Wikimedia Deutschland/new-editors/Key Findings and Results/2019 Key Findings and Results

From Meta, a Wikimedia project coordination wiki

 
Overview
   
Activities
   
Associated Research
   
Available Materials for Adoption


Go back to 'Key Findings and Results' overview page.

'Thank You!' Banner Campaign 2019 ("Dankesbanner-Aktion 2019")[edit]

What and why?[edit]

With our 'Thank You!' Banner Campaign 2019 ("Dankesbanner-Aktion 2019") following the 2019 fundraising campaign, we once again called upon donors to become active editors in Wikipedia. Like the campaigns of the past two years, the 2019 campaign is part of our efforts to recruit and retain new volunteers.

In the ‘Thank You!’ campaign, we thanked all donors who donated during the 2018 fundraising campaign and called upon them to become active for Free Knowledge on a long-term basis: either as a supporting member of the nonprofit Wikimedia Deutschland or as an editor on Wikipedia. The objectives of the ‘Thank You!’ campaign were on the one hand to recruit new editors and on the other hand to recruit new supporting members. Hence, the campaign was being organized across teams at WMDE.

The goal of the ‘Thank You!’ Banner Campaign for the New Editors team was to motivate as many people as possible through the banner to register a user account and to start editing. In addition, we intended to evaluate once more the usefulness of our online courses for new editors.

Summary[edit]

Overall, the 'Thank You!' Banner Campaign 2019 was a success. In particular, the number of new registered users was very high: 615 new users created an account. This number is especially remarkable because the number of page impressions was not much higher than in previous campaigns. In comparison, this campaign has resulted in the highest conversion rate so far: 2.9901 percent. The editing rate of new users lay slightly below average compared with previous campaigns. In absolute numbers, however, a great many new users began to edit: 115. A very interesting finding was that those users who had completed a training module submitted more edits on average compared with those users who had not completed a module or aborted a module.

Wikipedia Action Day 2019 ("Aktionstag 'Wikipedia vor Ort' 2019")[edit]

What and why?[edit]

The Wikipedia Action Day 2018 managed to create a great deal of publicity for Wikipedia, both within the community and the wider public. This year, on November 10, 2019, local spaces and groups in Austria, Germany, and Switzerland opened their doors once more to interested visitors in order to inform about Wikipedia, showcase their own work, and point out the manifold opportunities to volunteer for Free Knowledge.

A takeaway from last year’s Wikipedia Action Day was that comparatively large crowds are drawn to the events. The high number of visitors, however, made it more difficult to accompany each person individually into the world of Wikipedia.

We aimed to achieve the following objectives with the Wikipedia Action Day 2019:

  • Recruiting new community members who participate in editing Wikipedia
  • Increasing the sustainability / long-term effects of the Action Day through follow-up offers and audience-specific appeals
  • Visibility and recognition of the community among the wider public
  • Measuring success: Evaluation of the visitor numbers, as we did in the previous year
  • Accomplishing 250 new user registrations through the online campaign, which runs parallel to the online campaign

Summary[edit]

The Wikipedia Action Day 2019 ("Aktionstag 'Wikipedia vor Ort' 2019") is seen as a success by both the participating local groups and WMDE. In this second iteration of the event after 2018, some previous findings were confirmed, yet some changes in the planning and realization of the event also lead to some new insights. Here are the most important results in brief:

  • The display rate of the web banner had an effect on the number of page views and ultimately also on the number of registrations: The low display rate in 2019 caused a lower number of page views and user registrations.
  • At this point, already two thirds of all banner ads are displayed on mobile devices. Therefore, the optimization of landing pages for smartphones etc. is becoming ever more important.
  • A total of 1,021 people visited the Wikipedia Action Day 2019, which, considering the 18 participating groups across 17 cities, had a broader impact compared with the previous year. Per local space, the number of visitors was slightly lower, which enabled a more in-depth support for individual guests.
  • The percentage of edits by new users was similar to the average of WMDE banner campaigns.
  • The visitor survey brought us very helpful insights. Most guests visited the event because they saw a banner or an invitation for members of the non-profit association (WMDE).
  • Flyers and posters had a very low impact on activating guests both online and offline – at least at the small scale of this year’s print run and distribution.
  • Twenty-two guests of the Wikipedia Action Day 2019 returned to subsequent events in the respective local groups.
  • At the same time, more than half of the respondents to our visitor survey said that they could picture themselves becoming active in Wikipedia. This leads us to the conclusion that there are potential long-term effects of the Action Day, which are unfortunately difficult to measure.


