The great decline in Wikipedia pageviews (full version)

A shorter version of this post is available on LessWrong. The LessWrong post will remain static whereas this version will be updated with more examples as they occur to me.

Last year, during the months of June and July, as my work for MIRI was wrapping up and I hadn’t started my full-time job, I worked on the Wikipedia Views website, aimed at easier tabulation of the pageviews for multiple Wikipedia pages over several months and years. It relies on a statistics tool called, created by Domas Mituzas, and maintained by Henrik.

One of the interesting things I noted as I tabulated pageviews for many different pages was that the pageview counts for many already popular pages were in decline. Pages of various kinds peaked at different historical points. For instance, colors have been in decline since early 2013. The world’s most populous countries have been in decline since as far back as 2010!

Defining the problem

The first thing to be clear about is what these pageviews count and what they don’t. The pageview measures are taken from, which in turn uses the pagecounts-raw dump provided hourly by the Wikimedia Foundation’s Analytics team, which in turn is obtained by processing raw user activity logs. The pagecounts-raw measure is flawed in two ways:

  • It only counts pageviews on the main Wikipedia website and not pageviews on the mobile Wikipedia website or through Wikipedia Zero (a pared down version of the mobile site that some carriers offer at zero bandwidth costs to their customers, particularly in developing countries). To remedy these problems, a new dump called pagecounts-all-sites was introduced in September 2014. We simply don’t have data for views of mobile domains or of Wikipedia Zero at the level of individual pages for before then. Moreover, still uses pagecounts-raw (this was pointed to me in a mailing list message after I circulated an early version of the post).
  • The pageview count includes views by bots. The official estimate is that about 15% of pageviews are due to bots. However, the percentage is likely higher for pages with fewer overall pageviews, because bots have a minimum crawling frequency. So every page might have at least 3 bot crawls a day, resulting in a minimum of 90 bot pageviews even if there are only a handful of human pageviews.

Therefore, the trends I discuss will refer to trends in total pageviews for the main Wikipedia website, including page requests by bots, but excluding visits to mobile domains. Note that visits from mobile devices to the main site will be included, but mobile devices are by default redirected to the mobile site.

How reliable are the metrics?

As noted above, the metrics are unreliable because of the bot problem and the issue of counting only non-mobile traffic. German Wikipedia user Atlasowa left a message on my talk page pointing me to an email thread suggesting that about 40% of pageviews may be bot-related, and discussing some interesting examples.

Relationship with the overall numbers

I’ll show that for many pages of interest, the number of pageviews as measured above (non-mobile) has declined recently, with a clear decline from 2013 to 2014. What about the total?

We have overall numbers for non-mobile, mobile, and combined. The combined number has largely held steady, whereas the non-mobile number has declined and the mobile number has risen.

What we’ll find is that the decline for most pages that have been around for a while is even sharper than the overall decline. One reason overall pageviews haven’t declined so fast is the creation of new pages. To give an idea, non-mobile traffic dropped by about 1/3 from January 2013 to December 2014, but for many leading categories of pages, traffic dropped by about 1/2-2/3.

Why is this important? First reason: better context for understanding trends for individual pages

People’s behavior on Wikipedia is a barometer of what they’re interested in learning about. An analysis of trends in the views of pages can provide an important window into how people’s curiosity, and the way they satisfy this curiosity, is evolving. To take an example, some people have proposed using Wikipedia pageview trends to predict flu outbreaks. I myself have tried to use relative Wikipedia pageview counts to gauge changing interests in many topics, ranging from visa categories to technology companies.

My initial interest in pageview numbers arose because I wanted to track my own influence as a Wikipedia content creator. In fact, that was my original motivation with creating Wikipedia Views. (You can see more information about my Wikipedia content contributions on my site page about Wikipedia).

Now, when doing this sort of analysis for individual pages, one needs to account for, and control for, overall trends in the views of Wikipedia pages that are occurring for reasons other than a change in people’s intrinsic interest in the subject. Otherwise, we might falsely conclude from a pageview count decline that a topic is falling in popularity, whereas what’s really happening is an overall decline in the use of (the non-mobile version of) Wikipedia to satisfy one’s curiosity about the topic.

Why is this important? Second reason: a better understanding of the overall size and growth of the Internet.

Wikipedia has been relatively mature and has had the top spot as an information source for at least the last six years. Moreover, unlike almost all other top websites, Wikipedia doesn’t try hard to market or optimize itself, so trends in it reflect a relatively untarnished view of how the Internet and the World Wide Web as a whole are growing, independent of deliberate efforts to manipulate and doctor metrics.

The case of colors

Let’s look at Wikipedia pages on some of the most viewed colors (I’ve removed the 2015 and 2007 columns because we don’t have the entirety of these years). Colors are interesting because the degree of human interest in colors in general, and in particular colors, is unlikely to change much in response to news or current events. So one would at least a priori expect colors to offer a perspective into Wikipedia trends with fewer external complicating factors. If we see a clear decline here, then that’s strong evidence in favor of a genuine decline.

