Telepocalypse

Your name in pixels

Ooh look, someone's cobbled together a fun article from my blog posts on UMA.

It's a 95% accurate precis of what I said, although there are a few minor snafus in the journalistic precis (e.g. I don't yet do any work for telcos -- I rarely visit the dark side -- although it's on the cards).

It wasn't until I started getting quoted in mainstream press that I discovered that reported speech in quotes doesn't necessarily correspond to the actual sequence of words said by the speaker.

Anyhow, whilst I'm on the subject of mainstream media... In case any of you think dead trees won't make it through the digital revolution, here's a counter-example:

Exciting, huh? It's from my parents' house. Apologies for the chintzy furniture and tasteless lampshade. I'm hoping they live to be 120 so I never have to inherit it.

But look closer.

See the copy of British Airways News in the magazine rack? I'm guessing my Dad was probably given it by a friend who still works at BA. But before he retired, he'd bring home a copy every week. And my mother would read it, because it's there. There's the usual internal editorial, but also lots of adverts aimed at airline employees, plus listings of destinations with low loads good for staff travel. My Dad even got kudos for remembering to bring it home for her -- a small token of caring.

Now think how you would achieve this in an all-digital world. It just doesn't work. The casual veg-out in front of the TV with the paper on the coffee table. Beyond replication. No amount of directories, federated ID, security, syndication and wishful thinking is going to deliver the same result in the near future. There's no firewall keeping copies of BA News away from spouses of BA employees. But the digital gulf between BA and those people is immense.

The lesson? Beware of technology as a panacea. The real world is complex, messy and analogue.

Your aircraft is waiting, Mr Geddes

The airline industry has all sorts of structural similarities with telecom. Capital-intensive, prone to big busts, and obsessed with price discrimination. Both have also suffered intense government interference for decades. Bums on seats, packets through routers.

My parents are due to fly up to Scotland tomorrow to help my wife look after the kids whilst I'm away at Freedom to Connect. (Be brief, Martin - you've still got to write your keynote speech...) As retired former British Airways staff, they're flying up standby. There are about 60 flights each way every weekday between Edinburgh and London, and lots of competition between 5 airlines. But there are probably fewer BA flights each day than when I first flew the route about 25 years ago, when BA was a state monopoly.

Anyhow, I was just reminiscing on this, when I remembered that BA used to guarantee that if you turned up, they would sell you a seat and fly you. Apparently the feature started in 1974. They branded it as their Shuttle service, and I remember TV ads where there was an empty plane save the one excess passenger. Yes, they had spare aircraft at the ready at each end just in case a flight became oversold.

So the expensive monopoly was using guaranteed quality of service to justify its unique position. Indeed, they could threaten that if competition was allowed then such benefits would have to go.

Sound familiar? Give me 60 flights a day and rock-bottom fares anyday!

Anyhow, we were once the beneficiaries of this policy. A flight that we were listed for sold out, so they laid on an extra plane for the overflow passengers, which we staffers on standby also got on. I think it was a 757, quite a big plane for 10 passengers. When we got to Heathrow, the pilot had a bit of a heavy landing, but the plane was so light that we bounced off the runway!

Quality of service really isn't good for you, after all.

Don't mail me, I'll poll you

Email is broken. I'll write a rant on it another day. But if you ever see me with a Blackberry, please increase my medication dosage until I stop mumbling and lie down quietly.

My basic view is "every email is a failure to have found a more appropriate contextual communications channel".

But, what if we tried to rescue something from the attention wreckage of email? Here's one idea. Why don't we have a standard where mailing lists can tag their messages with a header like "RSS-Tag: http://www.telepocalypse.net/". The email reader would do two things:

• It would have an option to by default re-direct such emails into your feed reader. See it, read it, goodbye. No need to delete or file. No need to treat it as a "to do" item, which is how most people use their inbox.
• It would use the autodiscovery mechanism from the URL given to offer the user a choice of additional, real, RSS-based feeds.

Is such a thing really so complex? Surely the benefit outweighs the cost? And it's entirely edge-based -- no need to upgrade any email servers or relay points in the middle. I already manually use email-to-RSS, but if it was automated, it'd be so much easier.

