Mailing List Archive

[Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter
Dear all,
I write to send a small update on this. In a message about the methodology followed to measure success (https://meta.wikimedia.org/wiki/Talk:Social_media#Re:_Twitter_engagement_questions), Laura Dickinson posted this: "According to its 2022 report<https://www.rivaliq.com/blog/social-media-industry-benchmark-report/>, the median Twitter engagement rate for brands across all industries is 0.037%; for nonprofits specifically, it is 0.054% [our engagement] over the last 28 day period is 2.7%."

I have measured the engagement with that methodology (https://www.rivaliq.com/blog/social-media-industry-benchmark-report/#title-methodology) for @Wikipedia in January (Likes+RT+Comments / Number of followers) and the result is: 0.012%, three times lower than the industry standard and 4.5 lower than for non-profits. For context, Basque Wikipedia had 0.055%, Catalan Viquipedia 0.060% and Indonesian Wikipedia an astonishing 2.79%. (You can check the numbers here: https://www.rivaliq.com/free-social-media-analytics/twitter-head-to-head)

There's an open question about the strategy followed and a sincere proposal of opening this account to a shared volunteers/WMF administration.

Sincerely,

Galder

________________________________
From: ?lex Hinojo <alexhinojo@gmail.com>
Sent: Thursday, January 19, 2023 7:42 AM
To: Wikimedia Mailing List <wikimedia-l@lists.wikimedia.org>
Subject: [Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter

+1

On Thu, 19 Jan 2023 at 07:13, Peter Southwood <peter.southwood@telkomsa.net<mailto:peter.southwood@telkomsa.net>> wrote:

A Wikipedia account should be under the control of Wikipedians, following the editorial policy for Wikipedia, but they could let WMF do the technical work if such exists. WMF can and should run Wikimedia accounts. WMF running a Wikipedia account could be misrepresentation.

Cheers,

Peter



From: Andreas Kolbe [mailto:jayen466@gmail.com<mailto:jayen466@gmail.com>]
Sent: 19 January 2023 02:46
To: Wikimedia Mailing List
Cc: F. Xavier Dengra i Grau
Subject: [Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter



Dear all,



The obvious question surely is: Why not let volunteers (co-)run the Wikipedia Twitter account?



A number of Wikipedia language versions (French, Catalan, Portuguese, Basque, Waray, etc.) seem to have volunteer-managed Twitter accounts that are doing fine. If volunteers are good enough to write the encyclopedia and curate the main page of each language version, aren't they good enough to write (or suggest) the occasional tweet?



Andreas



On Wed, Jan 18, 2023 at 11:20 PM F. Xavier Dengra i Grau via Wikimedia-l <wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>> wrote:

Hi/Bona nit,



This last tweet from @Wikipedia is a good example of what some of us have been mentioning in this list during the past days:



https://twitter.com/wikipedia/status/1615756186640334848?s=46&t=7wB7VI4gwISyFjo-X2jZvQ



Despite the fact that many Wikipedias have already had this new skin deployed since months ago as voluntary testers, not a single mention on their huge contribution was explained on Twitter (neither back then nor today?). We need to go to the 8th tweet of today's publication to read something like "The new features, which start rolling out on English Wikipedia today, were built in collaboration with Wikipedia volunteers worldwide."



If this is the situation in which the main account is monopolized only to the English version and its news/articles, why not specifying it as "English Wikipedia" in the profile and in the main link?



Days pass by and we keep sharing to this list proofs, data and justified arguments (even collagues offering themselves and willing to trace a joint planning!), but still not a word or single thought from the Comms department. Disappointing, I am sad to say.



Kind regards/Salutacions



Xavier Dengra



El ds, 14 gen., 2023 a 09:52, Galder Gonzalez Larra?aga <galder158@hotmail.com<mailto:galder158@hotmail.com>> va escriure:

Egun on Boodarwun/Gnangarra,

You are righth in one thing: it is very difficult to prove a point only from one puntual statistic. That's why I have been tracking statistics for a long time, because patterns are here the most important thing. Neverthless, there is only one way to know if the point me and some other users in this thread are rising is valid: experimenting. @Wikipedia should try something: tweeting 6-7 times a day, with varied topics, "on this day" like tweets, varying timezones and even curiosities about how Wikipedia works (https://twitter.com/depthsofwiki/status/1614045362985082881 2 million impressions in 9 hours). Then, after -let's say- one month, if the results (engagement, followers, retention) are better, it would be quite obvious that there's a point changing the social media strategy. If not, if engagement is the same, no obvious uprise in followers or RTs is visible, the current strategy could be validated.



