After an initial career as a stockbroker with Merrill Lynch in the early 1980s, Mary Meeker has been a technology analyst, starting with work at Salomon Brothers in 1986. In 1991, she moved to Morgan Stanley, where over a period of 20 years, she was the lead research analyst on a set of remarkable big-picture reports on the growth of both the Internet and mobile computing. In December 2010, she left Morgan Stanley to become a partner in the Silicon Valley venture capital firm Kleiner Perkins Caufield & Byers (KPCB), where she participates in technology deals and continues to write one or two industry reports every year.
Given her long view of the ICT industry worldwide, this is a person to pay attention to whenever she publishes a report or gives a major speech. However, that doesn’t mean that everything she says is right, or even well thought out. But, because Meeker gets noticed, people tend to pay attention to what she says even though some of what she delivers is already “old news” or based on little evidence.
Meeker’s latest report, KPCB Internet Trends 2013 (written with colleague Liang Wu), is a case in point. While it contains useful and thought-provoking data and analysis, it should not be taken as gospel truth or the freshest thinking on the topics she raises.
Presented to the D11 Conference on Internet Trends on May 29, 2013, the highlights of this KPCB report is available as a deck on SlideShare. If you look at the reports from KPCB on this topic from previous years, you will quickly see that many of the slides are updates of previous slides, making the conclusions less startling. Nevertheless, this presentation has many interesting points for the mobile learning community.
As Meeker and Wu note, while there are well over 6 billion mobile phone subscribers in the world, there are a billion or so less mobile phone users, because many people have multiple phones. The International Telecommunications Union (ITU) reports that “there are over 100 countries with mobile-cellular penetration exceeding 100% (meaning there are more mobile cellular subscriptions than inhabitants), and in 7 economies worldwide, mobile cellular penetration is over 200%.” In the KPCB report, mobile is described as having “aggressive momentum” with tremendous upside potential. But, on the hardware side, much of that potential growth is in developing countries, particularly in Asia.
One controversy that has arisen from this report is from Slide 52, where Meeker and Wu state that mobile users reach to phone around 150 times a day. Jeff Elder of the San Francisco Chronicle questions this statistic and reports as follows:
But there was never any solid data to back it up. The numbers cited in the presentation were taken from opinions posted on a blog. And the blog post wasn’t even discussing smartphones, but non-smartphones… The person Meeker listed as the source for the slide, mobile consultant Tomi Ahonen, says he never saw any data to back up the 150 times-a-day number. It was a figure he had heard cited at conferences since 2010. The figures presented in Meeker’s bar chart were “definitely only opinion,” and “not intended in any way as a ‘study’ of consumer behavior”… Asked about the discrepancies, Meeker’s team at Kleiner Perkins argued that the slide was correct overall. They said data from a key secondary source – Danielle Levitas of International Data Corp. – backed up the information Meeker presented, but that they didn’t have rights to use it at the time of the presentation or they would have listed it as a source. Levitas disagreed. Contacted by The Chronicle, she said her data, which tracked social media use on smartphones, could not be used to back up the 150 times-a-day statistic in any reasonable way. She also said her data had been public since March, so anyone could have used it in a presentation.
Four Data Types
On Slide 12, Meeker and Wu show a succession of four types of digital files from mobile devices:
- Photos – explosive growth, but still early stage
- Video – ramping very fast
- Sound – emerging
- Data – emerging
I think that these are in the wrong order, and all are well underway.
Data collection and manipulation, especially as numbers and text, has been there from the beginning of computing, including with mobile devices such as calculators, personal data assistants (PDAs), and early phones equipped with SMS.
What is new is the fact we are storing so much data, so that over time we can analyze patterns, and that we have started the process of “datafication” (as opposed to simple digitization) of much of the world’s store of information. New “Big Data” applications are emerging for sure, but the collection of data using mobile devices goes back to their beginnings.
Similarly, the sound has been a mobile application from the beginning of the use of mobile phones, but especially with the iPod and MP3 applications we call “podcasting.” That revolution has come and already faded into the background.
Photos have been there ever since digital cameras were invented, and with cheap storage, could be uploaded to the web. Flickr.com and Facebook have billions of photos stored on their servers, at no cost to users.
The video is the one media that is recently ramping up fast, again because of cheap storage with services such as YouTube, and because of the spread of wired and wireless broadband services. What Meeker and Wu correctly identify is the emergence of new players such as Instagram and Snapchat into these exploding markets.
“I Won’t Tell You That!”
Slide 28, titled “Americans = Sharing Underachievers” is also problematic for me. It suggests that sharing is good, without describing what is being shared. If it is filed, that is one thing…but, if it is deeply personal information and feelings, then this graph may be reflective of cultural differences in terms of the value of privacy. I find this slide to be too ambiguous to be useful.
Extraordinary Attributes of Smartphone Users
Slide 31 introduces us to a new and emerging field of analytics – sentiment analysis. This assesses the moods and feelings of groups of people, often to predict what they are going to do.
