones, so I'm guessing this is an anomaly.
- Fring & Truphone. Unfortunately, neither of these had E71-optimised VoIP clients when I looked a couple of weeks ago. I'll check back at another point.
- Yahoo Go! Almost exactly a year ago, I gave the big Y! a hard kicking over its mobile capabilities. The good news is that the current version of Y! Go is a bit better. The bad news is that it's still pretty awful. There's this awful rotating-carousel of icons/services/widgets, most of which are useless & non-deletable (how can I get rid of the Sports & Entertainment widgets? And I don't want Flickr. And I can't add Y! Messenger or My Yahoo). At least the email inbox actually refreshes itself, unlike my experience from a year ago. Although reading a mail on the device doesn't change the read/unread status on my PC. Overall it's far too much of an attempt to be an all-round on device portal, which I don't want or need.
- Nokia Maps - OK but a bit fiddly, and keeps trying to sell me add on services. Scrolls too slowly too - needs some work on the user interface. So instead, I use...
- Google Maps - I'm fast moving to the view that G Maps is the one application I'd be prepared to swallow my pride about downloads, and put on any device, work or personal, as long as I had a flatrate data plan. It just works, even without GPS switched on. It gives me a great map of where I am by using Cell ID - I don't need turn by turn directions, just an electronic equivalent of an A-Z atlas. (Note: I'm almost always walking when I need directions, not driving - I don't need SatNav in the car given my travel patterns).

One question which I'm sure will crop up is "Why don't you use an Apple iPhone or a BlackBerry instead?". The main reason is this: The Nokia is a "vanilla" device. It's not provided with an operator subscription at all, nor any customisations or lock-downs. While this has proved a bit of a negative because of my email hassles, it's completely outweighed by the fact that I'm free to choose whichever service provider & method of billing I like.

In my case, I'm using a prepay SIM from 3, which enables me to get flatrate data for £5 per month. I don't use the thing for voice or SMS, so that's all I'm paying. $10 / Euro7 per month for full HSDPA mobile web browsing & the ability to mess around with assorted other applications. No blocking of VoIP or anything else, just a pure pipe. Bargain. None of the other UK operators (I think) permit a similar deal - certainly not on an iPhone or BlackBerry.

Overall - it's very useful as a mobile Web device, but I'm still left cold by most mobile applications, with the honourable exception of Google Maps. And for a QWERTY device, it's aesthetically pleasing. Yes, an iPhone would be better, but it's certainly not worth the extra cash & carrier lock-in. For the growing % of users who want to disaggregate device purchase from access subscription, Nokia "gets it" more than most of its peers.

That said, I'm still not sure I'd really want to pay the £355 that it retails at - I'd much rather lose some of the "smartness" and just have a simpler £100-150 data-only tablet-type device made up of HSPA chipset + screen + browser + battery. There's so much unnecessary software in the thing that must add to the bill of materials & time/cost of testing and integration.

Friday, July 18, 2008

Double-counting and half-counting mobile broadband

I'm up into my neck in a spreadsheet model of mobile broadband adoption & device usage at the moment.

An interesting paradox is currently bending my mind:

- some mobile broadband subscriptions will be "double-counted". One subscription might cover multiple devices, particularly for non-SIM technologies like Mobile WiMAX. So one individual user might have a single WiMAX "account", but have 2 or 3 separate devices permitted to use it (eg laptop + personal media player + in-car system) . A similar scenario might be people who SIM-swap between multiple devices, or share a 3G USB modem or mobile-backhauled local router.

- some other mobile broadband subscriptions will be "half-counted". One device might have access to multiple network accounts or technologies. So a laptop might have embedded WiMAX, plus an HSPA modem. Or a single operator may operate multiple network technologies, and use 802.21 in the connection software, to attach the device to whichever is available at a given location.

The easy option is to design the model to ignore these effects, and assume they're just minor or will cancel each other out. But the current roughly 30% mismatch between cellular user & subscriber data suggests that's an oversimplification too far.

Edit: The more I think about this, the more complex it gets. One of the other issues is going to be the possibility to get "ad hoc" access to WiMAX, and maybe even HSPA/LTE in the future, in the same sort of way that you can get one-off access to WiFi hotspots. If I go on a trip to 5 countries with a mobile-enabled laptop, and sign up for 2 days' WiMAX with local providers in each place - or buy a local prepay data SIM for HSPA - does that count as 5 "subscriptions"?

