driving churn....

...overheard a quick comment from a fellow conference goer today, which underscores some of my issues with the opportunity for operator-controlled IMS services:

"Well, obviously for my work life I use my company email address at ___.com ... but personally, I've been a Yahoo! Mail user for years - I even pay for it - and if my carrier tried to block access to it from my mobile, I'd obviously move to another provider"

Amusingly ___.com is a very large European cellular equipment vendor, that is currently pitching operator IMS solutions very loudly indeed....

killing time?

I'm in Korea for a conference at the moment, and spent a few days in Seoul before coming down to the island of Jeju for the event. I've been keeping an eye on the way the Koreans buy and use their cellphones and other bits of technology.

One of the things that I've noticed is just how much "personal technology" there is around. Not just cellphones, but also huge numbers of MP3 players (but few iPods), and also "proper" digital cameras. It's also telling that this is a very low-crime society, so nobody worries about the risks of wearing a few hundred dollars-worth of gizmos on a lanyard around their necks or in their handbags.

Also, as is well-documented elsewhere, there are hundreds of PC gaming / Internet shops known as PC-Bangs (as well as DVD-Bangs and other "outside of home" pay-per-hour technology locations that would normally be covered by home consumer electronics usage in Europe or North America)

Koreans use their phones a lot. Sit (or more likely stand) on the Seoul subway, and there is a large percentage of people pecking away at keys (mostly games or SMS), and a substantially smaller number actually talking (there's decent cellular coverage in the metro). Interestingly, there are far fewer people actually talking on their phones on the streets & other locations as well.

This has got me thinking about the way some of the much-vaunted new applications like mobile-TV and gaming are being pitched in Europe. Quite often, I've heard terms like "ways of killing time", "info-snacking", "mobisodes" and so on. The idea being that when people have a couple of spare minutes, they could use these new mobile apps rather than perhaps read a newspaper.

But in the UK certainly, and probably elsewhere in "chatty" countries with few taboos about talking on the phone in public (Italy, for example), I'm wondering if larger bundles of voice minutes & better in-building coverage would just make people use more voice instead. If you've got a few minutes free - why not call your friends or family? why bother trying to download music onto your phone (and paying through the nose for it), when you could have a good gossip instead "for free"?

xMax.... there's something there, but I can't be sure exactly what....

Met up with xG Technology the other day - those guys with the seemingly-unbelievable wireless broadband technology that they reckon is orders of magnitude more efficient (power, range etc) than WiMAX or other emerging technologies.

Unlike some journalists, I haven't had a demo, but I got to chat with their execs in some detail. They answered lot of my questions about latency, indoor performance and so on, but it seems like a lot of this will be up to exactly how the radio is implemented by any future licencees. There's nothing obviously untoward in the basic technology itself which would act as a limiting factor.

Apparently it works by modulating a signal onto a single wave cycle, rather than the 100s / 1000s of cycles more commonly used in other types of radio. (The long-buried physicist in me does wonder if this might induce any odd quantum effects, by trying to interpret a single wave, rather than averaging out a property measured over 100s - anyone else out there have a view on what Heisenberg & co might have to say about all this?)

More interesting was the commercial status & focus of the company. Essentially, they seem to be fending off various approaches from all and sundry in the telecoms industry, from chipmakers to equipment suppliers and carriers. Apparently, they don't need cash - they've been in stealth mode for 5 years & are quite happy on their own.

They are involved in an interesting game of chess - they want to create enough interest in their technology that someone agrees to licence it... but they don't want to give enough short-term concrete proof points that other companies try and use 1000s of engineers to reverse-engineer it themselves, emulate the idea and then adopt a "Go, on, try & sue us, how many lawyers have you got?" strategy.

Basically, their position is "we reckon we've got something cool, we want to create some buzz & speculation & excitement.... but we're quite happy if everyone else rubbishes it for a while longer, as it gives us a chance to navigate the mountains of unrealistic NDAs and liability legalese that prospective licencees are trying to foist on us".

It still sounds interesting, though.... and I guess there's half a chance some chipset manufacturer (Intel? Qualcomm?) might take a punt....

Mobile TV: good idea..... but will it work?

There seems to be a groundswell of naysayers around Mobile TV. I've been at a couple of conferences recently where the mantra of "why would anyone pay for it?" seems to have been chanted.

Is it a technology in search of a market?

I actually think these observers may have it the wrong way round. For me, it's one of the few recent inventions in the mobile business that actually seems to make sense. People know what TV is, and many of them pay for it already, so they understand intuitively about mobile TV.

The marketing message is pretty simple "It's telly. But on your phone" . Job done.