I’ve restricted attention to a small subset of the colors, that includes the most common ones but isn’t comprehensive. But it should be enough to get a sense of the trends. And you can add in your own colors and check that the trends hold up.

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
Black 431K 1.5M 1.3M 778K 900K 1M 958K 6.9M 16.1 Colors
Blue 710K 1.3M 1M 987K 1.2M 1.2M 1.1M 7.6M 17.8 Colors
Brown 192K 284K 318K 292K 308K 300K 277K 2M 4.6 Colors
Green 422K 844K 779K 707K 882K 885K 733K 5.3M 12.3 Colors
Orange 133K 181K 251K 259K 275K 313K 318K 1.7M 4 Colors
Purple 524K 906K 847K 895K 865K 841K 592K 5.5M 12.8 Colors
Red 568K 797K 912K 1M 1.1M 873K 938K 6.2M 14.6 Colors
Violet 56K 96K 75K 77K 69K 71K 65K 509K 1.2 Colors
White 301K 795K 615K 545K 788K 575K 581K 4.2M 9.8 Colors
Yellow 304K 424K 453K 433K 452K 427K 398K 2.9M 6.8 Colors
Total 3.6M 7.1M 6.6M 6M 6.9M 6.5M 6M 43M 100
Percentage 8.5 16.7 15.4 14 16 15.3 14 100

Since the decline appears to have happened between 2013 and 2014, let’s examine the 24 months from January 2013 to December 2014:

Month Views of page Black Views of page Blue Views of page Brown Views of page Green Views of page Orange Views of page Purple Views of page Red Views of page Violet Views of page White Views of page Yellow Total Percentage
201412 30K 41K 14K 27K 9.6K 28K 67K 3.1K 21K 19K 260K 2.4
201411 36K 46K 15K 31K 10K 35K 50K 3.7K 23K 22K 273K 2.5
201410 37K 52K 16K 34K 10K 34K 51K 4.5K 25K 26K 289K 2.7
201409 37K 57K 16K 35K 9.9K 37K 45K 4.8K 27K 29K 298K 2.8
201408 33K 47K 14K 34K 8.5K 31K 38K 3.9K 21K 22K 253K 2.4
201407 33K 47K 14K 30K 9.3K 31K 37K 4.2K 22K 22K 250K 2.3
201406 32K 49K 14K 31K 10K 34K 39K 4.9K 23K 22K 259K 2.4
201405 44K 55K 17K 37K 10K 51K 42K 5.2K 26K 26K 314K 2.9
201404 34K 60K 17K 36K 14K 38K 47K 5.8K 27K 28K 306K 2.8
201403 37K 136K 19K 51K 14K 123K 52K 5.5K 30K 31K 497K 4.6
201402 38K 58K 19K 39K 13K 41K 49K 5.6K 29K 29K 321K 3
201401 40K 60K 19K 36K 14K 40K 50K 4.4K 27K 28K 319K 3
201312 62K 67K 17K 44K 12K 48K 48K 4.4K 42K 26K 372K 3.5
201311 141K 96K 20K 65K 11K 68K 55K 5.3K 71K 34K 566K 5.3
201310 145K 102K 21K 69K 11K 77K 59K 5.7K 71K 36K 598K 5.6
201309 98K 80K 17K 60K 11K 53K 51K 4.9K 45K 30K 450K 4.2
201308 109K 87K 20K 57K 20K 57K 60K 4.6K 53K 28K 497K 4.6
201307 107K 92K 21K 61K 11K 66K 65K 4.6K 61K 30K 520K 4.8
201306 115K 106K 22K 69K 13K 73K 64K 5.5K 70K 33K 571K 5.3
201305 158K 122K 24K 79K 14K 83K 69K 11K 77K 39K 677K 6.3
201304 151K 127K 28K 83K 14K 86K 74K 12K 78K 40K 694K 6.4
201303 155K 135K 31K 92K 15K 99K 84K 12K 80K 43K 746K 6.9
201302 152K 131K 31K 84K 28K 95K 84K 17K 77K 41K 740K 6.9
201301 129K 126K 32K 81K 19K 99K 84K 9.6K 70K 42K 691K 6.4
Total 2M 2M 476K 1.3M 314K 1.4M 1.4M 152K 1.1M 728K 11M 100
Percentage 18.1 18.4 4.4 11.8 2.9 13.3 12.7 1.4 10.2 6.8 100
Tags Colors Colors Colors Colors Colors Colors Colors Colors Colors Colors

As we can see, the decline appears to have begun around March 2013 and then continued steadily till about June 2014, at which numbers stabilized to their lower levels.

