One thing interesting #3: Re-framing problems

A company like Airbnb could be described in many different ways; a hotel chain, a marketplace for spare rooms, or more recently, a travel agent. It depends on how you look at them.

Another less likely description would be a 'data processing and communications' company. Their business is clearly more specific than this. Yet Airbnb is built on the these fundamental inputs.

Advancements in IT and communications have been so significant such that it has made it possible to build a hotel chain in a completely different way. Unlike their predecessors, owning the infrastructure was not necessary for Airbnb. It could be taken off the balance sheet. Existing infrastructure - i.e. spare rooms - could be put to use if individuals outside the company could play a role in providing the service.

This made Airbnb's idea possible but it's not the reason for their success. They've succeeded by fostering a loyal community of hosts in order to provide a great customer experience. A combination of great brand design, a compelling value proposition and powerful network effects.

But without the underlying technology being in place, it wouldn't have been possible for Airbnb's founders to even start.

We're in a phase with computing and the internet where all the obvious applications have been built. Any industry that had arithmetic as a fundamental input (e.g. data analysis, finance, accounting) very quickly felt the effects of computers. Similarly, any industry that succeeded by owning some form of communication channel (e.g. journalism, media distribution, personal communication) very much felt the effects of the internet.

After those initial applications; and once we've had chance to understand the technology and have it permeate our lives, we find less obvious ways to apply it. This can be witnessed with the most recent wave of 'internet technology companies' and the industries they are entering. For industry incumbents, the disruptive market entrants are more likely to come from an unexpected place.

Which is why it would have been difficult for established hotel chains to have seen Airbnb coming.

Providing accommodation wasn't viewed as a 'communication' problem before the internet; but Airbnb re-framed it as one.

For Hilton, IHG and Marriott, winning involved expertise in real estate and brand differentiation; not software and marketplace building. Advancements in communication technology meant the game could be changed. So dramatically in fact that a fundamental input to providing a hotel service could be conducted by individuals outside of the organisation. Before the internet these 'coordination costs' would've simply been too high. An unthinkable business model for market incumbents.

Looking ahead, we're at the precipice of a similar technological revolution with machine intelligence. A moment that perhaps mirrors that of early 90s internet. The technology sounds promising but we're not quite sure how it will materialise and enrich our lives.

I really like how this article from Joshua Gans frames these possibilities of machine intelligence. 

If computing and the internet impacted human abilities related to data and communication, machine intelligence may have a similar impact on anything where human prediction is a fundamental activity. 

The obvious applications are emerging today. Predicting products you might want to buy on Amazon. Predicting where to send the Uber driver based on likely location of the next passenger. Predicting what you might want to read next time you open Facebook.

What will be interesting to observe will be the applications that emerge because they've re-framed a particular industry or category as being a "prediction" problem rather than how it's framed today. Similar to how Airbnb reframed the way to provide accommodation and hospitality because of better data and communication abilities. 

Assuming Gans is correct in his framing, it will be interesting to observe which industries could be turned on their head because we gained superpowers in our ability to predict.