Contrast that with, say, the MMS proposition, which is trying to train users to do something utterly new:

"Yes! You can take lousy low-quality images, compress the living daylights out of them further, put them in a cumbersomely-constructed message, and spend lots of money to send them to someone else, who might occasionally receive them OK. No! Don't just upload the image to a PC and email it for free more easily instead"

Or Push-to-Talk:

"Yes! It's Voice-over-IP! But you pay more for it, not less! And, er, don't ask us about the latency"

And I'm still unconvinced by the whole download-music-over-the-air thing as well, but at least the idea is relatively easy to understand, even if the pricing is silly. (But at least it's only a factor of 2-3x silly, so that's not too bad, when you consider that international mobile data roaming is often priced 10x-1000x silly)

So, for once, I'm a believer. Mobile TV makes sense, conceptually. People will "get it". I don't know if I'd use it personally (I don't pay for cable or satellite TV at home), but I can understand the millions who might want football / soaps / news / "adult" content on their 2.5-inch screen.

One fly in the ointment, though.... (leaving aside the usual theme of handset price & user experience) is in-building penetration. I did a quick scan of some studies about DVB-H and the general consensus seems to be that if you want decent indoor coverage, you'll need transmitters on cell sites as well as existing TV transmission towers. Which changes the cost of deployment very significantly. I haven't had a chance to find out if Qualcomm's Media-FLO and the satellite DMB approaches have the same shortcomings yet.

Bottom line in this case, for me at least, seems to the reverse of the usual:

Is it a market in search of a technology?

Service, application or feature?

I'm doing quite a bit of research on IMS at the moment. Apart from the head-spinning array of acronyms, one of the things that jumps out at me is how differently the IP and mobile worlds treat "things that people can do with their phone & the network".

There seem to be three main schools of thought:

- "It's a service" - which means "we, your carrier, will install a bunch of complex kit, maybe customise your end device, and bill you for using this thing every time/month/per-byte/etc. We might also try & charge you for using someone else's service, and bill you on their behalf."

- "It's an application" - which means "we, your carrier, will try and sell you some software that makes this thing work, and maybe even host it for you. However, you will 'own' it, which means that any problems are yours to solve"

- "It's a feature" - "we, your carrier, are annoyed that Microsoft / Symbian / someone else has bundled this in with the software of the phone, and that it connects to some server outside our control, relegating us to the role of 'bit pipe', without any chance to earn extra revenues, even if we're adding no appreciable extra value".

There's also the fourth school of thought, which is "It's bait for out community/advertisers/OS, and if it's not good enough, we'll keep adding extra bits until it is". Which translates as "we, Google / Microsoft / Yahoo / eBay will gladly give you a whole bunch of stuff for free, because we know it's actually pretty cheap to provide, even though you value it quite highly"

It strikes me that one of the things that IMS will enable is much simpler service vs. application vs. feature arbitrage. Of course, this isn't the IMS intention at all, which is purely about services - but to make IMS work, it seems likely that phones (and networks) will need such a lot of upgrade in terms of "smartness" and performance, that the feature & application stuff will probably sneak in regardless.

It will only get worse.....

Vodafone's gloomy outlook today does not really surprise me. What surprises me is that the City doesn't appear to have priced in what seems apparent to everyone I meet: that margins in the cellular business are going to face continuing pressure, despite the promise of whiz-bang new services and networks.

Voda is blaming its future difficulties on high levels of mobile penetration (exactly why this is surprising is beyond me, it's hardly as if its suddenly happened), and falling termination rates. Its continuing problems in Japan aren't helping in the short-term either.

Which is interesting, as it tells me that the likely additional pain from IP-derived pressure on roaming, plus price erosion for indoor and fixed "nomadic" use of mobile phones is not yet priced in.

Basically, customers (and competitors / substitutes) are waking up to the fact that the "mobile premium" on voice pricing is only acceptable when you're actually mobile - ie moving around. If you're stationary, sitting at home in or in the office, all that cool cell-to-cell handover technology in the network has zero value to you. And, increasingly, there will be ways to avoid paying for it in those cases.

At the moment, the IP-based options to play arbitrage games (like dual-mode phones) are still clunky, but they're evolving in various guises over the next few years. Other cellular-only options will also give lower-priced indoor phone calls. Voda itself has a "HomeZone" type service in Germany, offering low-cost calls ostensibly to substitute for fixed-line voice, but also to compete with O2's successful Genion product.

This is why conferences like last week's Wireless VoIP event are so telling - it's the cellular carriers that have the defensive presentations, talking about blocking VoIP, or pricing data traffic to mitigate the risk of Skype. It's also why I think the more aggressive fixed/mobile hybrid operators will be the ones to benefit from new network architectures like IMS.

I'm chairing this event on next-generation networks tonight, so if anyone's around, say hello.

