A World without Wikipedia


Image from: Gayvoronskaya_yana / Shutterstock

During the SOPA and PIPA protests last Wednesday, we got a glimpse of an alternative universe:  A world without Wikipedia.

Granted, there were ways around Wikipedia’s blockade.  And granted, Wikipedia’s site had a strong message on it that probably affected Internet behavior unto itself. That is, many visitors to Wikipedia’s site probably turned to news sources to learn more about why their encyclopedia was unavailable. Moreover, there were many other major Internet companies (and smaller firms, too!) that participated in Wednesday’s protest; so Wikipedia’s blackout was not the only change in an otherwise “normal” world wide web. Yet the fact remains that Wednesday’s blackout forced most Wikipedia users to search for information elsewhere. It created a great ripple in the typical flow of Internet traffic.

The chart below offers some evidence of that. It shows the daily reach and daily attention on Wikipedia.org last Wednesday and every other Tuesday or Wednesday after a federal holiday over the past six months, with the exceptions of the weeks of Thanksgiving and Christmas.* Daily reach, the percentage of Internet users that day who went to Wikipedia.org, is within line, or perhaps slightly higher, than one would expect given the pattern over prior months. The drop in attention – the percentage of total time spent on the Internet that day that was spent on Wikipedia.org – is stark. This shows that most users did not circumvent the blackout to read Wikipedia as they normally would.  The question is, where did they go?

Daily Metrics on Wednesdays

The most obvious place to look for increases in traffic is on other encyclopedias across the web. Compete categorizes ninety of the top 5,000 domains as reference “Dictionaries, Thesauruses, or Encyclopedias.” On January 17th, the day before the protests, Wikipedia.org received 0.32% of total daily attention. The other 89 sites collectively received 0.20%. If the other references provided a good substitute to Wikipedia, then one would expect the drop in attention in Wikipedia to coincide with a rise in attention to the other domains.

The chart below shows that even as attention to Wikipedia.org reduced by 0.18 percentage points during the protest, a full 58% of the attention to the domain the day before, attention to the other 89 domains in the same category only increased by 0.01 percentage points. This shows that Internet users did not consistently substitute other encyclopedias for Wikipedia while the site was blacked out.

Daily Attention

That is not to say that no encyclopedia received increased attention during the SOPA and PIPA protests. Perhaps the most notable jump in attention to any reference domain that day came to Britannica.com. Both reach and attention “spiked” on the eighteenth, as seen in the chart below.

Daily Reach and Attention

Nevertheless, the evidence at large shows that users did not typically turn to other reference domains when Wikipedia blacked out. Given this information, a more likely scenario is that users that normally would read articles on Wikipedia.org instead went to topic-specific sites for their information. So a user who normally would read up on Meat Loaf on Wikipedia might instead go to meatloaf.net, the result that appears after the Wikipedia article on the Google search results page for “meat loaf.”

So there you have it, no reference substitute for the largest reference domain on the web. To me, the lack of a proper substitute to Wikipedia adds to the strength of their form of protest. Users who take the non-profit for granted truly felt the effects of its absence. I suppose that may be part of the reason that so many users responded to Wikipedia’s call to action during the protests last week.

* Daily reach and daily attention to Wikipedia.org follow a weekly cycle, with the highest values on Sunday and the lowest on Friday. When a federal holiday falls on a Monday, Monday’s value tends to be very close to Sunday’s, Tuesday looks more like Monday typically would, and Wednesday looks more like Tuesday. Tuesday and Wednesday tend to be close in attention and reach, anyway, so a chart comparing last Wednesday to all other Wednesdays would show a very similar pattern.

About Jody Schechter:
As an Analyst for Compete, Jody manipulates data to address the information needs of our Online Media and Search clients. Outside of work, her interests include arts and crafts, developmental economics, Scrabble, and dance. Jody has a Bachelor of Science in Economics, with minors in Mathematics and International Studies, from the University of Michigan.