Web 3.1 beta 2

I think this list of Web 3.0 features is really worth absorbing. As a recap, "Web 2.0" is a shorthand for dynamic user interfaces and a first cut of de-coupling application from data using web services. I'd re-phrase their "Web 3.0" vision as follows:

- No user data is hostage to any one service silo. (Web 2.0 isn't there yet.)
- Aggregators will squeeze value out of the juxtaposition of any and all data. "No value left behind."

- You have a single, integrated historical view of your digital life.

- All data is perfectly liquid, loosely coupled and re-combinable. (DRM on your metadata = commercial suicide.)

But there's one thing missing, and maybe it'll have to wait until "Web 4.0". It's what we drew on our platform roadmaps at Sprint, which was "Commercial Web Services". In this world, money is able to flow in the opposite direction to the bits. We foresaw that Sprint was in a great position to become a commercial web services directory and broker. Indeed, all those "billable events" and pricing plans suddently became relevant again; although the end-user experience might be subscription all-you-can-eat, there was underlying accounting needed for the disparate data resources that the applications drew together. And Sprint was in a perfect intermediary position, because it could use it existing wireless and fixed customer bases as a bootstrap to launch a kernel of commercially viable web services for identity, messaging, provisioning, authentication, etc.

There's an existential danger to the telecom industry that the hub for passing money in exchange for bits falls outside their domain. For "telecom" is really only a bundle of vaguely associated things brought together by technological and historical accident. The network can be used as a loss leader to bring people into your marketplace hub. The complex and expensive billing system can be turned from a bug in the business model to a feature, as long as that billing and settlement is "downwards" to atomic service providers, and not "upwards" to users.

But fail to build such hubs, and all you got is loss, rather than a loss-leader.

Dead and buried

I'm too squeezed for time from the day job to blog much at the moment, but here's a gem I can't miss sharing.

Handsets get taken to the grave

More people than ever are asking to be buried or cremated with their mobile phones when they die, say researchers. ... Martin Raymond, director of international trend-spotting think-tank, The Future Laboratory said that this had started off "in the realm of the urban myth", but was fast becoming fact.

Now, what's strategically important about this? Well, the telcos are all desperate to finesse in some way the Paradox of the Best Network, which says that the stupidest dumb pipe is the best network for users, but the hardest one to make money from*.

Oh, and the asterisk?

"All other things being equal."

There's your get-out clause. Internet Protocol is just an abstraction that makes it easy for any networks to peer, and any device to attach to any network. But the abstraction doesn't necessarily match reality ("it fibs"), and it also doesn't cover all aspects of the real world ("it's incomplete"). This is fine, because the aim is to hide some of the complexity of real network topologies, varying pipe capacities, etc.

This "bury me with my phone" is an example of the latter category. IP has nothing to say about the user's affinity with their communications devices. So one way in which mobile carriers can search for new business opportunities is to hunt for more "personal affinity re-inforcement" businesses. In some ways the ringtone market is a shallow example of this.

By exploiting the mismatch between the abstraction and the real world, carriers can sneak between the chinks in the stupid network's armour to find new value sources that don't require price discrimination of every packet and network neutrality battles.

Full details for the usual immodest consulting fee ;)

Yahoooops!

Three things about the new Yahoo messenger beta that's desperately struggling to reach parity with Skype by adding PSTN in and out services and the same wideband GIPS codec for PC-to-PC use. Two bad, one good.

Number one. Telling me my ordinary, standard Firefox-on-XP setup doesn't meet your all-Microsoft standards doesn't endear you to anyone:

Secondly, putting the address book in a handy accessible tab with all the contact methods visible is good. This isn't a new feature, but it's still good. I'm not convinced anyone has really yet got the right model for buddies, acquaintances and contacts. My Skype buddy list grows and grows, despite many of those interactions being one-off or temporary. The tool isn't mirroring the real social relationships I have. The dissolution of a "buddy" relationship is too socially fraught. Yahoo's model isn't perfect either, but a tighter buddy list with a separate contact space is probably somewhat closer. The difference is minor, though.

