Talk:CentralNotice

WMCH_survey_2025

Latest comment: 3 hours ago6 comments3 people in discussion

@Ilario: Can you please elaborate why this campaign doesn't follow any of these guidelines? And not all of these are new.

It is not linked in CentralNotice/Request, only CentralNotice/Calendar.
The campaign was not created/requested 14 days in advance.
Any kind of surveys have to be discussed at least 90 days in advance with WMF (pinging TAndic (WMF)) and publicly confirmed. The landing page fully lacks data protection notes which is highly concerning and likely would not have resulted in a confirmation.
I cannot find any community communication about it before the start of the campaign. In parallel, campaigns for Wiki Loves Folklore and Steward Elections are running in your target area. I cannot find communication with these campaigns either, especially not that the WMCH campaign was given higher priority than the others.
The staff-created campaign of WMCH was not approved by a volunteer CN admin (and no, in this regard your volunteer account cannot do this).

For that reason, I have disabled the campaign and I strongly encourage you to follow CN guidelines first. — Due to the lack of a request page, I'm using this space here. Certainly, we can move it elsewhere. Pinging active CN admins and WMF staff for checking my actions: @Ciell, Martin Urbanec, Vermont, Johannnes89, Astinson (WMF), and SWangari-WMF: —DerHexer (Talk) 11:43, 19 February 2025 (UTC)Reply

The survey has been a standard practice since Wikimedia CH received funds through the FDC and remains an important tool, particularly as we conduct the fundraising campaign for the Wikimedia Foundation in Switzerland. It is not an exceptional event but rather a recurring practice that has been carried out for several years. For this reason, I included it in the calendar and activated the banner in my capacity as Wikimedia CH staff, as it was a planned activity aligned with the fundraising strategy (and conducted after the fundraising campaign). So it's s recurrening campaign and not a new campaign.

I apologize, as I should have formally opened a new request. However, I am unsure how to proceed in cases where activities are already implicitly covered by an "informal" agreement with the WMF.

Regarding privacy, the WMCH privacy policy applies, as no private data is collected. Participants may provide their username only if they wish to do so, but no IP addresses or other personal data are stored.

As you know there was an impression diet exactly to don't impact with existing campaigns.

On the other hand, I find it inconsistent that the Wiki Loves Folklore banner is currently displayed in Switzerland, despite no one from the Swiss community organizing the event here and Wikimedia CH not supporting it due to the short notice. Moreover, the banner is highly invasive—particularly in the Italian community, where it appears twice (both in the centralnotice and the sitenotice).

That said, there is no issue from my side and no need to add two days for approval. I will report this discussion internally and clarify how to proceed in the future to ensure alignment with the process. --Ilario (talk) 09:59, 26 February 2025 (UTC)Reply

@Ilario: Thanks for sheding some light into this, although I'd still prefer to have a CentralNotice request page for this campaign where we can proceed with the conversation.

The usage guidelines do not distinguish between new and returning campaigns. In fact, the last WMCH survey I could find was from 2022. Processes at CentralNotice as well as the WMF research/survey team have changed significantly in the past few years. As far as I know, every survey (including returning surveys) have to be approved by the WMF, at least the WMDE community survey will do. — Under the impression of the new information you provided, I can believe you that the WMF has informally approved the 2025 survey or a free pass for your community surveys. However, I'd prefer that TAndic (WMF) would also approve that publicly so that we can document it for this and possible future campaigns as this is the formal process according to the usage guidelines.
With regards to privacy, I am not fully convinced. Your landing page says “The record of your survey responses does not contain any identifying information about you, unless a specific survey question explicitly asked for it.” (marked by me). Free text answers can contain private data so that this should be clarified. Especially since this campaign is also running in Liechtenstein which is European Economic Area where the GDPR applies.
Wiki Loves Folklore runs on a lower diet in non-participating countries per this discussion. However, I agree that the campaign was not sufficiently planned. That was stated in the campaign and will hopefully taken into concern by the organizers. I encourage you to share your feedback to these either directly or at the request page. — Having the campaign shown twice in Italian Wikipedia makes no sense at all. It would have been nice if you or any other itwiki user would have reported that to either the organizers or the CentralNotice admins (it would have been fully appropriate to disable it for itwiki and drop a not on the CentralNotice request page, too).