A few sanity checks on these numbers:

  • The trends appear to be similar for different colors, with the notable difference that the proportional drop was higher for the more viewed color pages. Thus, for instance, black and blue saw declines from 129K and 126K to 30K and 41K respectively (factors of four and three respectively) from January 2013 to December 2014. Orange and yellow, on the other hand, dropped by factors of close to two. The only color that didn’t drop significantly was red (it dropped from 84K to 67K, as opposed to factors of two or more for other colors), but this seems to have been partly due to an unusually large amount of traffic in the end of 2014. The trend even for red seems to suggest a drop similar to that for orange.
  • The overall proportion of views for different colors comports with our overall knowledge of people’s color preferences: blue is overall a favorite color, and this is reflected in its getting the top spot with respect to pageviews.
  • The pageview decline followed a relatively steady trend, with the exception of some unusual seasonal fluctuation (including an increase in October and November 2013).

One hypothesis that some people might come up with is inter-language substitution: people are substituting away from reading articles in the English-language Wikipedia to other language Wikipedias. But the downward trend is present in many other major language Wikipedias, none of which have the second language status of English.

Here are the numbers for four colors in Spanish (negro = black, azul = blue, rojo = red, blanco = white) for the years 2009-2014 (we exclude 2008 because tracking for the Spanish Wikipedia began only in February 2008):

Year Views of page Negro (color) Views of page Azul Views of page Rojo Views of page Blanco (color) Total Percentage
2014 82K 143K 137K 87K 449K 12.7
2013 123K 275K 274K 123K 795K 22.5
2012 132K 301K 208K 109K 751K 21.2
2011 122K 248K 181K 81K 633K 17.9
2010 94K 213K 142K 61K 510K 14.4
2009 79K 174K 93K 52K 398K 11.3
Total 632K 1.4M 1M 512K 3.5M 100
Percentage 17.9 38.3 29.3 14.5 100
The months of 2013 and 2014:
Month Views of page Negro (color) Views of page Azul Views of page Rojo Views of page Blanco (color) Total Percentage
201412 4.7K 7.5K 7K 4.9K 24K 1.9
201411 7K 9.9K 18K 6.6K 41K 3.3
201410 7.5K 12K 12K 7.8K 39K 3.2
201409 7.7K 13K 13K 8.6K 42K 3.4
201408 6.3K 11K 9K 6.4K 32K 2.6
201407 5.9K 10K 9.9K 6K 32K 2.6
201406 6.7K 12K 11K 7K 37K 2.9
201405 8.2K 14K 13K 8.4K 44K 3.5
201404 7.2K 13K 12K 7.6K 40K 3.2
201403 8.2K 15K 13K 9K 45K 3.6
201402 7.1K 14K 12K 8.4K 41K 3.3
201401 5.3K 11K 9K 6.1K 31K 2.5
201312 5.2K 13K 12K 6.4K 36K 2.9
201311 9.2K 23K 23K 12K 67K 5.4
201310 10K 25K 25K 13K 73K 5.9
201309 8.3K 17K 17K 9.3K 51K 4.1
201308 8.9K 18K 18K 9.6K 55K 4.4
201307 9.9K 20K 21K 10K 61K 4.9
201306 12K 25K 25K 12K 74K 5.9
201305 14K 29K 29K 12K 85K 6.8
201304 13K 29K 30K 3K 75K 6
201303 12K 26K 27K 11K 76K 6.1
201302 11K 26K 26K 13K 76K 6.1
201301 9.9K 25K 21K 10K 66K 5.3
Total 204K 418K 412K 210K 1.2M 100
Percentage 16.4 33.6 33.1 16.9 100

Similarly, here are the pageview counts for the same four colors in French (noir = black, bleu = blue, rouge = red, blanc = white):

The years 2009-2014:

Year Views of page Noir Views of page Bleu Views of page Rouge Views of page Blanc Total Percentage
2014 65K 108K 86K 68K 327K 10.7
2013 157K 184K 125K 110K 576K 18.9
2012 149K 209K 169K 103K 630K 20.6
2011 103K 194K 143K 85K 525K 17.2
2010 116K 195K 122K 100K 534K 17.5
2009 126K 168K 81K 84K 458K 15
Total 716K 1.1M 725K 550K 3M 100
Percentage 23.5 34.7 23.8 18 100

All months in 2013 and 2014:

Month Views of page Noir Views of page Bleu Views of page Rouge Views of page Blanc Total Percentage
201412 4.4K 6.9K 6.6K 4.8K 23K 2.5
201411 4.9K 8.3K 6.8K 5.2K 25K 2.8
201410 7K 9.4K 6.9K 6.6K 30K 3.3
201409 6.5K 9.5K 6.8K 5.9K 29K 3.2
201408 3.9K 6.5K 4.6K 4.1K 19K 2.1
201407 4.3K 7.3K 5.3K 4.6K 22K 2.4
201406 4.4K 8.2K 5.8K 4.6K 23K 2.5
201405 5.9K 9.9K 7.8K 6.9K 31K 3.4
201404 5.6K 10K 7.7K 6.1K 30K 3.3
201403 6.4K 11K 8.9K 6.5K 33K 3.6
201402 5.9K 10K 9.1K 6K 31K 3.5
201401 6K 11K 9.4K 6.7K 33K 3.7
201312 7.1K 11K 9.3K 7.3K 35K 3.9
201311 13K 15K 12K 10K 50K 5.6
201310 12K 15K 11K 9K 47K 5.2
201309 7.9K 11K 7.8K 6.4K 33K 3.7
201308 9.2K 11K 7.1K 5.8K 33K 3.7
201307 11K 12K 7.8K 6.9K 37K 4.1
201306 15K 16K 10K 9.3K 50K 5.5
201305 17K 19K 11K 11K 58K 6.4
201304 16K 19K 7.3K 10K 52K 5.8
201303 17K 20K 14K 11K 62K 6.8
201302 16K 17K 13K 11K 57K 6.3
201301 16K 19K 15K 12K 62K 6.8
Total 222K 293K 210K 178K 903K 100
Percentage 24.6 32.4 23.3 19.7 100

Cognitive biases (a small subset thereof)

For a more niche but still timeless set of pages, I picked cognitive biases:

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
Accentuation effect 3.8K 3.4K 1.3K 0 1 4 2 8.5K 0.2 Cognitive biases
Attentional bias 28K 28K 21K 16K 8.9K 7.1K 5.1K 113K 3.1 Cognitive biases
Availability heuristic 174K 199K 194K 180K 125K 106K 85K 1.1M 29.4 Cognitive biases
Bandwagon effect 173K 258K 231K 196K 217K 189K 136K 1.4M 38.7 Cognitive biases
Base rate fallacy 79K 79K 67K 45K 43K 43K 38K 394K 10.9 Cognitive biases
Choice-supportive bias 23K 25K 18K 14K 13K 11K 9.6K 114K 3.1 Cognitive biases
Clustering illusion 30K 26K 28K 28K 33K 40K 40K 226K 6.3 Cognitive biases
Conjunction fallacy 57K 59K 57K 39K 33K 29K 24K 298K 8.2 Cognitive biases
Total 567K 677K 617K 519K 473K 426K 338K 3.6M 100
Percentage 15.7 18.7 17.1 14.4 13.1 11.8 9.3 100

The situation for cognitive biases is qualitatively similar to that for colors: the peak occurred in 2013, and the decline appears to have been between the early months of 2013 and the middle of 2014. Unlike colors, cognitive biases in 2014 still did a lot better than they had done three years ago. Note that, with one exception, all the pages selected here have been around since 2008, and the sole exception doesn’t account for enough pageviews to affect the overall trend.

Geography: continents and subcontinents, countries, and cities

Here are the views of some of the world’s most populated countries between 2008 and 2014, showing that the peak happened as far back as 2010:

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
China 5.7M 6.8M 7.8M 6.1M 6.9M 5.7M 6.1M 45M 9 Countries
India 8.8M 12M 12M 11M 14M 8.8M 7.6M 73M 14.5 Countries
United States 13M 15M 18M 18M 34M 16M 15M 129M 25.7 Countries
Indonesia 5.3M 5.2M 3.7M 3.6M 4.2M 3.1M 2.5M 28M 5.5 Countries
Brazil 4.8M 4.9M 5.3M 5.5M 7.5M 4.9M 4.3M 37M 7.4 Countries
Pakistan 2.9M 4.5M 4.4M 4.3M 5.2M 4M 3.2M 28M 5.7 Countries
Bangladesh 2.2M 2.9M 3M 2.8M 2.9M 2.2M 1.7M 18M 3.5 Countries
Russia 5.6M 5.6M 6.5M 6.8M 8.6M 5.4M 5.8M 44M 8.8 Countries
Nigeria 2.6M 2.6M 2.9M 3M 3.5M 2.6M 2M 19M 3.8 Countries
Japan 4.8M 6.4M 6.5M 8.3M 10M 7.3M 6.6M 50M 10 Countries
Mexico 3.1M 3.9M 4.3M 4.3M 5.9M 4.7M 4.5M 31M 6.1 Countries
Total 59M 69M 74M 74M 103M 65M 59M 502M 100
Percentage 11.7 13.8 14.7 14.7 20.4 12.9 11.8 100

Of these countries, China, India and the United States are the most notable. China is the world’s most populous. India has the largest population with some minimal English knowledge and legally (largely) unrestricted Internet access to Wikipedia, while the United States has the largest population with quality Internet connectivity and good English knowledge. Moreover, in China and India, Internet use and access have been growing considerably in the last few years, whereas it has been relatively stable in the United States.