Back to top
Go back to 'Key Findings and Results' overview page.

Social Media Campaign 2019 ("Social Media Kampagne 2019")[edit]

What and why?[edit]

For the re:publica19 conference, we developed a type test that recommends to participants where and how they can become active in the cause of Free Knowledge on the basis of their answers to test questions. The test offers a previously non-existent overview of different ways to participate and entry points into the Wikimedia ecosystem, including Wikipedia, Wikidata, Wikimedia Commons, and becoming active for Open Data and Free Knowledge. You can find a more detailed project description here.

The concept was received very positively by members of the Wikipedia community who were introduced to the test. This also brought about the suggestion from within the community to extend the test on Facebook and make it playable.

So far (May 2019), the landing page and the test had only been presented at re:publica19 and were not known among the wider public. Therefore, we took up the suggestion from the community to expand the idea into a social media campaign.

Through our social media campaign, we aimed to motivate people to become new active volunteer editors on Wikipedia. Towards this objective, we purposefully addressed potential editors outside of Wikipedia and approached them where they were: on Social Media.

Campaign Objectives[edit]

We wanted to raise awareness about the type test and invite Social Media users to complete the test. These were our communication goals:

  • Address new groups of users and target audiences
  • Raise awareness about the opportunity to participate in Wikimedia projects
  • Increase identification with the projects around Free Knowledge
  • Point out concrete areas for participation and showcase the breadth of participation opportunities
  • Interaction between users and the Wikimedia Facebook account beyond the campaign is desired and might bring about a growth in the number of followers.

Campaign Results[edit]

This campaign was the first online campaign of Team Volunteer Support outside of the platform Wikipedia and the banner campaigns displayed there. One goal was to test whether advertisements on Social Media would be of interest to Wikimedia.

The “Participate”-test was very well received on Social Media. Over the duration of around four weeks, ads were booked alternately on Facebook and Instagram. As a result, over 1,200 participants were motivated to complete the “Participate”-test. The test gained additional reach through embedded ‘Share’ buttons for Facebook and other Social Media platforms.

The characters that were getting the most test results were ’String Puller,’ ’Documentalist,’ and ‘Star Author’.

Besides positive comments about the “Participate”-test, likes, and new fans of the Wikipedia Facebook page, there were also some critiques and comments regarding internal conduct among the community. Due to overwhelmingly positive feedback concerning the test and the great number of successful completions of the test, it will be continuously used in the future as part of events and on the Web to activate users for the Wikimedia ecosystem.

Targeting / playout[edit]

Contrasting with WMDE’s previous banner campaigns on Wikipedia, Facebook, through its AdManager, offers the possibility to narrowly target audiences and therefore potentially reach people for whom the campaign is relevant. Based on the test result types (the illustrated characters), different target audiences for the ads were defined:

On the one hand, people who are interested in non-profit organizations, were addressed with a relatively broad message. On the other hand, we also worked with much more specific target audiences: For example, the campaign was targeting groups of people who had a recognizable technology background or whose profile revealed a cultural interest.

Impressions and Conversions[edit]

Compared to banner campaigns, significantly less impressions were achieved: 164,156 impressions on Facebook versus 23 to 168m impressions that are served through campaigns on Wikipedia. Due to the precise targeting, however, the conversion rate is much higher: In this social media campaign, 2,525 page views and 6 registrations on Wikipedia were achieved. The conversion rate of impressions to page views was 1.54%. In comparison, in our banner campaigns, the percentages lay between 0.01 and 0.76%. In regards to the conversion rate from page views to registrations, the facebook campaign’s results were mediocre (0.2376%). Due to the low number of just six newly registered Wikipedia accounts, it is impossible to make any conclusive statements regarding the retention rate at this point. Notes regarding the test: The calls-to-action in the test results are not specifically designed to recruit new editors. Only in two of nine test results, there is a call-to-action to participate in Wikipedia.