Thanks for taking this feedback into consideration with WMCH. Please let me know if I can be of any help. Best, —DerHexer (Talk) 18:14, 28 February 2025 (UTC)Reply

Dear @DerHexer:,

Thank you for your thoughtful feedback. I truly appreciate your insights and fully understand your concerns. However, I would like to offer some clarifications:

Survey Continuity and CentralNotice Guidelines: The survey has remained largely unchanged over the years, as it is essential for tracking the evolution of responses and ensuring meaningful comparisons over time. It serves as an important opportunity for the Swiss Wikimedia community to provide feedback. If the questions were significantly altered, it would be impossible to measure trends accurately. However, I acknowledge that the CentralNotice guidelines do not explicitly differentiate between recurring and new campaigns, particularly when the survey is part of a fundraising campaign, as is the case here (I understand the guidelines state that exceptions only for the donation banners and not to the banners related to fundraising campaigns, but it's not clear). Given this, I will report your concerns and ensure that future procedures align as closely as possible with the guidelines.
Privacy Considerations and Compliance: The survey strictly collects only the data that respondents explicitly choose to provide (opt-in). As Wikimedia CH is responsible for fundraising campaigns, it ensures full compliance with both the GDPR and the nLPD (Swiss Data Protection Law). In fact, I have previously pointed out cases where non-Swiss campaigns operated in Switzerland solely under GDPR, without considering nLPD requirements. To address such concerns, Wikimedia CH has actively worked with legal offices to review and update its policies. Additionally, Wikimedia CH has an officially recognized DPO (Data Protection Officer) registered under Swiss law. The survey does not retain identifiable data—so much so that the same individual can participate multiple times without any record being kept.
CentralNotice Usage for Photo Contests: I have been part of the international team for Wiki Loves Monuments and have organized Wiki Loves Earth multiple times. Additionally, Wikimedia CH actively supports the Wiki Science Competition. However, for all these photo contests, the CentralNotice banner is only displayed in countries where a local team is organizing the event. In Switzerland, no local team is running Wiki Loves Folklore, which may indicate a lack of local consensus for its promotion. On the other hand, this survey is a special case, designed to gather community-wide feedback in line with Switzerland’s strong tradition of direct democracy. As a result, some community members have inquired about its visibility. To ensure broad participation, we are actively sharing the survey link through social media channels and the mailman mailing lists of the Swiss communities.
Understanding Swiss Community Consensus: We also need to assess the consensus of the Swiss Wikimedia community. However, since there is no dedicated Swiss Wikipedia, the geolocated banner is our only viable solution. Unlike a message on the village pump, where we cannot verify whether respondents are actually from Switzerland, the banner ensures that only users within the country receive the invitation. This is essential because the survey does not collect personal data, meaning we have no other way to distinguish between responses from Swiss contributors and those from other countries.

My intention is not to assert a definitive stance but to provide context for this approach. The current guidelines appear to be more tailored to global or multi-national campaigns, whereas their applicability to campaigns focused on small communities or localized areas is less clear and supportive. I absolutely understand your actions and the reasoning behind them, however I find it difficult to fully comprehend such a radical measure (I would have appreciated more a simple personal email or a message on the discussion page). This situation could instead be seen as an opportunity to refine and enhance these guidelines. Let me know if you have any further suggestions. --Ilario (talk) 22:19, 28 February 2025 (UTC)Reply

@Ilario Specifically about WLF: we previously discussed that we (CN admins) would like to have the option to turn off ineffective banners before their end date. We did not come to a conclusion in this discussion (if my memory serves me right), but with the scheduled campaigns for March coming up, I looked up the results from the WLF non-participating countries banner in the Netherlands, and in the whole month of February there have been 2 uploads for NL. For me, this means the banner is not effective enough for my community and I decided to make more room for other, newer banners, and I therefore turned off NL targeting in WLF, and left a note about this on the WLF request page. Looking at the WLF uploads from Switzerland, I think these are 36 images by one user, on one day.

Just to illustrate: I think we need to continue the conversation about banner effectiveness at some point. Ciell (talk) 11:53, 2 March 2025 (UTC)Reply

Yes, I have noticed this as well and had the same consideration, especially when taking into account the number of participants. Over the course of a month, the overall efficiency seems quite low. However, I have also observed similar banner duplication in other Wikipedias, such as here, where one appears in the CentralNotice and another in the SiteNotice. Running such an invasive campaign for two months borders on spamming. Surely if the sitenotice is used, there is no sense to run the centralnotice as it vampirizes other campiagns. Ilario (talk) 16:12, 2 March 2025 (UTC)Reply