It is interesting that the year with the maximum total pageview count was as far back as 2010. In fact, 2010 was so significantly better than the other years that the numbers beg for an explanation. I don’t have one, but even excluding 2010, we see a declining trend: gradual growth from 2008 to 2011, and then a symmetrically gradual decline. Both the growth trend and the decline trend are quite similar across countries.

We see a similar trend for continents and subcontinents, with a clear peak in 2010 and an otherwise roughly symmetric rise and fall:

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
Africa 2.1M 3.2M 3.6M 3.5M 3.7M 3.5M 3.3M 23M 12.8 Continents and subcontinents
Antarctica 1.6M 2.2M 2.5M 2.3M 2.5M 2.2M 1.9M 15M 8.5 Continents and subcontinents
Asia 1.6M 2.1M 2.1M 2.4M 2.6M 2M 2.1M 15M 8.3 Continents and subcontinents
Asia-Pacific 438K 509K 436K 342K 285K 203K 155K 2.4M 1.3 Continents and subcontinents
Australia 5.6M 7.5M 7.8M 8.7M 11M 7.2M 6.4M 55M 30.5 Countries Continents and subcontinents
Central America 637K 768K 754K 702K 871K 642K 626K 5M 2.8 Continents and subcontinents
East Asia 448K 503K 527K 511K 589K 419K 370K 3.4M 1.9 Continents and subcontinents
Europe 3.3M 3.7M 4.3M 4.4M 6.1M 4.1M 4M 30M 16.8 Continents and subcontinents
North America 1.7M 1.9M 2.2M 2.3M 2.9M 1.9M 1.9M 15M 8.3 Continents and subcontinents
South America 1.2M 1.7M 1.6M 1.6M 2M 1.4M 1.6M 11M 6.2 Continents and subcontinents
South Asia 635K 715K 794K 737K 775K 535K 444K 4.6M 2.6 Continents and subcontinents
Total 19M 25M 27M 27M 34M 24M 23M 179M 100
Percentage 10.8 14 14.8 15.3 18.9 13.5 12.7 100

In contrast with these large geographic entities, their smaller counterparts, such as cities, peaked in 2013, similarly to colors, and the drop, though somewhat less steep than with colors, has been quite significant. Below is a list for Indian cities:

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
Ahmedabad 416K 756K 629K 422K 359K 258K 235K 3.1M 3.3 Cities Indian cities
Ajmer 152K 223K 257K 133K 86K 62K 53K 966K 1 Indian cities
Allahabad 221K 391K 325K 209K 188K 126K 92K 1.6M 1.7 Indian cities
Ambala 74K 101K 87K 53K 41K 34K 27K 417K 0.4 Indian cities
Bangalore 1.2M 2M 1.7M 1.3M 1.1M 985K 825K 9.1M 9.7 Cities Indian cities
Bhopal 254K 355K 316K 259K 218K 159K 123K 1.7M 1.8 Indian cities
Chandigarh 425K 624K 596K 406K 341K 253K 219K 2.9M 3.1 Indian cities
Chennai 975K 1.5M 1.3M 1.1M 980K 702K 683K 7.3M 7.8 Cities Indian cities
Dehradun 210K 391K 348K 175K 136K 105K 91K 1.5M 1.6 Indian cities
Delhi 1.1M 1.5M 1.6M 1.3M 1.2M 863K 888K 8.6M 9.1 Cities Indian cities
Gandhinagar 108K 160K 134K 86K 70K 47K 34K 639K 0.7 Indian cities
Guwahati 231K 328K 279K 172K 121K 82K 69K 1.3M 1.4 Indian cities
Hyderabad 752K 394K 79K 86K 90K 120K 161K 1.7M 1.8 Cities Indian cities
Jaipur 456K 700K 693K 472K 372K 294K 269K 3.3M 3.5 Indian cities
Jamshedpur 188K 276K 258K 179K 132K 85K 69K 1.2M 1.3 Indian cities
Kanpur 223K 272K 311K 241K 175K 113K 91K 1.4M 1.5 Cities Indian cities
Kochi 308K 528K 352K 245K 96K 27K 27K 1.6M 1.7 Indian cities
Kolkata 965K 1.3M 1.2M 901K 791K 534K 482K 6.2M 6.6 Cities Indian cities
Lucknow 381K 515K 503K 376K 308K 218K 191K 2.5M 2.7 Indian cities
Mumbai 1.8M 2.6M 2.8M 2.9M 2.2M 1.9M 2.4M 17M 17.7 Cities Indian cities
Nagpur 252K 393K 364K 238K 177K 120K 112K 1.7M 1.8 Indian cities
Patna 192K 272K 270K 251K 166K 113K 95K 1.4M 1.5 Indian cities
Pondicherry 283K 497K 628K 111K 84K 103K 96K 1.8M 1.9 Indian cities
Pune 570K 1M 850K 625K 538K 382K 359K 4.4M 4.7 Cities Indian cities
Ranchi 140K 231K 186K 110K 86K 59K 48K 860K 0.9 Indian cities
Shimla 360K 510K 570K 289K 210K 155K 126K 2.2M 2.4 Indian cities
Srinagar 210K 297K 294K 149K 136K 93K 81K 1.3M 1.3 Indian cities
Surat 252K 364K 327K 232K 194K 133K 109K 1.6M 1.7 Cities Indian cities
Thiruvananthapuram 252K 435K 359K 255K 211K 132K 107K 1.8M 1.9 Indian cities
Varanasi 597K 710K 713K 457K 370K 301K 241K 3.4M 3.6 Indian cities
Total 13M 20M 18M 14M 11M 8.6M 8.4M 94M 100
Percentage 14.4 21.1 19.6 14.7 12 9.2 9 100