One thing's for sure... we're all going to have to be very wary of reported statistics on mobile broadband. I've just realised I've already got 4 subs myself - a 3 HSDPA dongle, a 3 prepay SIM in a smartphone, a new HSDPA featurephone with flatrate data on O2, and a T-Mobile 3G SIM which is actually in a 2G device.

Edit 2: No, I've got 5 subs. I forgot I'd been given another temporary week-long HSDPA modem & SIM in February by the GSMA in Barcelona. Although that one may no longer be counted as "active" depending on Telefonica's definition of churn.... See what I mean about the stats?

Mobile Wallet? Whose money is it?

My brother works for a large investment bank & spends a lot of time thinking about the finer details of various financial instruments.

He raised an interesting issue, specifically about London's Oyster card (stored-value card for Tube & Bus travel), but I'm wondering if it applies generically to mobile payments as well.

It's a simple question which is surprisingly difficult to answer: Does the money on a stored-value card belong to the user, or the provider? If Transport for London (in the Oyster example) or a mobile operator (for a mobile wallet or even prepay credit) goes bust, can you get the money back? Or are you at the back of a long list of other creditors? Related to this is what happens if the phone or card breaks.

Given the huge % of the world that uses prepay accounts for mobile - and presumably given that operators will want to avoid credit risk for postpaid users with mobile wallets in future - this is far from an academic question. I've seen some estimates recently that suggest $10's of billions may get spent annually on mobile payments with NFC.

At that level, it's certainly relevant to ask who it belongs to, and what happens when things go wrong. Let's face it, given recent events it seems that normal banks aren't as safe as we used to think - so would you trust a mobile operator to give you better protection for your cash deposits?

(Incidentally - my brother tried to get an answer about Oyster credit, but eventually gave up. He couldn't even find someone who understood the question).

Great article on Web vs. Native mobile apps

I'd heartily recommend reading this post by David Wood (of Symbian, so not exactly unbiased, but very even-handed on this topic). It looks at the very hot topic of whether mobile app development will avoid the headaches of fragmentation and operator lock-downs, and just migrate to the Web and cool AJAX-type tools instead.

My view is similar to David's - for many applications, Mobile Web will be the way to go, for ease of development, cross-platform support, rapid update and so on.

But for some the most important and demanding applications, there will still be a need for native development, even if it comes with a dose of pain.

The same is true on the PC - most things work fine in the browser. But for Skype or anti-virus or serious corporate applications, there's advantages to being able to write the most efficient, powerful, complex code to the native platform if you're capable of doing so.

I'm sure the balance between outright optimisation and convenience will shift over time, and across different application categories and networks. But there isn't "one answer", and developers and service providers will need to work out ways to blend the two paradigms.

A mischievous suggestion - how about a WiMAX/UMB hybrid?

I see that Sprint has left the NGMN (Next Gen Mobile Networks) consortium in something of a huff, after that organisation endorsed LTE as "the answer" for 4G, leaving WiMAX on the table despite scrutinising it for a long time.

To be honest, NGMN's "inclusive" policy of other networks always struck me as a thinly veiled stick with which to prod Ericsson, NSN et al into a more benign IPR and interoperability direction.

So, in the spirit of end-of-the week "What if......?" thinking, I'm speculating about a possible next step. NGMN's other jilted lover is Qualcomm's CDMA-UMB, which is now looking almost completely friendless.

So.... how about a WiMAX / LTE hybrid? Could Intel and Qualcomm bury the hatchet and play nicely together?

What if... Sprint got together with KDDI in Japan (also a CDMA / WiMAX operator) , and locked the 2 silicon companies in a room together with representatives from CDG and the WiMAX Forum, and told them to turn UMB into WiMAX's FDD profile? And make it all neatly backward-compatible & roam-able with EVDO? It might even tempt the Koreans back to the CDMA camp.....

Yes, I know it's probably not practical on technological, commercial or political grounds. But it's fun to speculate anyway.......

Thursday, July 17, 2008

HSUPA stalling?

I've been looking at some of the data on HSUPA deployments, and frankly it's a bit disappointing. There was a news article yesterday quoting Qualcomm saying "This is the year for HSUPA" and "it's now being included as standard for chipsets aimed at 'mid range' devices and by the end of the year will appear as standard on low end phones as well".

To me, that last comment rang alarm bells - it sounds like a marketing pitch "Hey, everyone! HSUPA's the next must-have! Get involved! No, really, everyone's gonna be doing it! Honest....."