Me, personally, I'm ready to help the Communications Team with this task, proposing intercultural items that could be tweeted and promoted. If they want help, they know where to go for it. Again, I think that following the same pattern is a bad communication strategy (as we can see by our own eyes) and trying something new could be better. Is up to the communications team to aknowledge this and give a try.



Sincerely,

Galder



________________________________

From: Gnangarra <gnangarra@gmail.com<mailto:gnangarra@gmail.com>>
Sent: Saturday, January 14, 2023 6:00 AM
To: Wikimedia Mailing List <wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>>
Subject: [Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter



Kaya Galder



The assumption that despite there being a wider audience the interests of those audience members is exactly the same, if that was true why have multiple channels. What I am saying is that in different communities that doesnt and will never hold true. Using statistics to compare the two is the issue and then complaining about different audience responses to the same event being caused by those posting to the channel. Its not the channel operators, it's the underlying expectation that all audiences are the same and react exactly the same way every time even as the audience is increasing by many orders of magnitude.



Boodarwun



On Sat, 14 Jan 2023 at 02:06, Galder Gonzalez Larra?aga <galder158@hotmail.com<mailto:galder158@hotmail.com>> wrote:

@Gnangarra: I would doubt on the idea that Pel? is not relevant to the English audience, as it was the most visited article by far that day (https://pageviews.wmcloud.org/topviews/?project=en.wikipedia.org&platform=all-access&date=2022-12-29&excludes=), and the second most visited next day, just after the less known Andrew Tate. Also, the account is not ENGLISH Wikipedia. Is called Wikipedia, so it should take into account, even if it tweets only about English Wikipedia (as pointed by @Xavier Dengra) a global audience. Because, again, the goal is "By 2030, Wikimedia is to become the central infrastructure for Free Knowledge on the Internet.". Not only for US centered people, but by a global audience. Even with that in mind, Pel? was the most visited article in English Wikipedia.



@Yaroslav: Basque Wikipedia is not one of the few accounts tweeting about Pel?, and in perspective, there are more Basque tweeting accounts per speaker, than there are for other larger languages. We are not competing with major news outlets; we are competing to be "the central infrastructure for Free Knowledge on the Internet". Wikipedia is doing well on that: nearly 2,5 million visits in two days for the article about Pel? only in English. I think that there may be very few web services having 2,5 million visits for a page about Pel? in two days, if there's any. Also, next day the most visited article was about Andrew Tate. So, you are right: we are not a news outlet, but we are visited according to the news. Any strategy that doesn't have this in mind, will fail.



You also ask how many tweets a day would be enough. I don't have an answer for this. I would like the communications team to come with one, but they don't seem either to have one. I don't think that tweeting every hour is better, but I'll explain why one tweet per day is a bad strategy, based only in what we know about the Twitter algorithm:

* The Twitter algorithm tends to show a tweet to followers and others more often if it gets more engagements (RTs, likes, comments...). So, maximizing engagements seems a something positive if we want to reach to new people.
* It also shows an account more often if the user interacts with it. If someone likes, RTs or comments a tweet, it seems that this account will be shown again soon. That's why you see more often tweets from your friends than others. And that's why ideological bubbles are created.
* If people are engaged with a tweet, it will be shown more regularly after a tweet by other people you follow once you scroll down. This is why if you open a tweet by a far-right politician, you will see below other tweets by far-right sided politicians and the opposite for left, libertarian, green or vegans. It shows you similar content, based on people's interaction.

So, tweeting more doesn't maximize engagement (if you tweet every minute, you will lose it), but tweeting less minimizes engagement. If you only tweet once a day, and you don't get too much attention, your next tweet will be less important for the algorithm, and so on. The only valid strategy is one that gets people engaged to your tweet, so you get more impressions, and this drives more interactions, and this drives more followers. Because, at the end of the day, we want to be "the central infrastructure for Free Knowledge on the Internet".



