Header image
Insider News
Every Sunday
Octymo Consultancy Ad

Fast Code


I once had a business lunch with a very important VC. We went to  Billy Goat Tavern, a cheap burger bar. Beer in bottles and plastic plates. The food was fine and the meeting a success.

The Billy Goat is famous in Chicago, it’s also unusual in that the food, while served fast, isn’t fast in its nature.

Mobile phone applications are built on the principals of fast food.  The most distressing statistic about them is  huge number of people who’ve only ever downloaded one application or game: They know how to do it. They have a phone that has been set up. The only reason they didn’t buy a second app was that the first one was rubbish.

Operators point to the Apple Apps store and its ease of use as the reason why the iPhone has made apps a success. I’d argue that there is more to it than that. The Apple Apps store is a success because the apps are good. And the apps are good because they are written for one device.

Developers complain about device fragmentation and how hard it is to write for all the different types of phone but they are looking at the problem the wrong way.

No. They are looking at the wrong problem.

The reason why there are so many different types of phone is that there are lots of different types of people.  Networks and handset manufacturers have sophisticated segmentation models for things we think of as early adopters, practical mums, wrinklies, flash chavs and the like. They give them names. Like High Street Urban Tribes and make lots of PowerPoint presentations about them.

While I’m being facetious here, the principal of making the product right for the kind of person who will use it is spot on. So why does the whole ecosystem try so hard to make applications where one size fits all? It doesn’t work because the applications then fail to take advantage of the devices. Some have special gaming buttons above the screen, but applications, even those that come pre-installed on some of those phones, don’t use them.

Developers should sit back and look at the phone to decide what will work. Something with a track ball suits centipede, a jog dial for a driving game and a multi-touch screen dance dance revolution.

There is a reason why games now go to the lowest common denominator and it’s money. While with an iPhone the developers get 70% of the revenue, with the traditional system the operator gets half. But developers don’t get the other half. With the iPhone you submit your application to Apple, and for all its well reported problems, Apple evaluates the program and puts it in the Apps Store. There is a cost to be a registered developer but app approval is free.

Getting your application onto an operators portal is very much harder. Operators don’t want to deal with individual developers, or handle approval so you have to go to an aggregator. These are companies which do the work the operators should be doing in the first place. Organising contracts with developers, running the databases and all portal provision. For an extra fee they will sort out marketing on the portal and for more money they will certify the application.

So let’s look at what this costs in money terms. The operator typically takes 50%. The aggregator between 20% and 30%, plus will charge up to $5,000 an application for certification.

It’s no wonder that developers have to cast their net very wide. Imagine you have a music application.  You could develop it for the Nokia 5800 or the iPhone.  If the application costs $50,000 to develop and you sell it for $5, there may be a cost of $1 per unit for the music you licence.

Under the iPhone model you need to sell 20,000 copies to make your money back. Under the traditional model you need to sell over 200,000 copies. No wonder developers feel they need to address every phone going but can’t afford to spend any time on making them optimal.

It’s a vicious quality spiral. The consumer gets a worse product and actually the network makes less money.

It doesn’t look like the networks have learned the lessons. They still think 50% is reasonable

The big shame is that Apple isn’t the first to have shown this. Long before there was NTT DoCoMo with iMode. The revenue share between DoCoMo and the developer is 18% to the network, 78% to the developer. And guess what. The iAppli programs are successful and lucrative. The shame is that when the networks tried iMode outside of Japan they didn’t bring the pricing system and it failed so iMode was written off as a “Japanese thing”.

If the networks want to make some money out of applications they need to look to the quality of what they are selling and understand that the developer needs to make the bulk of the money to invest in future projects.

Without that we are stuck with a model less like the Billy Goat Tavern and more like Pizza delivery. The cost of the ingredients is dwarfed by the costs of marketing, margin, paying the delivery driver and high street frontages.

That’s fine for a quick snack but it’s not what the developers need if they want to sell enough to be able to afford to eat anything better than moped pizza.

Cat Keynes publishes her thoughts on the mobile phone industry every Sunday at www.catkeynes.com you can read the column  the previous Friday by subscribing here. Follow me on Twitter here.

<< Previous Sunday's



[Home] [Archive] [Subscribe] [Advertise] [About Me] [Contact Me]