I’ve been introduced to a blog written by Benedict Evans who works at Andreessen Horowitz. And in particular I’ve been reflecting on a recent piece he wrote about the next phase of the mobile internet. Here’s the link: http://ben-evans.com/benedictevans/2014/11/20/time-for-new-questions-in-mobile
It’s a thought-provoking piece which starts to open some doors to where the next phase of the development online is coming from.
There has already been a transition in the internet through two ‘states’ of power:
At first it was the telco who controlled access to what at the time was a limited and ‘luxurious’ environment of online information. They effectively acted as the ‘control-layer’. The world was focused on pure forms of communication, message boards, email and very simple content forms. And that spawned Netscape and AOL.
But what we’ve seen over the last 10 years is a gradual shift of power to the ‘utility’ layer as access became ever more ubiquitous (even legislated), content has become more prevalent and a new breed of utility provider was needed to help audiences navigate through a plethora of websites. There have been further developments too, the biggest of which being ‘social’, but fundamentally it is the utility players who today own the customer relationship – the “aggregators” and “discovery engines”.
The third ‘state’ of the internet is going to be a real power struggle, and in my book it’s all going to revolve around data and who owns that data. We’re already starting to see the first moves – the Government’s smart metering programme, Facebook’s tightening of it’s data rules, roll out of wearable devices, Google Street View, you might even put ‘social’ in there, etc.
But in my mind there are three key phases to move through, and today we’re only a few yards out of the starting blocks.
- Data collection – today there is very little data that is actually collected, properly tagged/structured and stored. Some industries are better than others (for legacy reasons) but there are huge variances in the quality and volume of data collected. So in the first phase of this next evolution in the internet, I see a land grab for data sets across the board – personal data, financials data, logistics data, etc. Hence the focus on wearable devices, so the providers of those services can own the datasets they yield. And we’re already starting to see evidence of this in the valuations of early stage businesses who convince investors of the perceived future value of the data they are collecting.
- Interpretation and understanding – the problem with (1) is no one actually knows the value of the data collected until someone comes up with an algorithm to interpret, draw conclusions and action upon what the data is saying. Here’s where the next phase of development comes in. There’s very little today actually doing it – most analytics businesses create a fancy dashboard and defer the ‘interpretation’ component to the end user who has to come up with the answer themselves. But as these algorithms develop, we’ll see a shift in value as data interpretation yields greater levels of automation.
- Emergence of the data superpowers – only once we’ve got to a point where the data collected in (1) and interpreted in (2) has delivered real value, will we know which of the data owners will have succeeded. For example, would it be helpful for Philips to install a chip in every lightbulb it sells? On the face of it probably not, until an application is built using the data to inform security systems, insurance companies, the power grid, or whatever. It’s a bit like the move we’ve just been through in mobile: first you build the playing field; Nokia had a go, Vodafone had a go, then Apple came along, and Google; then you fill it with apps, test out the merits of each approach; and finally the developers and consumers work out which ecosystem they like best. In this case it might be because the data is all-encompassing, because it is better structured, because it’s collection is more accurate, or just simply because there’s just more of it. But because that dataset is so fundamentally important to the development of the resultant application layer, the owner of the data assumes a position of unassailable power, as Apple and Google have done today.
The reason I think this is such a hard one to predict is that the shift from Netscape/AOL to Apple/Google took place in the internet’s “growth phase”, both in the number of people connected to the Internet and the volume of content on it. This next shift, at least in the Western world, takes place when the Internet is nearing saturation when it typically takes fundamental disruption for new players to break through, far more than in growth markets. Couple that with greater intervention by regulators and Government, particularly when it comes to personal data (arguably is the most valuable dataset), and we’re in for quite a ride.
Fun times ahead!