Normally that sort of pitch is made when something isn't happening. "Standard on low end phones"...er, well given that WCDMA still isn't standard on low-end devices, I'd be pretty surprised if the world's $30 handsets suddenly skipped straight over HSDPA and went all the way to full HSUPA from GSM.

In fact, there's still a real dearth of HSUPA phones. Although there are 60-odd devices in total, most are modems, dongles, modules & other non-phone products. There's a handful of Toshiba phones in Japan, and some high-end Windows devices from HTC, I-Mate and a couple of other vendors. There are no high-end Nokias with high-speed uplink yet (including the new N96), the iPhone is HSDPA not HSUPA and so on.

So clearly, there's some level of desperate hyperbole.... but the question is how much? Looking at the GSA's statistics for deployments and "commitments", it seems that the wheels have come off the "Uplink" bandwagon a bit. While 51 HSUPA networks have launched, there's only another 17 known to be "committed". Yes, the number of live networks has almost doubled in 6 months, but the future rollout schedule hasn't kept pace. Looking back at some statistics for HSDPA from 2006, when it reached 50 deployments, there was a pipeline of 60+ on the way.

Now don't get me wrong, I'm all in favour of HSUPA. It's one of the things that makes VoIPo3G much more practical. But I suspect that it's that, coupled with the risks of P2P traffic or people running PC's as web servers over the mobile network, that are leading operators to avoid specifying UPA as a must-have. They'd rather have faster generations of HSDPA, or support for different frequency bands like 900MHz. Despite some of the blather about social networking, there isn't really a killer revenue-generating app for HSUPA services outside of a few road warriors and photographers needing to upload big image & presentation files from PCs. I'm not convinced operators really want to host billions of uncompressed 5MP cameraphone images on their networks, even if the radio network can support it.

I have a suspicion that most operators will probably wait for HSPA+ / HSPA-Evolved before upgrading their uplink, particular for handset users. There will be a slow creep of HSUPA into 3G dongle modems for those markets where shipments have exploded, and you'll start see smartphones gradually coming into the market at end-2008, but you can forget about truly massmarket HSUPA phones until late 2009 or early 2010.

Tuesday, July 15, 2008

BT's Fibre announcement - initial thoughts on impact on mobile

BT has this morning announced a £1.5bn ($3bn) planned rollout of fibre in the UK, principally in urban areas. It is targetting "as many as 10m homes" by 2012 (out of a total of about 25m households, so 40%). The plan is for a mix of fibre running to street-side cabinets (FTTC), plus some fibre-to-the-premise for new-build homes and I'd imagine also some businesses or apartment blocks (FTTP).

There's going to be a lot of discussion about the general proposition and some of the regulatory impact elsewhere, so I thought I'd focus & have a few initial thoughts on the possible impact on the UK mobile industry.

  • Fibre announcements like this put the claims of a narrowing gap between fixed vs. mobile broadband into context. The figures in BT's announcement suggest a range of initial speeds from40Mbit/s for FTTC and 100Mbit/s for FTTP, with roadmaps to 60Mbit/s and 1000Mbit/s respectively. Unlike the "headline" numbers for HSPA and LTE, these are not "shared" capacity or the theoretical speed achieved by a sole user in a cell sitting next to the base station.
  • Related to this point, BT also talks about upgrading its other ADSL customers to ADSL2+. In reference to the current discussion about real vs. advertised speeds, it says that the "majority of customers should enjoy speeds of 10Mbit/s or above".
  • Coupled with moves by competitors in local loop unbundling, and the UK cable operators, this suggests to me that real-world sustainable average broadband speeds for fixed services should remain 2-5x that of mobile broadband. Yes, there will be specific locations & services where it's occasionally possible to get 10MBit/s+ via a dongle with HSPA or WiMAX or early LTE rollouts, but they'll be offset by other areas of weaker signal strength, congestion from density of users, & limited backhaul capacity.
  • On the topic of backhaul, although BT doesn't mention it, presumably they'll be putting fibre to (or near) almost every cell site in the country in the expectation of continued growth in demand for wholesale capacity (or outright dark fibre rental).
  • Depending on what happens in the UK's delayed 2.6GHz auction, it seems reasonable to assume that any future BT deployment of WiMAX will also be supported by fibre.
  • The statement this morning has some interesting comments about the regulatory changes that BT would hope to see from Ofcom. Although much of its comment initially seems to be aimed at getting the cable & other fixed NGN networks opened up for wholesale, the line "BT’s firm belief is that all next generation networks in the UK should be open as this approach will boost competition and consumers and businesses will benefit." suggests to me that it has its sights set on mandatory wireless wholesale / MVNO access too. It will be extremely interesting to see if this ends up going in the direction of structural separation (NetCo vs ServiceCo) for mobile operators.
  • ... on the other hand, if open mobile networks are mandated in the UK, will it make sense for some/all of the MNOs to club together and build a big shared network, rather than competing overlapping ones? That has some interesting implications for spectrum policy too
  • FTTH & ADSL2+ should make it possible to run LTE femtos at decent speeds in (some) UK households. Should be interesting to see whether this catalyses a view that UK deployments of LTE could be based initially around a large in-home radio build out.
One last thought - I was at a debate last week during which it was suggested that the appropriate number for a UK-wide FTTH deployment would be about 10x this level of investment - the figure of £15bn was thrown around. To me, this compromise looks like a fairly prudent risk-management approach on BT's part given the current economic situation. Sufficiently aggressive to catalyse some regulatory change - and maintain clear water between fixed & mobile broadband - but not large enough to act as a boat anchor on the company as the UK economic falters.