Now for the big'un. I think Yahoo is making a terrible strategic blunder and is failing to leverage its own advantage in community. Yahoo is about the triumvirate of media, community and communications. They're all supposed to benefit from each other.

I'm a member of several Yahoo groups. In particular, there's one for my younger daughter's rare medical condition. (Her business, not yours. Don't ask.) But the new client doesn't integrate these communities. There's no "there" there in which we can meet. The electronic world gives me the opportunity to be in multiple places at once. One place I want to sometimes be is "hanging out" in the virtual meeting space of people concerned with this medical condition. But there's nothing like "3 members of XYZ Group are online", and then invite me to enter that space and negotiate the privacy and permission stuff.

Yahoo's job is to broker great conversations. Adding radio into the IM client isn't nearly as powerful as bringing disparate folks across the world with a common interest.

PS - When I click on "What is Yahoo! 360?" I shouldn't get a blank web page. And when I try to log in to Yahoo! 360, I'm not supposed to get this either.

Skype is successful because (a) it works, (b) everyone has access, not just Americans. If you're looking to build some complementary good like a Wi-Fi phone, Skype's still the more attractive option.

Spindrift

The significance of new giant solid-state hard drives is probably higher than many readers imagine.

I used to work at Oracle from '97 to '01 as a server technology specialist, mostly poking around high availability and very big transaction systems. Databases like Oracle perform two main functions:

- They manage concurrent access to data from contending users.
- They ration out I/O, and abstract that process away from users.

The rest is just features, as they say.

The end of the spinning hard drive will be disruptive, as it undermines ones of the two central pillars of our information storage tools. I wouldn't rush out and short Oracle stock right away, though -- they've diversified into apps and this is a slow process unfolding over a decade or more.

But it will cause seismic change -- an equivalent of the invention of optical networking, say. Whilst CPU, transmission and storage have become cheap, I/O itself has been relaively expensive. Many disk arrays attached to Oracle databases are mostly empty. They're being bought for multiple spindles, not storage area. Order-of-magnitude shifts in the relative costs cause structural industry changes.

The telecom angle is that faster I/O means we've got an even greater ability to soak up connectivity. Your kids will laugh that you called a 1 megabit connection "broadband", and even had to pay for such meagre resources. Roll out that fibre, folks! You're gonna need it.

One, two, many

Whilst I'm neatly balanced between flu and pharmaceuticals and enjoying some fading lucidity, one parting thought for today.

I can't name my accomplice, 'cos it busts his girl's privacy without her permission. But he tells me that his tweenage daughter has taken to a new form of social communications media (reproduced with permission, though):

one little strange thing my daughter and her friends do with skype is they put their best friend's picture instead of their own in the profile, strange huh? It seems that part of their identity is their best friend and they advertise who this is to each other.

Wowee. Cool! The users will always re-invent your products in unexpected ways.

Funnily enough, some telcos did great with "family plan" products. Some even managed minor product innovation as well as billing changes. But none seem to have produced group-centric products. You don't have to be a business genius to see that the social group dominates the life of tweens and teens. Admittedly, it's a back-office nightmare with current systems to build such products, because identity management isn't up to scratch. (Lots of dirty telco secrets, but it takes good wine to extract them.) Then again, that isn't a problem for those aiming to win the whole schebang from scratch.

Bonus offer: Telepocalyptic stock tips!

I once claimed I had the drawing ability of a three year old, with suitable evidence.

I'm going to have to revise that claim. My 2½ year old first daughter is showing signs of rapidly overtaking me.

Anyhow, the keen-eyed of you with a real browser (i.e. anything but Internet Explorer) may have noticed in your that there's an icon for Telepocalypse next to the URL, and it's also shown in the tab header. In keeping with my artistic disability level and web design budget, it's a pure-black square. Kinda tele-apocolypticish, maybe. (Promise never hire me to do your brand management or graphic design, OK?)

This thing is stored in a teeny file called favicon.ico. If I can make one, so can you. There's lots of free online sites to help. Artistic ability costs extra.