International roaming and VoIP

Does anyone know what proportion of carriers' international roaming calls is accounted for by long, outbound calls? It strikes me that it's probably quite high - business people phoning into conference calls, ringing back to head office or to clients - or even tourists phoning home.

I think this section is uniquely vulnerable to PC-based VoIP substitution, unlike inbound calls transferred to the user, quick outbound calls for voicemail and so on.

At a conference the other day, a substantial section of the audience (admittedly attending a Mobile VoIP event) had been using Skype or some other form of VoIP over in-room broadband or hotel WiFi whilst they had been there.

I did myself - I had to have a 30min+ conference call with someone in the US, while I was myself roaming in Belgium. Given a choice between my mobile, the hotel phone, or the (relatively painful) €20 WiFi charge + pennies on SkypeOut, it was no contest.

Add in the fact that the timing of the call is often known in advance, the user (especially a business traveller) will likely have a PC anyway, the desirability of having web access to Google or a client's website while on the phone, and the ergonomics of a headset, and it's a bit of a no-brainer.

What's not clear to me is just how much of a typical carrier's voice roaming falls into this category - but I suspect it's a pretty sizeable chunk.

Wireless VoIP

I'm currently at a conference on the technology, applications and commercial issues around Wireless VoIP. I've been speaking myself (on the topic of enterprise use of VoWLAN), and generally harassing most of the other speakers with questions.

Some interesting snippets that are emerging:

- Skype is already doing VoIPo3G with E-Plus in Germany, initially from laptops with UMTS data cards. It's looking at other operators, and also at working over EDGE. At present it creates about 1MB traffic a minute, so it's not useful on the hideously overpriced 3G data services available in Europe (often $1-3 per MB, and much worse on international roaming). E-Plus is one of the first operators with a sensible flat-rate data pricing strategy.
- Interestingly, another analyst at the event (coming from a very defensive and cellular-operator centric position) suggested carriers should maintain this type of pricing structure to limit the threat of wireless VoIP. Given that my view that wireless data transport in Europe is 2-6 orders of magnitude too expensive, it seems that this is a massive opportunity either for new entrant cellular "challengers" or WiMAX / TDD / other wireless IP broadband operators. Back to my rallying call of "Just give me a pipe!"....
- Disagreement on the timelines, volumes and impact of dual-mode WLAN/cellular devices and converged FMC services. My view is that volumes will be lower than many think (I've been forecasting complex stuff for too long to draw oversteep hockey-stick uptake curves - I think there are lots of practicalities that will dampen the market growth rate), but the wider impact on cellular pricing and strategic realignments in the industry will be greater. I'm also ever more convinced that UMA is a non-starter - not so much because of the technology, but because of the commercials and ways of building the user interface, customer support and billing etc. I'm also starting to wonder about the timelines on SIP-based dual-mode, and whether my (relatively low-ball) predictions may be too bullish. I'm particularly skeptical about the market opportunity for non-smartOS phones with WiFi
- Starting to see more people talking about using cellular over low-power GSM/3G picocells as an alternative to VoWLAN in-building. I've been mentioning these at FMC conferences for at least a year, but it's the first time I've seen so many other people appear to take them seriously. This fits in with trends & opportunities observed by these guys and also the recent announcement from these other guys, that could help drive wider development of picocells, and even the new "flavour of the month", home "femtocells" hanging off DSL lines. I'm already looking forward to the Daily Mail's headlines about base stations in your living room.....

Too good to be true?

I've woken up this morning to several news stories, emails & blogs telling me about the "new new thing" in wireless broadband - a proprietary technology being called xMax by its inventors xG Technology . In fact, the technology is supposedly applicable to wired applications as well (although that's true of UWB as well)

It all reminds me a little of other "You what?!" broadband innovations - sending signals through gas pipelines , or via hovering airships. There are always stealth startups popping up, claiming to be about to change the world. Sometimes I guess it comes true.

It certainly sounds clever. And it may well work as claimed for certain usage cases. Whether it's commercial or not is another question. On the other hand, it certainly addresses some of my usual questions asked to mobile broadband evangelists: "will it work in-building?", "will it be low-powered enough" and "will it work internationally?".

But ever the cynic, I'm also looking for any possible stumbling blocks. I have to admit that some of the fine-grain discussion around RF modulation and encoding is beyond me, so many of my questions may have simple and obvious-to-some answers. But.....


- Are there any issues around latency? Does lots of clever signal-processing and error-correction add so much time to a "roundtrip" that it's useless for realtime apps like voice?
- Is there anything about xMax that means it doesn't work well with IP or ethernet protocols?
- Is there anything else new or around-the-corner that could interfere with it? Other types of UWB, for instance?
- Can the receivers be small enough to fit into phones?
- How well does it work with moving devices / users? At what speeds?

... and so on. Clearly, there are 100s of other questions that need answers.

