Black Swan Farming

How to make billions of dollars from spare bits

Happy 20th Birthday, SMS.

Wow, haven’t you grown?!  Sure, you’re still only 160 characters long, but none of us would have guessed 20 years ago how much a part of our lives you would become – the most widely used data application in the world with 3.6 billion active users.  By utilising a bit of spare space (in the control channel used to send information about cell strength and signal incoming calls) we’ve been able to communicate with each other in a whole new way, even spawning a new vowel-challenged lexicon to make you a little faster, eke a bit more from you.  Last year we sent 7.8 trillion messages.  This year it’s predicted that we will send nearly 10 trillion messages – thanks to you.  That’s over 300,000 per second!

You were first a twinkle in the eye at a pizza joint in Copenhagen.  When you were eventually born, on 3rd December 1992, we wanted the best for you, but we didn’t expect you to change the world quite like you have.  Nor did we realise then how incredibly valuable you would become!  By next year, worldwide revenues from SMS are forecast to break the USD 150 billion per annum mark. What’s truly astonishing is that billions of mobile phone subscribers are willing to pay more for you per kB than it costs to transmit data from Mars to Earth.

The thing is, whilst you’ve grown into one of the largest asymmetric payoffs the world has seen – where a relatively small investment has a chance of a huge upside – features a like you abound in product development.  I spoke at XPDay in London last week about how we can go hunting for “Black Swan” style features like you by using economics and Cost of Delay at the feature level.  The results I shared show that the top 25% of features are 1000 times more valuable than the bottom 25%.  This pattern repeats for different applications and different demand streams across a portfolio of development worth £100m each year.

What this means is that, just like you, a relatively small number of features deliver incredible value in amongst a sea of relatively low value (but still profitable) features.  Given this, you’d think we might put a bit more effort into working out at least a rough prediction of the value of the features we’re thinking of building.  Sadly, most product development pipelines abound with large batches that hide and conceal these diamonds.  Combine this with a prioritisation method based on MoSCoW wish lists or the “HiPPO“, and no wonder we deliver such a high proportion of “requirements” no one actually uses.

We must not forget that value estimates are merely predictions at this point.  These are ideas that have been evaluated and triaged based on expected economic outcome before they are refined and realised.  As we know from you, my dearest SMS, these things are often not predictable.  But surely that makes it even more important to at least attempt to separate the wheat from the chaff in the first instance – but then go further and shorten the lead-time to delivery, speeding up the feedback loop about what actually works and what doesn’t?

Why not improve our chances of discovering the next billion-dollar idea that not only changes our lives, but the fortunes of a whole industry?

Merry Christmas, SMS.