If my web site didn't have one, it would grate with me. What's the point of a blog and domain if you're not going to push your dodgy global microbrand to the limit?

I see it as kind of a sign of "We get the Net!". So here's my investment portfolio advice for those wishing to lose money on telecom rather than horses. Just pick the ones with favicons. You'll lose less.

First in the doghouse is Vodafone, who couldn't quite manage to downsize their (actually rather good) speech quote logo. But then, you didn't need me to tell you they've had lacklustre returns.

On the other hand, step forward old, cuddly BT Group! Double points for making the icon the letters "B" and "T" and not the rapidly-dating psychedelic colour orgasm. And, guess what ... the investor returns (whilst still not pretty) look rather less drab. And it might just get better, too: they're organised right, and have some positive intentions.

But it only gets better! A follower of US telecom? Well, Bellsouth must have been cool dudes, 'cuz they've got one. But the new at&t doesn't even get the superficial integration of the web site landing page right. Boy, am I glad I don't work in their IT department right now.

We can safely predict that logo-fuelled Qwest will be bought out by Net-naive Verizon. The cablecos will rock: favicon logos all round. Sprint's new logo (tagline: "We're Midwestern, OK?") makes it up.

Who will succeed in the UK mobile market? Well, T-Mobile have staked out their "dumb pipes R us" with Web'N'Walk, and they've got one. Orange are going the opposite way: build up a trusted brand around not ripping people off, then have a cosy brand bonfire and squeeze them till they bleed. Who knows! But poor old O2, well, I'm sorry but Telefonica blundered. It's supposed to be dull, legacy icon-less acquiring companies that are supposed to be buying iconed prey, not the other way round!

Hours of fun for all the family, and saves you a ton of money in investment advice!

Uh-oh. Time for more paracetamol and sweaty body temperature undulations.

The body is sick, but that damned mind...

I've got the flu. Sick all weekend. I'll spare you the details. Praise be to paracetamol!

Nonetheless, the brain keeps ticking over sporadically. (Unfortunately, the guts churn over sporadicaly, too.) So here's a therapeutic kicking of a common telco piece of received wisdom. I know I'll feel better afterwards.

"Fast follower".

On average, it's rubbish.

I've heard it over and over again. "We're fast followers." Gee, well I guess they were too.

The problem is a narrowness of framing. Telcos still think in terms of competition with other telcos, particularly on the wireless side. The assumption remains that a service only pops up on the network once it's been given the telco blessing and launched. OK, you allow a bit of innovation at the fringes a-la i-mode and a thousand download portals. If a particular service becomes popular, you can always buy in a version of it. Picture messaging becomes fashionable? One picture messaging product, hold the fries please!

Not!

What it ignores is that the competition is aggregating huge, integrated hubs. The Myspaces and Flickrs of the world are sucked up into even bigger mega-hubs of Yahoo! and Newscorp. They don't do termination fees, obscure interconnect agreements, and state-by-state regulatory negotiation. They actually understand their customers, because they focus on doing a few things well. You're going to have to negotiate with them instead, and it won't be comfortable. They know time is on their side.

The alternatives? Plenty. For example, "integrate, participate or cultivate". Integrate is a product-centric strategy. Build your own hub where the products are truly integrated, and voice and messaging sit at the heart. The whole more than the sum of parts. Participate is a redefinition of the business model. Shred it into wee bits, stick (billable) APIs into your app and biz platforms, and see what other people find useful in picking over the telco carcass. You never know, they might find something useful. Cultivate is a customer intimacy play. Grow a closer relationship with some sub-segments. Re-org yourself to be less monolithic. Who's scared of being a dumb pipe when the capacity is being bought by some other more customer-centric part of your own business?

But "our competitive advantage is we like being second"? Commercial suicide, I say.

What you don't measure...

Imagine you're building a "Telco2.0" (ahem) with some kind of Internet-compatible business model. This means you're involved in one or more of dumb pipe deployment, making money from APIs to your business and IT platform, and some select apps which you do particularly well.