Friday, July 11, 2008

The upside-down pyramid of mobile broadband

I've been seeing an awful lot of PR spin recently about the numbers of supposed "mobile broadband" subscribers, and how this will likely "overtake" fixed broadband.

While I'm a big believer in 3G dongles, and also in using devices like high-end smartphones (Nokia, iPhone et al) with decent browsers & flatrate data, there's a huge amount of hype here.

A key thing to remember if you're scrutinising numbers is this:

- Pretty much anyone with a 3G modem for a PC & a data subscription will actually use it regularly (unless it's embedded) & probably generate a huge amount of traffic
- Pretty much anyone with fixed broadband will use it very regularly, and so will 1-3 other members of their family too.
- Quite a lot of people with an HSPA or EVDO Rev A handset will never use data at all, or will use it for very occasional downloads or web browsing. Plenty of Nokia N95 or Samsung Soul owners bought the device because it's got a 5MP camera. (I'd expect most iPhone 3G users to be pretty active with their data connection though).

The idea that there is 100m+ active users of "mobile broadband" in any meaningful sense is doubtful. The fact that someone can download the 50kb front page of their operator's WAP portal at 7.2Mbit/s is not exactly what "broadband" means to most casual observers.

I'm currently working on a model of mobile data users & devices & traffic & applications, but as a rule of thumb some averages might be

PC+dongle = 1GB / month
iPhone-type device with decent data plan =100MB / month
Blackberry-type device or smartphone with limited data plan = 10MB / month
Everything else = 1MB / month or less

In other words, there is what Boris Johnson might refer to as an "inverted pyramid of piffle" when it comes to discussion about mobile broadband. A few % of the users generate a huge % of the traffic, while a large chunk of supposed users (ie people with suitable phones & networks) generate none at all. This will change only slowly, as PC-based mobile broadband is still early in its growth cycle, while 3G is being pushed into handsets of people who still don't care about anything more than voice & SMS.

Bottom line - you'll be seeing a lot more press hyperbole about the numbers of "mobile broadband" users. They might be fruit, but they're not all Apples. But don't compare them with Oranges. (Puns intended, sorry, it's Friday afternoon).

Monday, July 07, 2008

Nokia Series 40 6th Edition - featurephone market direction

In my focus on femtocells and spectrum recently, I haven't had as much time to talk about handset software as I would normally. (I know that I'm "owing" a comment on the Symbian Foundation thing as well).

However, another announcement did catch my eye, as it's a bit less glamorous - but arguably more important. Although it's Nokia's high-end devices that get all the limelight, it's the more-basic Series 40 featurephone platform which dominates sales volumes - and which pretty much sets the scene for all low-to-midrange phones globally. Given that it accounts for 30%+ of all worldwide phone shipments - and probably 40%+ or 50%+ in certain markets - any developments in the platform pretty much determine the viability of future next-gen business models.

It's very easy to get hung up on Feature X or Application Y that is supported by S60 / Windows / Android / iPhone / whatever.... but if it ain't in S40, it ain't going to be that important in the really big, wide world.

There's a good general blog article on changes in S40 6th edition here, and the official page for developers is on Forum Nokia here

But I want to dig into this more deeply. Reading through the specs, one thing leaps out at me:

Nokia is turning S40 into a mobile web platform.