Advantages and Disadvantages of Campaign Channels[edit]

Impressions: In contrast to Wikipedia as a platform on which Wikimedia can place banners for free and can very easily achieve very high numbers of impressions, on Facebook, this is paid advertising associated with costs. In return, the impressions are served to previously determined target audiences. This leads to a higher conversion rate from impressions to page views.

Context: Users who are surfing Wikipedia and click on a WMDE banner are thematically (better) prepared by the context for the following pages. Users coming from Facebook need to be provided with more topical context: We found that the type test had a lower bounce rate when users made a detour to the landing page wikimedia.de/mitmachen first. A campaign landing page that is not located on wikipedia.org or other ‘in-house’ platforms needs to explain more.

Effort: A Facebook campaign can be created and administered relatively easily by oneself using Facebook’s AdManager. If needed, third parties (such as media agencies) can be granted administrative permissions to support the campaign. In comparison, a Wikipedia banner campaign requires more administrative input. In this case, potential campaign periods need to be checked for potential competition with other Wikimedia banners, e.g. Wiki Loves Monuments or the Wikipedia Action Day in the fall). Furthermore, WMDE cannot directly place banners on the German-language Wikipedia, but has to follow a community process.

Community/Audience: Online banners on Wikipedia are potentially viewed more critically by the community than advertisements on other platforms. The Facebook campaign was announced in the Wikipedia:Kurier (the German-language online community newspaper) but did not cause any notable reactions. In the case of placing ads on other platforms, we see the potential risk of negative comments regarding the use of donations for advertisements. During the Facebook campaign, there were generally few reactions or comments regarding the placed ads. No comments referred to the financial cost or the use of Facebook. The majority of comments were rather negative in nature. We decided not to react at this point. If there had been a greater number of reactions, especially significantly more negative ones, then a more intense community management process would have been needed on those channels. A shitstorm on Facebook as a public-facing channel would have very likely been received differently than one on Wikipedia project pages, which are barely known outside of the community.


Back to top
Go back to 'Key Findings and Results' overview page.

E-Mail Campaign 2019[edit]

What and Why?[edit]

Wikimedia Deutschland’s previous campaigning work for recruiting new editors concentrated on banners that are displayed on Wikipedia. Our practical experience in addressing and recruiting volunteers on other communication channels had hence been somewhat limited up to this point (fall of 2019). Therefore, this pilot project in October 2019 aimed at testing the potential and effectiveness of E-Mails for recruiting volunteers. For this purpose, in a first step, a small selection of donors was contacted to win them as volunteer editors.

The campaign idea was based on the assumption that donors had already displayed a heightened interest in Free Knowledge and therefore might also possess a higher readiness – besides their financial commitment – to take an active role in editing and the Wikipedia community. Therefore, we wanted to motivate as many donors as possible to register a user account and begin editing.

Campaign Objectives[edit]

The basic goal of the campaign was to recruit donors to become active editors via e-mail. We assumed that donors have a heightened awareness of and interest in the topic of Free Knowledge. Hence, we thought that there was a significant potential for activating donors to become new editors, besides their financial donations.

In light of future campaigns, we hoped to also become better able to provide answers to the following questions:

  • Do donors actually fulfill the high potential for being recruited as new editors, as we assume?
  • What is our general assessment of e-mail as a channel for future campaigns?
  • Can we further improve our landing page for future activities?

The objectives had to be evaluated by a set of key performance indicators. The following data items were collected:

  • Opening rate for sent mails
  • Click rate for sent mails
  • Number of visitors / impressions for the landing page
  • Number of new user registrations coming from the landing page
  • Number of edits made by newly registered users
  • Qualitative features: How is the campaign perceived by the target audience (praise, criticism, complaints, general feedback)

We did not collect data on the training modules.