Now, what metrics does the CEO want on his dashboard to know he's getting nearer to the "IP-compatible" business model? Hint: ARPU, CPGA, and churn are lagging indicators tied to the old business model. Business performance matters, but like flying through a storm there's a period during which a safe landing is more important than fuel economy or punctuality.

What metrics do you use to measure your "permission" assets from customers to access their attention?
How do you measure your product portfolio's "stupid network compatibility"?

How do you quantify your success in delivering compelling, sensuous presence experiences?

Rather than me give too much leading thought, it's over to you. What metrics should telcos be using to track whether they're getting closer to the goal of an IP-compatible business model?

Feel free to tailor your answers into specific segments like "wireless-only players", etc.

R U 0wn3d by ur telco?

Two related bits of pundutry out there that illustrate my belief that success in network operation comes from innovation in funding and pricing models.

First, the prolific Mr Frankston:

Instead of having the strange phenomenon of carriers spending billions and then arguing that they deserve to be paid we'd have them bidding on contracts to install and/or maintain connectivity to a marketplace that is buying capacity and making it available so value can be created without having to be captured within the network and thus taken out of the economy.

To achieve this we need a different "unbundling" regime. Maybe call it the "secession" regime. I don't believe the natural unit of purchase of connectivity is the household; the costs of marketing, billing and support are too high. Imagine that everyone connected to a central office exchange could vote to leave their "service provider" and collectively choose someone else to operate all the equipment in the switching centre. There are a number of possible fiscal models, such as absolute buy-out of the local loop assets, or rental and maintenance payments.

Here's the other bit of related "who owns what resources" news from Techdirt on unlicensed spectrum:

Last year, we were intrigued by the new spectrum allocation plan in the UK that would create a much more open market for spectrum. The idea was that rather than (as the FCC does) setting exactly what the spectrum must be used for and who can use it, the market is allowed to decide. That is, anyone can purchase the exclusive rights to the spectrum, but then, they can do with it as they want -- whether that's selling it to others, or making use of the spectrum.

Here's my idea. Some philanthropist with deep pockets wanting to make a lasting impact on society should just go buy up chunks of "private" spectrum, and then open it to the public in a managed way. In fact, we've got a great model for this already with physical cultural and landscape assets over here: the National Trust. (Hey, dig that commie tagline: for ever, for everyone.)

The good news is that there are enough tech billionaires out there who can pay as well as understand the benefit. I'm sure we can figure out a licensing scheme that immortalises your name onto every device operating in that spectrum band...

The bad news is I'm not one of them.

You can quote me on that

Thoughts for the day:

- In physical transport (atoms), we found that there was a complex business called "logistics" which is where the profit is; "trucking" is just a small function of logistics. With data networks (bits), it's the exact opposite; there once was a complex "bit management" industry called telecom. This is slowly, painfully fading away. It'll get replaced by a simpler industry that just teleports bits over geographic distance. We don't have a name for it yet.

- Telecom is just another utility delivering stuff through pipes. In fact, it's just like a water utility -- except you don't have the glamour of sewage disposal.

And a phone call? Priceless.

I had the pleasure in January of briefly visiting Seattle to see my old friend and mentor, David Anderson. On the Saturday I went with him and his elder daughter for a toddle around the canal locks in Ballard, followed by a coffee (what else?) at a local cafe. As David pointed out, this cafe specialises in cupcakes, thus differentiating itself from the ubiquitous Starbucks and Tullys -- a "plus one" marketing concept.

Being January, and Seattle, it wasn't exactly tropical outdoors. When we got in to the cafe there was quite a long queue, and all the seats were taken or reserved. (This latter practise of bagging a seat which stays empty whilst those ahead in the queue are then forced to stand and drink is not my idea of good manners.)

Anyhow, it occurred to me that the queue itself was part of the experience. You get to longingly gaze into the cabinet of cupcakes, discuss with your friends which type you want to guzzle. You're in the warm, out of the chilly wind. You can smell the aroma of the drinks being prepared; there's a sense of anticipation in the palate. You watch groovy Ballardites doing their Saturday thing. You aren't too hurried and flustered in selecting a drink and encoding your order in the local coffee dialect -- "A ventissimo ne plus ultra grande sucrified milky-milky cocoa beverage, please." (Handy hint: theobromine is a better mood-altering drug than caffeine.)