There has been a certain amount of speculation that Nokia would push Symbian and S60 down further into the featurephone space, perhaps even getting rid of S40. That's clearly nonsense - S40 has to sell at price points right down to the bottom end of the GSM market, and up to some pretty decently-featured higher-end 3G devices. Symbian won't scale down that low. S40 is still a proprietary OS, which doesn't allow developers to write "native" code.

Other rumours that Nokia is going to revamp S40 with an underlying Linux kernel are obviously wrong too - at least for another year or two.

Critically, there's also no Java multi-tasking capability. S40 remains a resolutely single-task platform. This has a number of side-effects - many of the "converged" concepts of mobile applications implicitly need to run background applications or have multiple 'live" services or windows. IMS really needs multitasking phones to be usable. WiFi in phones really needs multitasking to be useful for anything apart from basic UMA connectivity. Presence, push email, background downloads and a variety of other capabilities all benefit from proper multitasking, even if it possible to kludge workarounds.

There's no JSR281 Java API for IMS in the basic platform, nor JSR180 "Naked SIP". Although in the FAQ it mentions that specific phones might support extra APIs on a case-by-case basis, it singles out NFC support rather than other examples.

(Given that the S40 FAQ only mentions WiFi in the context of UMA, I think we can be fairly confident that notions of putting WiFi "into every handset" are total nonsense. I'd be surprised if more than 5% of S40 phones ever ship with WiFi).

On the other hand, there's lots of Web stuff in S40:

  • Webkit-based browser capable of rendering "the real Internet"
  • Support for Adobe Flash Lite 3.0
  • Suport for Javascript & Ajax
  • Developer support for Nokia's Widsets widget platform applies to S40 too

This fits with a lot of the painful truths I've been telling my network infrastructure advisory clients recently - web protocols like HTTP and AJAX are going to create more value for mobile than communications protocols like SIP. Given that architectures like IMS seem to assume that session-based SIP services are the be-all-and-end-all dominant future revenue generators, this poses some very awkward questions indeed. It also makes Ovi look much more important.

On the other hand, all this web goodness in S40 will be totally wasted without decent data plans - especially for the mass of global prepay users which I suspect account for the bulk of S40 phone buyers.

Thursday, July 03, 2008

A thought: HSPA/LTE in unlicenced spectrum? Ideal for femtos?

There's about a zillion frequency bands for 3GPP technologies these days - 700MHz, 850, 900, 1800, 2100, 2600, AWS, various Japanese ones, upcoming UHF bands and so on.

But there's some notable absences - unlicenced 2.4GHz and 5GHz bands. The ones that WiFi and Bluetooth and a million other gadgets use.

Sure, its unmanaged, congested, interference-prone.... but for short ranges, who cares? Like, for example, if you had a 2.4GHz HSPA femto at home which you could "roam" onto? (I'm assuming you wouldn't have your own SIM card & HLR).

Why would you want to do this? Different business models - lots of them. And a complete end to the need to put WiFi into phones as well, if you could have a single multi-band chip that supported private cellular as well as operator cellular.

Yes, I know it would have some horrible problems, and I certainly couldn't see the stuffy standards bodies daring to support something so lacking in QoS and control and operator involvement. But if cellular ever wants to stand a chance of competing in homes or enterprise networks, it needs to have an "owned" profile as well as a "service" profile.

Wednesday, July 02, 2008

Mobile Web 2.0 and Power Management

I've been playing around with some Web 2.0 services (Widsets, Yahoo Go etc) on a Nokia E71, and one thing I've noticed is that they kill the battery, as they keep going online looking for new stuff to pull down. Given that the device has a huge 1500MAh battery, the fact it can be depleted in a day (and gets pretty warm to the touch too) indicates to me that we've got a problem here.

Anyone else experienced this? Obviously the simple answer is to turn off "push" and "background apps" but that sort of misses the point.

The other option is to have something clever in the network, like BlackBerry or the new notifications server bit in the Apple system, which wakes up the phone when necessary, without needing it to keep "polling" the network.

I've also heard stories that power issues can completely derail the "mobile presence" model - particularly if you have a "live" address book with hundreds of contacts.

I haven't had a chance to look in detail at yesterday's announcement by OMTP of their BONDI platform for Web 2.0 on phones (I have a meeting with them scheduled for next week), but hopefully there's something in there which could help solve the problem.