Results[edit]

As a whole, the campaign was very successful and achieved high conversion and interaction rates. Below are the detailed results.

Results from the first mailing[edit]

In both e-mails, the recipients were asked to become active in Wikipedia. The first e-mail (Template A) postulated that less than 1% of readers are active in Wikipedia and that more people are needed to maintain the world’s largest online encyclopedia. The second e-mail (Template B) focused on motivating recipients to participate by mentioning interesting topics (local peculiarities). The results show that the differences in the interaction rates between the two e-mails were insignificant – both appeals achieved good results. In regards to registrations, which were predominantly achieved through e-mails B and C, we assume that a topical and personal appeal works better. Nevertheless, this working hypothesis needs to be verified through subsequent tests.

TABLE 1

E-Mails (total) Template A Template B
Reach ca. 2,087 ca. 2,072
Opening Rate 39.91 % [833] 38.17 % [791]
Click Rate 5.51 % [115] 4.54 % [94]
Click Through Rate 13.8 % 11.8 %
Registrations Clicking Users: 5.22 % [6] Clicking Users: 10.63 % [10]
Top Links
  1. Get to know more ("Jetzt mehr erfahren") [77%]
  2. Wiki Calendar ("Wiki Kalender") [15%]
  3. Wikipedia Training [8%]
  1. Get to know more ("Jetzt mehr erfahren") [71%]
  2. Gender [17%]
  3. Create Places ("Orte anlegen") [12%]

Nearly 40% of recipients opened and read the e-mail. Across different sectors, this is an above-average opening rate. Roughly 5% clicked on one of the links, and a total of 16 recipients of these two e-mails straight away registered as editors on Wikipedia.

Regarding the number of recipients, the conversion rate of recipients to registrations is 0.3846%.

Results from the Follow-up Mailing[edit]

As the e-mail campaign format is very new for us, in the aftermath of the campaign, we sent out a second e-mail asking all recipients for feedback on the first e-mail. With this e-mail, we wanted to find out how the addressed donors perceived our e-mail and which improvements we can deduct for future mailings.

Following our inquiry, we received around 200 responses, which greatly exceeded our expectations. The feedback was very helpful and at the same time very personal, formulated with great passion for the work done at Wikimedia Deutschland. One quarter of all users indicated that they can very well imagine themselves participating as an editor and that they were already making concrete plans for this. Roughly half of respondents said that their non-participation was due to a lack of free time or interest. Two percent indicated that they had previously had negative experiences. Finally, sixteen percent responded that while they have no interest in active participation, they intend to continue supporting Wikipedia financially.

In addition, we also achieved some page views and registrations through this e-mail. A total of 14 new Wikipedia accounts were registered from this e-mail.


Results of all e-mails
Page views in total

TABLE 2

Page Views
Wikipedia:Treffen_der_Wikipedianer (“Meetings of Wikimedians”) 28
Wikipedia:Wikimedia_Deutschland/LerneWikipedia (“Learn Wikipedia”) 290
Total 318

Registrations in total
TABLE 3

E-Mail Registrations
E-Mail A 6
E-Mail B 10
E-Mail C 14
Total 30

The conversion rate of mailings to page views is 0.8% and therefore higher than the average value of banner campaigns of 0.44899%.

The conversion rate of page views to registrations is 9.4340% and therefore significantly above the average for banner campaigns of 0.71842. This is very promising. In addition, the high response rate for the follow-up mailing points to a close attachment to Wikipedia and/or Wikimedia Deutschland.


Edits by newly-registered users

The activity of newly-registered accounts was relatively low. Just two users made edits.

TABLE 4

Number of Edits Wikipedia Accounts
1 1
2-4 1
5-9 0
10 and more 0
Total 2

The influence of the campaign on donor behavior will be examined for a longer period of time after the campaign. As soon as we have new results, they will be added here.



Back to top
Go back to 'Key Findings and Results' overview page.