I believe I once read of studies of how much people's enjoyment of Disneyland rides varied with how long they queued. The result? Zero queue was not the best! Even the newer fast ticket bypass system still makes you wait in anticipation for your allotted time on the big ride, just without the pain of standing in line.

The lesson from all this is that value in the eyes of the customer is not some simple end-product of an industrial process. Indeed, it can be hidden and subtle. The activity is not the same as the user's goal. At the extreme, what appears to detract value may be an intrinsic part of the experience; and adding more features and capabilities in fact decreases value.

With various consulting clients, I've been unpicking the value proposition of messaging and telephony. You can read about some of the results in my ETel ketnote. But here's an example that doesn't appear in any of my client reports of how value is created and destroyed. (Disclosure: I think I got the insipration for this one from an article I read once, but I've lost all references.)

When you send an SMS to someone, you have almost total certainty that it will be consumed on a mobile handset. Furthermore, that handset is a personal device used by one person alone; and there is a social taboo of messing with other people's handsets and checking their messages. So there is a strong privacy value element to SMS, because the sender can make assumptions about how the message will be received and consumed.

Now, imagine for some moment that an overexcited telco doing some triple/quadrouple/whatever play bundle decides it would be a great idea to integrate messaging and telephony. Now, if your TV is on then your text messages also are displayed there in a little scrolly box at the bottom of the screen.

Does this new feature create or destroy user value? My belief is that it destroys more privacy value than it creates in distribution value. No longer can senders be sure that messages will be consumed on a small, personal, private 1.5-inch screen; rather, they may be broadcast to the whole household or visitors. The "feature" has net-negative value. But the only way you'll work this out is by going through the intellectual exercise of decomposing the activities involved in the product and reverse-engineering out the underlying value to the user.

Sadly, my experience in telcoland tells me you're liable to be labelled a freak for even trying.

Internet as consumer surplus engine

I've long described the Net as a "consumer suplus engine" to anyone who'll listen. It takes previously bundled billable services like voice telephony, strips out the service from the connectivity components, and stops you being over-charged for the service because of lack of competition on the connectivity.

Anyhow, here's the verbatim quote from Paul Kedrosky, who has found an academic paper that puts the numbers to what's obvious to everyone:

Only about 0.2% of consumer spending in the U.S. ... went for Internet access in 2004 yet time use data indicates that people spend around 10% of their entire leisure time going online... Based on expenditure and time use data and our elasticity estimate, we calculate that consumer surplus from the Internet may be around 2% of full-income, or several thousand dollars per user.

Wow, that's pretty awsome, isn't it? You're only paying for about 10% of the value you get from your Internet connection. That other 90% becomes a budget for other, exciting services and activities. Who knows, your non-telco job may be funded by the consumer suplus of the Net!

Personally, I suspect that this report will under-estimate the real value, because it only considers monetised value, and users as consumers of content and service. How can you put a price on the value of open, democratic discussion, for instance?

PS - I haven't got time to read the paper; anyone who wants to write a guest blog post on it, feel free to pitch me!

It's good to talk ... M&A

Apparently there's a rumour that BT is a possible acquisition target.

Of course it is. It's a great property company. Even comes with a useful little telco attached.

Look! Here's a recent picture I took of Cherry Hinton telephone exchange, near Cambridge in England:

Everything that went in the big building should now fit neatly inside that small green box outside. There's even a wopping great exchange building just down my road. At the prices we have over here right now, who wouldn't want to get into the used building and scrap copper business?

Telecom training

Promise you won't tell? Really?

It's been quite nice being away from family duty on a business trip to Paris and London for two nights.

Ssssshh. It's just a secret between you and me, OK? I've always wanted to try out the Eurostar train from Paris to London, and it's certainly quick. Although someone needs to buy them a few sponges to clean the windows a bit.