I don't know how much is the ideal thing. In Basque Wikipedia our strategy is to publish 5-6 tweets every day, and then also interact with people talking about Wikipedia or speaking about articles they have created (like @viquipedia does, with great success). Our topics from the 5-6 daily tweets now (2023) are like this: every morning (yes, most of our followers live in the same time-zone) a biography of someone who was born/died on this day; then, something that happened 100 years ago. At noon, an artwork. If the artwork is depicting something interesting, a second tweet linked to that explaining the artwork itself. Two tweets in the afternoon: the first one, optional, about something related to Wikipedia itself (Statistics, projects, some user who has created something cool...) and then science/technology in a broad sense. At evening, we like to tweet something related to current events, if this is interesting. We have a shared doc with the daily tweets and we program them some days in advance. Also, we use MOA to have them copied to Mastodon.



I don't know, again, if this is the optimal. I know that is better than one-per-day, because data is obviously better. Engagements, followers and interactions are better this way, as I have proved above.



Best,

Galder









________________________________

From: F. Xavier Dengra i Grau via Wikimedia-l <wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>>
Sent: Friday, January 13, 2023 3:37 PM
To: Wikimedia Mailing List <wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>>
Cc: F. Xavier Dengra i Grau <xavier.dengra@protonmail.com<mailto:xavier.dengra@protonmail.com>>
Subject: [Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter



Hi/Bon dia



Yaroslav: Also, you say one tweet per day is too little, how may do you think is normal? If I personally see an account which tweets more than say 10 per day (not counting threads) I start thinking may be it is a spam generator.



Since 4 years ago we updated the social media methodology for the Catalan Wikipedia Twitter account (approx 4.5M native speakers, 10M audience), we boosted from 15.3K to 45.4K speakers, now being the 4th most followed language of Wikipedia.



Our method in a nutshell: we have up to 23 knowledge themes that we oblige ourselves to post at least once every week. The number of our daily tweets vary from 6 to 10 only in content (i.e., articles). This depends on, ofc, whether it's a working day vs a weekend or other time aspects (peak hours). Plus the interactions (RT+kudos) with our wikipedians that share their new articles tagging us, which has been a massive way to appreciate their task and to visibilize to others the task of being a volunteer in Wikipedia. In fact, the latter has been especially critical to bring us huge additional views and to renew a few of our new, most active editing community (especially young users!).



If our account, managed by volunteers, can conduct this organized work for a small-medium size language, why should we accept that a whole staffed team from the WMF, firstly, rejects to provide engagement data on our common, biggest handle? And secondly, why should we give up on them preparing a strategy to improve its scope and objectives?



Regarding the last question, I'd like to add a last thought: never ever in the 4 years that I've been upfront in the handles in my language, the @Wikipedia account has given a simple, courtesy RT of any knowledge content (articles) from the Arabic, Bahasa Indonesia, Basque, Catalan, Galician, French, Suda or Portuguese (etc.) existing handles. That should be a key aspect in our debate.



Because if @Wikipedia is mostly used as the ?central account? for the project, then it should also be very careful 1) to not always post in English and give some room to interact with the other language handles, 2) to stop centering their tweets on English-speaking culture, and 3) to post without clear range of topics to stay balanced. Oppositely, if it is decided that @Wikipedia is only the English-language handle, then it may change its profile name to "English Wikipedia" and not continue as the reference speaker either for the WMF nor for significant news or events.



Best/Salutacions,



Xavier Dengra

------- Original Message -------
On divendres, 13 de gener 2023 a les 14:56, Yaroslav Blanter <ymbalt@gmail.com<mailto:ymbalt@gmail.com>> wrote:



Hi Galder,



on the other hand.. Basque Wikipedia is one of very few accounts twitting on the Pele death in Basque, whereas a lot was twitted in English. I do not think English Wikipedia twitter can compete with major news outlets, they operate on a completely different scale.The low-hanging fruit would be twitting DYKs, FAs, GAs, or may be some other randomly picked stuff. Also, you say one tweet per day is too little, how may do you think is normal? If I personally see an account which tweets more than say 10 per day (not counting threads) I start thinking may be it is a spam generator.