Some niche topics where pageviews haven’t declined

So far, we’ve looked at topics where pageviews have been declining since at least 2013, and some that peaked as far back as 2010. There are, however, many relatively niche topics where the number of pageviews has stayed roughly constant. But this stability itself is a sign of decay, because other metrics suggest that the topics have experienced tremendous growth in interest. In fact, the stability is even less impressive when we notice that it’s a result of a cancellation between slight declines in views of established pages in the genre, and traffic going to new pages.

For instance, consider some charity-related pages:

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
Against Malaria Foundation 5.9K 6.3K 4.3K 1.4K 2 0 0 18K 15.6 Charities
Development Media International 757 0 0 0 0 0 0 757 0.7 Pages created by Vipul Naik Charities
Deworm the World Initiative 2.3K 277 0 0 0 0 0 2.6K 2.3 Charities Pages created by Vipul Naik
GiveDirectly 11K 8.3K 2.6K 442 0 0 0 22K 19.2 Charities Pages created by Vipul Naik
International Council for the Control of Iodine Deficiency Disorders 1.2K 1 2 2 0 1 2 1.2K 1.1 Charities Pages created by Vipul Naik
Nothing But Nets 5.9K 6.6K 6.6K 5.1K 4.4K 4.7K 6.1K 39K 34.2 Charities
Nurse-Family Partnership 2.9K 2.8K 909 30 8 72 63 6.8K 5.9 Pages created by Vipul Naik Charities
Root Capital 3K 2.5K 414 155 51 1.2K 21 7.3K 6.3 Charities Pages created by Vipul Naik
Schistosomiasis Control Initiative 4K 2.7K 1.6K 191 0 0 0 8.5K 7.4 Charities Pages created by Vipul Naik
VillageReach 1.7K 1.9K 2.2K 2.6K 97 3 15 8.4K 7.3 Charities Pages created by Vipul Naik
Total 38K 31K 19K 9.9K 4.6K 5.9K 6.2K 115K 100
Percentage 33.4 27.3 16.3 8.6 4 5.1 5.4 100

For this particular cluster of pages, we see the totals growing robustly year-on-year. But a closer look shows that the growth isn’t that impressive. Whereas earlier, views were doubling every year from 2010 to 2013 (this was the take-off period for GiveWell and effective altruism), the growth from 2013 to 2014 was relatively small. And about half the growth from 2013 to 2014 was powered by the creation of new pages (including some pages created after the beginning of 2013, so they had more months in a mature state in 2014 than in 2013), while the other half was powered by growth in traffic to existing pages.

The data for philanthropic foundations demonstrates a fairly slow and steady growth (about 5% a year), partly due to the creation of new pages. This 5% hides a lot of variation between individual pages:

Page name Pageviews in year 2014 Pageviews in year 2013 Pageviews in year 2012 Pageviews in year 2011 Pageviews in year 2010 Pageviews in year 2009 Pageviews in year 2008 Total Percentage Tags
Atlantic Philanthropies 11K 11K 12K 10K 9.8K 8K 5.8K 67K 2.1 Philanthropic foundations
Bill & Melinda Gates Foundation 336K 353K 335K 315K 266K 240K 237K 2.1M 64.9 Philanthropic foundations
Draper Richards Kaplan Foundation 1.2K 25 9 0 0 0 0 1.2K 0 Philanthropic foundations Pages created by Vipul Naik
Ford Foundation 110K 91K 100K 90K 100K 73K 61K 625K 19.5 Philanthropic foundations
Good Ventures 9.9K 8.6K 3K 0 0 0 0 21K 0.7 Philanthropic foundations Pages created by Vipul Naik
Jasmine Social Investments 2.3K 1.8K 846 0 0 0 0 5K 0.2 Philanthropic foundations Pages created by Vipul Naik
Laura and John Arnold Foundation 3.7K 13 0 1 0 0 0 3.7K 0.1 Philanthropic foundations Pages created by Vipul Naik
Mulago Foundation 2.4K 2.3K 921 0 1 1 10 5.6K 0.2 Philanthropic foundations Pages created by Vipul Naik
Omidyar Network 26K 23K 19K 17K 19K 13K 11K 129K 4 Philanthropic foundations
Peery Foundation 1.8K 1.6K 436 0 0 0 0 3.9K 0.1 Philanthropic foundations Pages created by Vipul Naik
Robert Wood Johnson Foundation 26K 26K 26K 22K 27K 22K 17K 167K 5.2 Philanthropic foundations
Skoll Foundation 13K 11K 9.2K 7.8K 9.6K 5.8K 4.3K 60K 1.9 Philanthropic foundations
Smith Richardson Foundation 8.7K 3.5K 3.8K 3.6K 3.7K 3.5K 2.9K 30K 0.9 Philanthropic foundations
Thiel Foundation 3.6K 1.5K 1.1K 47 26 1 0 6.3K 0.2 Philanthropic foundations Pages created by Vipul Naik
Total 556K 533K 511K 466K 435K 365K 340K 3.2M 100
Percentage 17.3 16.6 15.9 14.5 13.6 11.4 10.6 100
Here’s a closer look at this same list, restricted to the months from January 2013 to December 2014:
Month Views of page Atlantic Philanthropies Views of page Bill & Melinda Gates Foundation Views of page Draper Richards Kaplan Foundation Views of page Ford Foundation Views of page Good Ventures Views of page Jasmine Social Investments Views of page Laura and John Arnold Foundation Views of page Mulago Foundation Views of page Omidyar Network Views of page Peery Foundation Views of page Robert Wood Johnson Foundation Views of page Skoll Foundation Views of page Smith Richardson Foundation Views of page Thiel Foundation Total Percentage
201412 753 25K 139 6.4K 1.7K 171 382 184 1.5K 138 1.9K 907 343 399 39K 3.6
201411 886 27K 137 7.3K 679 179 448 200 2.2K 147 1.9K 1K 629 483 43K 4
201410 821 32K 138 7.1K 910 181 437 226 1.6K 153 2.2K 1.1K 1.2K 450 48K 4.4
201409 1.3K 30K 126 7.3K 842 315 489 226 2.1K 141 2.4K 1.1K 1.4K 453 49K 4.5
201408 958 26K 122 5.6K 806 190 466 190 2K 141 1.9K 949 722 302 41K 3.7
201407 866 21K 161 6.9K 774 195 398 213 1.8K 200 2.1K 1.1K 757 267 37K 3.4
201406 856 21K 112 6.7K 714 193 371 241 2.4K 174 2.3K 1K 757 373 38K 3.5
201405 904 32K 78 9.1K 776 187 273 193 3.5K 135 2.4K 990 726 233 51K 4.7
201404 839 28K 87 10K 680 171 277 195 2.2K 159 2.5K 1.3K 772 278 48K 4.4
201403 847 33K 82 13K 941 249 158 275 2.9K 184 2.4K 1.2K 701 202 56K 5.2
201402 691 30K 2 11K 552 149 9 145 2K 130 2.3K 965 410 102 48K 4.4
201401 832 31K 1 20K 532 148 7 154 1.9K 119 2.2K 1.4K 274 77 59K 5.4
201312 812 25K 1 7.4K 563 138 4 140 1.8K 141 1.7K 893 262 95 38K 3.5
201311 691 27K 1 6.9K 470 137 2 136 1.8K 116 2.3K 1K 233 75 41K 3.8
201310 825 27K 0 6.9K 589 167 3 149 3.4K 162 2.1K 881 316 85 42K 3.9
201309 707 27K 3 6.5K 758 142 3 284 1.9K 133 1.8K 872 274 114 40K 3.7
201308 834 23K 4 6K 730 149 0 196 2.5K 124 1.8K 822 242 149 37K 3.4
201307 1.1K 22K 3 6.6K 672 152 0 194 2.2K 127 1.9K 817 289 167 36K 3.3
201306 944 26K 1 8.2K 493 148 0 169 1.9K 119 2K 865 256 155 42K 3.8
201305 861 38K 4 8.7K 1K 147 1 174 1.6K 104 2.4K 941 328 173 54K 5
201304 863 33K 2 8.5K 703 133 0 180 1.5K 130 2.5K 1.1K 341 121 49K 4.5
201303 946 37K 2 8.6K 1.3K 161 0 214 1.5K 183 2.5K 939 319 146 54K 4.9
201302 897 34K 1 7.9K 623 186 0 276 1.6K 156 2.3K 808 294 116 50K 4.5
201301 1.1K 34K 3 8.9K 568 173 0 160 1.6K 119 2.4K 763 371 102 50K 4.6
Total 21K 689K 1.2K 201K 18K 4.2K 3.7K 4.7K 50K 3.4K 52K 24K 12K 5.1K 1.1M 100
Percentage 1.9 63.2 0.1 18.5 1.7 0.4 0.3 0.4 4.6 0.3 4.8 2.2 1.1 0.5 100
Tags Philanthropic foundations Philanthropic foundations Philanthropic foundations Pages created by Vipul Naik Philanthropic foundations Philanthropic foundations Pages created by Vipul Naik Philanthropic foundations Pages created by Vipul Naik Philanthropic foundations Pages created by Vipul Naik Philanthropic foundations Pages created by Vipul Naik Philanthropic foundations Philanthropic foundations Pages created by Vipul Naik Philanthropic foundations Philanthropic foundations Philanthropic foundations Philanthropic foundations Pages created by Vipul Naik