Anyhow, I passed through London's Kings Cross station today on my way back up to Edinburgh. Here's a few vignettes from around the station I'd like to share with you.

Firstly, here's the automated ticket machines and an economics lesson for telcos.

Now, I've got an APEX ticket that only is valid on this 2pm service from London, and I collected my ticket from one of these machines. But I got to the station at 1pm, and there was a 1.30pm departure I could in principle have taken. The train operator, GNER, missed a chance at an up-sell here; they could have offered me (for a price) to go on the earlier service. There's always a price at which they won't undermine the price-discriminating effect of the original APEX ticket. It's free money! Likewise, they could try to do impulse up-sell to first class. This train is also one of the older 125mph diesel ones, as it carries on from Edinburgh to Aberdeen. The earlier (all-electric) service probably had Wi-Fi on board, so they've also missed an opportunity to see me connectivity.

I've been putting in my client reports some examples of how telcos fail to manage up-sell, cross-sell and impulse buys. For instance:

• You can normally call someone back from an SMS by just pressing the green call button; so why not at the bottom of each SMS draw a horizonal bar, and then say "To call Bob back press [green button icon]"?
• The phone and network knows who I SMS and MMS the most. Where's the one-click impulse "Send this message to Jane's mobile" menu option? Amazon don't own a patent on all impulse buying, you know!

Next up, here's a fuzzy O2 advert up above the entrance to the newsagents:

And the message, in close-up:

This is an iMode promo with a a picture of a field and the slogan "bid on eBay from here". Now, I'm sure there were some good expensed lunches for the biz dev teams, but this model is ye olde school "Look mama! I've got a link on their portal!". Will you be able to register your mobile number on eBay and have a seamless federated identity experience? Probably not. And even if you can, it's not enough. Because eBay is only one of many, many services with which you probably want to interact. If the mobile industry is going to have a future doing anything apart from bit pipes, it can't be this. Other people are better at portals, and can integrate more media properties together and scale their partnerships better. But if O2 offered a more open platform with a bunch of APIs, and don't put up hurdles, then there's a real chance that the public will find lots of services that have their mobile number automagically become easier to use on O2. And not just the few that they do deals with. Gatekeepers, burn your gates!

Next is this snippet from another illuminated advert.

The track infrastructure in the UK is operated by a nationalised body, Network Rail. (There was a literally catastrophic period when it was privatised; another telco infrastructure funding lesson for another day.) The trains are operated by various private companies. In some places, there's competition on parts of routes where the franchises of multiple train operators overlap. (The original rail infrastructure in the 19th century had multiple vertically integrated operators competing with different routes to the north and west!)

This ad is telling commuters from distant Peterborough that by putting up with the slower stopping service that regional operator WAGN offers, they can save £1000 a year (about US$1750) over the long-distance GNER service.

Competition works.

This brings me back to my old bug bear of network neutrality. This tries to get the outputs of competition without the inputs on, um, multiple competing services. Unsurprisingly, it's a dud. Suppose you pass such a law, and outlaw all "packet discrimination". You enforce it. Success? No banana, babe.

Say I'm a small business customer, and I don't care too much about how fast my connection is. I even don't care too much if I have to pay a bit more to get a "VoIP-compatible" jitter-free unblocked connection. But I really care if the service goes down for more than a few hours. A domestic DSL contract promises a fix to a service outage within a "reasonable" period, where "reasonable" is defined by the telco. They offer a "business-class" service that is functionally identical at five times the price with a sensible SLA. I'm not happy, but I have no choice. Network neutrality merely entrenched the single-supplier structure, delivered benefits I cared little about, and didn't give me the outcome I wanted.

There are many facets to a connectivity service, and different people place varing values on each of those facets. Picking one of them, application price discrimination, and artificially mandating an outcome in that one facet, is a pale shadow of what competition or alternative network ownership structures might deliver. "Positive discrimination" policies have been a spectacular failure for people; why should we think they'll align behaviour and incentives for technology any better?

PS - This train has a super-cool feature. Look!

Someone at GNER understands they're in intermodal competition with the planes, and the plane folk don't get it. Push home your advantage, whatever it is!