The Internet; Past, Trends, Predictions
Mary Meeker's presentation offers a wealth of information and viewpoints; some of the most interesting center around mobility.
We collect enough data on traffic and usage on the Internet that trends are evident and predictions possible. As we look to the future we are not always right--consider some of the dying social media sites. However, looking back and forward is still worth the effort.
"Internet Trends@Stanford-Bases", a presentation by Mary Meeker of the venture capital firm Kleiner Perkins Caulfield & Byers does just that--looks backward and forward. There are 88 slides in her presentation--here is my take on a few of the highlights.
Meeker used data from many sources, including the International Telecommunication Union (ITU). Her slide, "Global Fixed Telephone Lines vs. Mobile Subscriptions, 1994-2009" reports from the ITU data that landlines peaked at 1.3 billion in 2006; by 2002, mobile subscriptions had caught up to landlines. An interesting fact is that the growth of mobile devices has slowed, and while the number of landlines has dropped, the decline has only been modest, not quite the death that was predicted by some pundits.
Another chart also based on ITU data looked at worldwide Internet growth--2.4 billion users in 2012, with an average growth of 8% year–to-year (Y/Y).
China has the highest number of Internet users but not the highest population penetration. The USA has the highest penetration rate, with Argentina close at 68% penetration. China is one of the lowest penetrations for the list of 15 countries reported. Iran had the highest Y/Y growth rate at 205%, with Indonesia at 58% and Argentina at 57% Y/Y growth.
The crossover point for desktop vs. mobile Internet users in India appears to have occurred this past May. Meeker's chart, "India Internet Traffic by Type, Desktop vs. Mobile, 12/08- 5/12," based on StatCounter data, shows a trend that many other countries will follow. Mobile Internet traffic will eventually dominate the market as a whole.
A disadvantage of this trend, at least from the businesses' point of view, is the Average Revenue per User (ARPU). The ARPU for desktop transactions is 1.7 to 5 times higher than for mobile transactions. This may be due to longer time on site for the desktop, or perhaps the mobile user wants to complete one or a few transactions and move on vs. longer onsite times for the desktop user. Meeker pointed out that for Google and Facebook, mobile traffic is boosted by the number of clicks, thereby reducing the cost of a click, which in turn limits the revenue attained.
Another of Meeker's slides, "Global Digital Information Created & Shared, 2005-2015E", based on IDC data, presented the growth of information volume. For 2012, about 3,800 zettabytes of information were created and shared. It is estimated that this will grow to about 8,000 zettabytes in 2015. (A zettabyte is 1 billion terabytes.) This information includes documents, tweets, pictures, personal profiles, and anything you could create.
A major portion of the presentation focused on "Re-Imagination". Meeker compared many "Then and Now" topics and what has changed for computing devices, connectivity, news delivery, note taking, drawing, photography, publications, music listening, video viewing and many many other trends.
The magnitude of the coming changes "will be stunning," according to Meeker. These include:
* "Nearly ubiquitous high speed wireless access in developed countries"
* "Unprecedented global technology innovation"
* "Fearless (& connected) entrepreneurs"
* "Available (& experienced) capital"
* "Fearless (& connected) consumer"
* "Inexpensive devices/access/services (apps)"
* "Ability to reach millions of new users in record (& accelerating) time"
These predictions may all come true. However the growth in poorer countries will depend on reliable and accessible power to charge the wireless devices. If I were to invest in any part of the mobile growth, I would invest in battery companies. Without reliable and long-lasting batteries, most of this growth will be limited.