The lessons for FMC operators from airphones

Today's announcement by Ryanair that they plan to enable GSM roaming on their flights provides a useful lesson to carriers thinking of deploying fixed-mobile convergence (FMC) solutions. Historically telephones on aircraft have been a commercial flop. As we will see below, they have failed to satisfy the user need. It is possible most FMC technology has the same problem, but from the opposite direction: attempting to solve every problem at great expense to ensure lock-in, rather than the parts of the puzzle of most value to the user.

Understanding the problems of airphones

Many reasons have been put forward for the failure of in-air phones, such as cost, lack of privacy, and audio quality issues (from background noise). No doubt these are factors, but that doesn't seem to stop people from talking on trains for the last decade or more. The Telco 2.0 approach is to decompose the value proposition the the user seeks, and then see how well the product satisfies that need. This is a core skill for the new all-IP world. When an industry goes from vertically integrated to horizontal specialism you need to be able to pick the unbundled parts of the value proposition you will specialise in.

In the case of a phone call, the user has to select their contactee from an address book; dial; and persuade the contactee to answer (is the caller ID someone I want to talk to)? Should the callee not answer, then there needs to be a means of leaving a forwarding identity (call me back at...). That means knowing your identity/number, and having the technical means to take an inbound call. Even if the call connects, there may be a need for the callee to call the caller back to complete the user's objective. This may need to occur after the flight has completed -- and the callee may not be aware of when you dip below the magical 10,000 feet and all powered electronic items suddenly turn into verboten objects of subversion.

Simply being able to place an outbound call only filled a small part of the user need. The number you need to dial is likely to be stored in the address book or call log of your phone, which naturally you've buried deep in your bag in the overhead locker, which you can't access from your window seat, and aren't allowed to turn on even if you can get it Even worse, the callee's number may be locked up in your operator voicemail box as a caller ID or dictated message, necessitating even more expensive air-to-ground calls. Even that assumes the user can remember the way you access your voicemail from handsets other than your own SIM-enabled one.

And that's just the beginning of the user's problems, as you can see from the above "lifecycle" of the conversation. The complete value proposition includes address book, outbound identity (who am I?), inbound identity (how to contact me back -- which may not be a phone call), outbound call handling, inbound calling and access to voice and other messaging services.

We didn't dig into the privacy issues during the conversation, either. The call need could also broadly fall into one of three categories: information, transaction, or chatter. Information can be relayed as long as it isn't some sensitive confidential business. Transactions are harder: do you really want to read your credit card details out to all seats between 31A and 33C? And personal chatter will lack emotional impact due to reticence amongst strangers.

Finally, payment isn't part of the value proposition (as it's something you take from the customer, not give to them). Still, having to get your credit card out is another opportunity to have a "can't be bothered" moment.

The Telco 2.0 prediction is that the Ryanair service will be a success because it fixes all these problems. Furthermore, we would expect the bulk of the revenue to come from SMS, then from inbound termination fees, and lastly from outbound call fees -- a complete reversal of the airphone model.

The FMC lessons

Much of the effort in deploying FMC technology has focused on one very narrow problem: hand-offs between fixed and mobile networks. With the airphone example, the operator didn't solve enough of the user's problem. In this case, it focuses on solving the least critical part of the puzzle, one that the user may care little about. The attraction to the operator is clear, in that they can offer a feature which is intimately tied to their network and handsets. The risk to FMC operators is bypass by nimbler players who cheaply pick off the piece-parts of the puzzle, without having to carry the costs of the FMC infrastructure.

Skype already offers me the chance to verify ownership of my mobile number, by sending me a security code to re-enter onto their web site. I can then send SMS messages with an outbound ID of my own cell phone number. Likewise, services can piggy-back onto existing channels to offer consolidated inbound and outbound identities. One unique solution for landlines is PhoneGnome, which can intercept inbound and outbound PSTN calls and bridge them with VoIP services to "steal dialtone" from the incumbent operator. [Disclosure: article author is an investor and advisor.] The same trick is harder to pull off on mobile devices.

Similarly, a consolidated address book could be offered using SIM card exchange, Bluetooth, Outlook sync, or network-managed address books. Today it's a pain to provision the address book in your home DECT phone, but there's no reason that should be the case in future.

FMC proponents also gloss over the loss of utility FMC may cause. Sometimes I may wish to call the place, not the person ("it's coming to rain and I left the window open!"). I may want to partition my life, and turn off my mobile at night, but still allow urgent inbound landline calls.

In summary, FMC may be as much about satisfying the needs of operators to lock customers in as it is about fixing the problems of users. In the current environment that's a very dangerous strategy given the multiplying number of choices users have and declining market power that network access ownership offers. Many users already have the FMC solution they seek: the standard mobile phone, and nothing else!

Managing social acceptability

There are some other noteworthy aspects of the Ryanair service.

Passengers on super-discount Ryanair aren't likely to complain too bitterly about disturbance from ringing phones and dull conversation. The airline has skilfully set the expectation of service low, and generally over-delivers on the core values of punctuality, schedule and price. If you arrive alive for your €0.99, it's a success. On more upscale airlines, there a potential problem from social nuisance. It wouldn't be surprising to see some airlines compromise on allowing SMS and voicemail access only. This suggests that for once we may see some user benefit from vertical integration of network and application. A pure "dumb pipe" leaves you at best in a cat-and-mouse game with passengers on deep packet inspection and blocking of "anti-social" application traffic. Forcing users through gateways where you get to inspect the message adds value. The elimination of externalities of use exceeds the loss of innovation, and ensures a sound economic model. Dumb pipes are only a means to an end, not the end itself.

Connectivity is King (but it's more than just a pipe)

Boeing recently shut up shop on it's in-flight broadband service. This suffered similar troubles to airphone: a $1000+ communications device with limited battery life required, to be operated in a space where even Cirque du Soleil would struggle to swinging a cat. The service price was quite reasonable: indeed, for long-haul I would say it was cheap. Quality was patchy but generally OK. I've received a Skype call this way, so it's real broadband.

The lesson is that, again, you need to look at the whole user experience and need. People want to communicate with people. That means giving them connectivity, service and devices. Connectivity alone is a raw potato: useful, but not tasty.

Mobility is Queen

Interestingly, similar GSM-in-plane technology deployed elsewhere could end up spelling bad news for Verizon, Sprint and KDDI, who use incompatible CDMA technology. If aircraft standardise on GSM, I can see a lot of churn ahead. Each then has a nationwide network in 2D, but nothing in the vertical Y axis! Coverage always beats features, speed and service in wireless wars. No wonder the bidding in recent ground-to-air spectrum auctions has been so well contested.