Engagement/MDN Durable Team/Community hypotheses
As part of 2016 research around community, the MDN team has tested some hypotheses. The first batch of hypotheses has been draft in October 2016 and the results are available below.
Contents
Summary of findings
See also this summary document.
| Tactics | Status | Result | Notes |
|---|---|---|---|
| Give away t-shirts (long time contributors) |
Done (November 2016) |
Negative | Our first findings indicate that giving away t-shirt to long time contributors has no effect. It doesn't change their contribution behaviors neither reengage them if they've stopped contributing.
It is recommended to run another test with new contributors to see the effect it can have on that different population |
| Give away t-shirt (New contributors) |
? | ? | |
| Highlight top of the month contributors | Done (December 2016) |
Positive | See Top contributors tracking spreadsheet
Public promotion on Twitter had a positive effect on contribution by users whose Twitter handles were known and used. It is recommended to find ways to encourage contributors to add their Twitter handles to their profiles, as well as to start a program of thanking contributors on Twitter using their handles. |
| Publicly thank first-time contributors whose profiles contain Twitter handles | Done (July 2017) |
Negative | Contributors thanked on Twitter did not contribute again after being thanked. |
| Engage “one shot” contributors one month after their last contribution | Done (April 2017) |
Negative | Unthanked contributors were more likely to contribute again than thanked ones. |
| Engage first-time contributors one week after their first contribution | Done (April 2017) |
Negative | Thanked
contributors were no more likely to contribute again than unthanked ones. |
| Send a thank-you that shows contributors that someone read their contribution | Done (March 2017) |
Negative | See Tracking spreadsheet; thanked users were less likely to contribute after being thanked than unthanked contributors; note that the sample size was very small (20/21 in each cohort) |
| Thank l10n contributors from the last month | Abandoned | ? | Many of the contributors in the one-month-later and one-week-later experiments were l10n contributors. |
| Create a page listing l10n prioritiess | Done (January 2017) |
Positive | See MDN L10n Priorities Status
Our findings tend to indicate that promoting such priorities can help contributor to focus. Even if the results are pretty small in number they demonstrate a positive trend on the period (against a global negative trend) for establish communities. However, as expected it has no effect on abandon communities or communities that do not interact with the team through the regular communication channel we are using. |
| Advertise l10n priorities on the MDN homepage | Pending | ? | On hold pending homepage redesign |
Hypotheses
The following hypotheses has been formulated based on the expectation we have on how to move our community metrics.
More interaction increases retention
In order to check that hypotheses we want to test the followings tactics:
- Engage “one shot” contributors one month after their last contribution
Personalized interaction increases retention
In order to check that hypotheses we want to test the followings tactics:
- Send a thank-you that shows contributors that someone read their contribution
The more we engage community the faster l10n content increases
In order to check that hypotheses we want to test the followings tactics:
- Thank l10n contributors from the last month
Increasing recognition will increase retention
In order to check that hypotheses we want to test the followings tactics:
- Highlight top of the month contributors
- Give away t-shirts
Provide content priorities to contributors increase content creation
In order to check that hypotheses we want to test the followings tactics:
- Create a page listing l10n priorities
- Advertise l10n priorities on the MDN homepage