Here we learn that smartphone users feel connected (7 out of 10), excited (4 out of 10), curious/interested (2 out of 10), and productive (2 out of 10). Except for feeling connected, these are not strong results when 0 means weakest and 10 means strongest.
Yet, because these four feelings rise to at least 2 out of 10 (all the other listed feelings are weaker), the slide calls them “extraordinary attributes.”
I don’t think so…
How We Access Information
Slide 32 tells us that mobile traffic has been steadily increasing from just less than 1% in May 2009 to 15% in May 2013.
While this is encouraging for those of us in the mobile learning industry, we need to remind ourselves that 85% of the Internet traffic today is not mobile, which may explain why the adoption of mLearning going slower than we may like.
While the raw number of users accessing the Web in China or other countries may be increased to surpass the number of users accessing the Web with a desktop PC, users may be spending much more time for each session on the Web compared with the in-and-out behavior of accessing information on a mobile phone.
I was really struck by the shift in smartphone market share from Q1:2010 to Q4:2012. Now there is a shift. Just take a look:
For me, the most dramatic slide in the deck was number 45 showing the truly explosive growth of tablets in such a short period of time. Here it is:
Why did this happen?
My own hunch is that tablets are a compromise technology – light enough to be portable, but not pocketable, and big enough to have a useful screen for presentations. Tablets, in my view, are a “getting out of jail card” for those who are not comfortable using smaller phone screens to interact with information. As I argued recently, new display technologies, such as micro-projection and flexible rollout screens may make tablets in their present form obsolete within five years.
Gaps Between Devices Will Decrease
I also disagree with the contention in the report that we are in two computing cycles – 1) smartphones and 2) tablets, and that we are “entering the third cycle”…3) Wearables/Drivables/Flyables/Scannables.
First, as I noted, tablets are a transitional technology, that will likely disappear as a separate category of mobile devices. We already see the gaps being filled in between laptops and tablets with “ultralights,” and the gap between tablets and smartphones with “phablets” such as the Samsung Note. There is now a continuum of devices of all sizes and configurations of features.
In any case, we at Float have never defined mobile learning as being about the device, but about the fact that learners are now mobile, compared with the immobility of the classroom and the desktop PC.
The category of Wearables/Drivables/Flyables/Scannable is not a single cycle of technology, but each of these terms describes a different arc of development.
Wearables go way back to the early 1980s with the pioneering work of Steve Mann at the University of Toronto – it is not even clear that most people will accept them when simply carrying a mobile phone on your belt will fill most computing and information needs in the near future.
Drivables are autonomous automobiles pioneered in the 1980s by Mercedes-Benz and Carnegie Mellon University and most recently developed by Google.
Flyables refer to remote-controlled drones that have been fitted with cameras.
Scannables are readable optical tags such as QR codes (invented in 1994) and radio frequency identification tags (invented in the 1970s) that can be used to convey information by being placed in an environment and scanned by devices with appropriate software. Meeker and Wu throw sensors into the category of scannables, but, in truth, there are many different types of sensors, most of which don’t involve scanning, which has their own developmental paths.
More importantly, I believe that Meeker and Wu miss the biggest technology change of all – the information tsunami that has taken place in the last 10 years, and the set of concepts, tools, and methods that are emerging under the category of “big data.”
Specifically, geospatial data from mobile phones is emerging as a special kind of data that links many other kinds of data together, integrating them into a holistic view of what a person is doing or has done. This built-in feature of mobile devices – the generation of geospatial data – is transforming learning in ways we are just starting to imagine. In 2012, Jeff Jonas, distinguished engineer and chief scientist at IBM described geospatial data as “analytic superfood.” “Geospatial data,” he says, ” is going to rip the lid off what’s computable.”
Because we are collecting new types of data and huge amounts of data, we are also evolving new approaches to measurement and analytics using this data. (This is what I am speaking about at mLearnCon 2013).
We are also starting to see that learning is not about memorizing a cannon of required knowledge based on curriculum or business objectives, but that it is about people becoming problem-solvers and answer-seekers based on the contextual needs of the environment or circumstances in which they are situated.
Meeker and Wu list the top 15 learning tools from Jane Hart’s list Top 100 Tools for Learning in Slide 108. Here they are:
All of the above tools can be built into or accessed by a mobile learning app, and we have already used several in Float apps. Our newest app (to be released soon) is a geolocational app called Wayfiler that alerts users to the presence of specific files in Dropbox when they reach a specific location.
The good news in this report for the mobile learning industry is that online learning, in general, is growing rapidly and that by 2012, 32% of post-secondary students were taking at least one online course (Meeker and Wu only have up to 2011 in Slide 99). As well, acceptability among academic leaders is also improving, as shown in the graph from Slide 100.
The KPCB report on Internet trends has both strengths and weaknesses. But, it’s an important document that should be reviewed by everyone in the mobile learning industry.
Latest posts by Gary Woodill (see all)
- Making the Business Case for a New Learning Technology - July 1, 2019
- Rapid Doubling of Knowledge Drives Change in How We Learn - January 23, 2018
- What Does AR for Learning Enable? - January 19, 2018