A little over six months ago, we deployed our first Wikipedia Zero partnership with Orange in Uganda, offering mobile internet users access to Wikipedia without data charges. Since then, nine additional operators in different countries have launched the program, and twenty-two others are in the current queue. We are now beginning to see the impact of the program, and the first numbers are encouraging: page views from our partners in Niger and Kenya have risen sharply, as have unique visitors in Malaysia.
Our stated mission for the Wikipedia Zero program is to reduce barriers for accessing Wikipedia on mobile devices. From the outset, it has been our goal to manage this program with an analytical lens. With the cost of data removed, we expect an increase in the amount of people accessing Wikipedia for the first time, and also hope that existing readers aren’t deterred from reading more articles. The best proxy to measure this is to look at the growth of Wikipedia page views that come from a mobile partner’s IP address range, and compare it to the growth rate of mobile page views from the rest of that same country. Ultimately, we also want to measure unique visitor additions from the program (to verify that new readers are indeed being introduced), but, internally, we can only measure page views for now.
Page View Growth
Two of our partners agreed to let us publish Wikipedia Zero traffic figures from their network for this blog post. Orange Niger and Orange Kenya both launched the program in July of this year. Both countries are extremely mobile-centric, with 41 percent of all Wikipedia page views in Kenya and 29 percent in Niger coming from mobile (compared to a global total of 12 percent). The chart below shows Wikipedia mobile page views for each partner and country respectively:
The news is good. Since the month prior to launch, we’ve seen 77 percent growth in page views through Orange Niger (compared to 7 percent for rest of Niger), and 88 percent for Orange Kenya (compared to -7 percent for rest of Kenya). For each of these two operators, their “Wikipedia share” (the percentage of mobile page views in that country from the partner’s customers) has nearly doubled in that time. Data sets for these two partners are both relatively small, so we’re careful to not to draw too many conclusions from them. However, we’re excited about what it might imply for the future impact of the program.
Unique Visitor Growth
As mentioned before, we’re unable to measure partner unique visitors yet through our internal analytics (our public, global figures for unique visitors are measured by comScore Mediamatrix). However, some partners, such as Digi (Telenor’s subsidiary in Malaysia), do have their own mechanisms for measurement. Since they launched the program in May, unique visitors to Wikipedia on their network have jumped 42 percent, from 91,000 to 131,000. Though it is not yet something we can measure on a recurring basis, it’s a telling indicator that Wikipedia Zero partnerships are successful in bringing new readers to the site.
These three data points make us really optimistic. They show growth, though we need to continually manage and measure to see if growth persists when we work with larger bases, and also need to test what happens over time. We’ll share more data as we can, and we also hope to deep dive into a few markets over the next several months to learn exactly what type of partner marketing activity is most effective in driving the growth we described. Stay tuned.
Senior Manager, Mobile Partnerships
1. ↑ From http://stats.wikimedia.org/wikimedia/squids/SquidReportCountryData.htm, Sept 2012 (mobile site page views / total page views): Kenya (4.1 M / 10.4 M), Niger (69 K / 244 K), World (2.2 B / 18.5 B)