We see that even though the overall number didn’t decline from 2013 to 2014, there was a decline from March 2013 to June 2014 for most pages, as we saw with colors.

The dominant hypothesis: shift from non-mobile to mobile Wikipedia use

The dominant hypothesis is that pageviews have simply migrated from non-mobile to mobile. This is most closely borne by the overall data: total pageviews have remained roughly constant, and the decline in total non-mobile pageviews has been roughly canceled by growth in mobile pageviews. However, the evidence for this substitution doesn’t exist at the level of individual pages because we don’t have pageview data for the mobile domain before September 2014, and much of the decline occurred between March 2013 and June 2014.

What would it mean if there were an approximate one-on-one substitution from non-mobile to mobile for the page types discussed above? For instance, non-mobile traffic to colors dropped to somewhere between 1/3 and 1/2 of their original traffic level between January 2013 and December 2014. This would mean that somewhere between 1/2 and 2/3 of the original non-mobile traffic to colors has shifted to mobile devices. This theory should be at least partly falsifiable: if the sum of traffic to non-mobile and mobile platforms today for colors is less than non-mobile-only traffic in January 2013, then clearly substitution is only part of the story.

Although the data is available, it’s not currently in an easily computable form, and I don’t currently have the time and energy to extract it. I’ll update this once the data on all pageviews since September 2014 is available on or a similar platform.

Other hypotheses

The following are some other hypotheses for the pageview decline:

  1. Google’s Knowledge Graph: This is the hypothesis raised in Wikipediocracy, the Daily Dot, and the Register. The Knowledge Graph was introduced in 2012. Through 2013, Google rolled out snippets (called Knowledge Cards and Knowledge Panels) based on the Knowledge Graph in its search results. So if, for instance, you only wanted the birth date and nationality of a musician, Googling would show you that information right in the search results and you wouldn’t need to click through to the Wikipedia page. I suspect that the Knowledge Graph played some role in the decline for colors seen between March 2013 and June 2014. On the other hand, many of the pages that saw a decline don’t have any search snippets based on the Knowledge Graph, and therefore the decline for those pages cannot be explained this way.
  2. Other means of accessing Wikipedia’s knowledge that don’t involve viewing it directly: For instance, Apple’s Siri tool uses data from Wikipedia, and people making queries to this tool may get information from Wikipedia without hitting the encyclopedia. The usage of such tools has increased greatly starting in late 2012. Siri itself was released with the third generation iPad in September 2012 and became part of the iPhone released the next month. Since then, it has shipped with all of Apple’s mobile devices and tablets.
  3. Substitution away from Wikipedia to other pages that are becoming more search-optimized and growing in number: For many topics, Wikipedia may have been clearly the best information source a few years back (as judged by Google), but the growth of niche information sources, as well as better search methods, have displaced it from its undisputed leadership position. I think there’s a lot of truth to this, but it’s hard to quantify.
  4. Substitution away from coarser, broader pages to finer, narrower pages within Wikipedia: While this cannot directly explain an overall decline in pageviews, it can explain a decline in pageviews for particular kinds of pages. Indeed, I suspect that this is partly what’s going on with the early decline of pageviews (e.g., the decline in pageviews of countries and continents starting around 2010, as people go directly to specialized articles related to the particular aspects of those countries or continents they are interested in).
  5. Substitution to Internet use in other languages: This hypothesis doesn’t seem borne out by the simultaneous decline in pageviews for the English, French, and Spanish Wikipedia, as documented for the color pages.

It’s still a mystery

I’d like to close by noting that the pageview decline is still very much a mystery as far as I am concerned. I hope I’ve convinced you that (a) the mystery is genuine, (b) it’s important, and (c) although the shift to mobile is probably the most likely explanation, we don’t yet have clear evidence. I’m interested in hearing whether people have alternative explanations, and/or whether they have more compelling arguments for some of the explanations proffered here.