My overall take is that xMax if probably now over the first hurdle - people are taking it seriously enough to at least sit up and notice it. I'll be watching it carefully, for one.

Just give me a wireless pipe....

Great piece by James Enck at EuroTelcoblog today . But my view is that the IP/Internet issue is going to hit the mobile operators even harder than the fixed guys, and that the cellular industry is even more ostrich-like in its state of denial at present.

I reckon a lot of the hybrid fixed/mobile operators like FT/Orange, TI/TIM and BT(+MVNO) now "get" the issues around VoIP, and the impact of MSN, Yahoo!, Google, Skype/Ebay et al.

They might not like it, but I reckon most of them understand its importance.

It is notable that BT is pretty cozy with Yahoo and FT with Microsoft.

I also think they have internalised that the way they will make money in the future is by becoming "smart pipes". They are rushing to deploy IMS or other IP-based NGN core transport networks, with various types of application platform layer sitting above it.

Sure, they will try & slow down the rate of change, but fundamentally, the fixed operators have long understood the value-added opportunities associated with network connectivity. This is because they have long dealt with selling wholesale services to other carriers, and above all the complexities of the enterprise communication world.

Fixed operators sell plenty of "smart pipes"already. They provide IP-VPNs, managed security services, they resell and maintain IP-PBXs and do all sorts of other complex nuts-and-bolts services with acronyms like MPLS. Increasingly, they are pushing into mainstream IT services, and enterprise LAN/WAN sales and management.

And although there have been some strange outbursts about MSN, Skype et al from some carriers, the fact remains that they have long accepted the existence of 3rd-party VoIP for millions of users. Fixed carriers aren't complaining that corporate firms are using "their pipes" for services delivered via grey boxes with Cisco and Avaya logos. Many of them sell & install IP-PBXs themselves.

None of this is glamorous stuff. It doesn't involve glossy TV advertising campaigns with pop stars & footballers. Their office Christmas parties will probably be full of boring people in boring clothes, talking about boring network widgets. But it's got a fighting chance of being profitable. It's difficult to commoditise hardcore network engineering expertise.

But the mobile operators are different. They are addicted to the consumer market, to glossy marketing, and the idea that absolutely everything is a "service" and therefore billable. Although some grudgingly wholesale capacity through MVNOs, they're still incredibly sensitive about being seen as pipes, especially "dumb" ones. Most don't have a sophisticated enterprise group that is educating the rest of the organisation that pipes can be "smart", albeit at the cost of being dull. Instead of trying to offer robust mobile IP-VPNs so users can deploy their own VoIP solution, they're still trying to pretend that "mobile PBX" solutions can succeed. Many also seem convinced they can offer billed IM services that don't interoperate well with MSN and Yahoo.

Many seem uncertain about deploying IMS, worrying that it could open the floodgates to 3rd-party IP applications. What they don't realise is that it's going to happen anyway, and that opening the floodgates is the only way to avoid drowning in the flood.

Presence or Absence?

I seem to be bombarded with companies wanting to talk about "federated presence" at the moment. There seem to be 100s of variations on a central theme:

"People have too many phone numbers / devices / services / IM accounts / all of these. Nobody can ever reach them without leaving messages like "I'll call you on your mobile instead" or "u there?" How can we simplify this & make some money as well?"

There are operator-based solutions, web-based ones, PC-based ones, handset-based ones, enterprise collaboration-based ones and IP-PBX based ones. Some of them rely on the user to update their presence information, others infer it from factors such as whether your mobile phone is switched on, your Skype client online and so on.

The idea is that you (or your service provider) should be able to set rules.... "don't phone after 8pm with work inquiries unless you're my boss"... "don't phone me when I'm travelling in a different timezone & it's 3am"... "I'm on a phone call, but email me if it's urgent"... "send my email headers to me via SMS except when I'm logged into MSN in which case IM me instead", or whatever.

All this is great in principle. But I have my doubts about just how viable this wonderful "unified presence" concept really is. Can people really be bothered to update their presence religiously? It doesn't "degrade gracefully" with non-compliance.

I reckon that if 90% of the users use the system 90% of the time, it'll be great.
If it's 80% of people & 80% of time, it'll be marginally useful.
But if it's 70% of people complying 70% of the time, it'll be an active pain in the backside & create more problems than it solves.

And if you're really important, do you actually care if other people have to jump through hoops to find you? Surely that's their problem, not yours?

Maybe "absence" will be simpler and more important then "presence". Rather than try & tell people the "best" way to reach you, why not just tell them which is worst? "Don't both calling my mobile, I'm on a plane". "I've got 1000 emails to reply to. Try something else."

Overall - it all sounds utopian. But I have a sneaking feeling that practicalities may get in the way rather a lot. The user interface is going to a major stumbling block, even if the back-end technology works.