Best

Yaroslav



On Fri, Jan 13, 2023 at 2:26 PM Galder Gonzalez Larra?aga <galder158@hotmail.com<mailto:galder158@hotmail.com>> wrote:

Some months have gone since I started this topic in this list, and still, we can't know how much engagement we have at Wikipedia, because data is not available. Twitter is now owned by Elon Musk, things are changing, there are more accounts in Mastodon daily, but still Twitter matters. I have been looking at the Twitter activity in the last days for @Wikipedia and I'm still very worried about the (lack of) strategy followed here. A full team, with staff members, which only produces one tweet per day, a lonely message in the vastness of the ocean, and gets really poor engagement numbers.



A couple of weeks ago Pel?, one of the greatest football players of all time, died. (English) Wikipedia Twitter account needed 7 days to tweet about it, even if the article was changed in a few minutes after the death (https://twitter.com/Wikipedia/status/1611363972174778368). The tweet had 13.729 impressions (now we can know the number of impressions), 14 RTs and 129 likes. Wikipedia account has nearly 644.000 followers. If we divide these two numbers, we get a rate of 2,13% of impressions per follower.



The same day Pel? died, Basque Wikipedia made a tweet. Not a week after, just when it was news (https://twitter.com/euwikipedia/status/1608541274491211776). The tweet had 964 impressions, 3 RTs and 2 likes. Basque Wikipedia account has 7,956 followers. This is a rate of 12,11% of impressions per follower. x5.68 times larger, relatively than (English) Wikipedia Twitter account.



(English) Wikipedia Twitter account has nearly 81 times more followers than the Basque one. English Wikipedia is more visible, because it has a (now golden) verified account symbol, so tweets are more often promoted. English has 1.500 million speakers around the world. Basque has fewer than one million. English Wikipedia should have around 1.000 more followers than Basque Wikipedia. English Wikipedia article about Pel? had 2,5 million pageviews in the two days after his death. Basque had 250 pageviews. This is 10.000 times more pageviews.



@Wikipedia has 644.000 followers, and @euwikipedia has nearly 8.000. Audience of English Wikipedia is 10.000 times larger for the same event. Why Wikipedia is not 10.000 times larger? Why doesn't Wikipedia account have 80 million followers? YouTube's Twitter account has 78 million followers. "By 2030, Wikimedia is to become the central infrastructure for Free Knowledge on the Internet.". How could we if Youtube's account has 100x more followers than we have? How can think that we are in a good shape if our tweets are only seen by less than 2% of our followers?



I hope that 2023 comes with a change. A change to open these accounts, have a fresh way of thinking on social media ,and building engagement, both with momentum, not losing opportunities, and promoting good content.



Sincerely



Galder





________________________________

From: Galder Gonzalez Larra?aga <galder158@hotmail.com<mailto:galder158@hotmail.com>>
Sent: Tuesday, August 16, 2022 3:21 PM
To: Wikimedia Mailing List <wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>>
Subject: Re: [Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter

Dear all,

Some weeks ago, we had a discussion here about the different approaches we have for the @wikipedia account at Twitter. We don't know yet how many interactions does the account has, but as I said in the discussion, we try to find ways to measure our work at @euwikipedia. Today I want to share with you that this account was ranked last week as the most influential social-movements account in Basque language (https://umap.eus/ranking/gizartea) and the 10th most influential account in all categories (https://umap.eus/ranking/orokorra). This is a good metric we use to know if we are doing fine or not.



Sincerely,

Galder



________________________________

From: Andy Mabbett <andy@pigsonthewing.org.uk<mailto:andy@pigsonthewing.org.uk>>
Sent: Friday, August 5, 2022 8:50 PM
To: Wikimedia Mailing List <wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>>
Subject: [Wikimedia-l] Re: @Wikipedia losing opportunities in Twitter

On Mon, 18 Jul 2022 at 18:48, Lauren Dickinson <ldickinson@wikimedia.org<mailto:ldickinson@wikimedia.org>> wrote:

> Also, Andy, we will follow up this week regarding your questions
> about the @WiktionaryUsers and @Wiktionary accounts.

Three working weeks have passed since the above was written; I've seen
no such follow-up. Have I missed something?

--
Andy Mabbett
@pigsonthewing
https://pigsonthewing.org.uk
_______________________________________________
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/ASHCU4Z7TN2Q5PJCZ6JAXHWJSJYI3BTG/
To unsubscribe send an email to wikimedia-l-leave@lists.wikimedia.org<mailto:wikimedia-l-leave@lists.wikimedia.org>

_______________________________________________
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/5MHFSBSKJSRIDF5TKH265YZRPOPTZPQA/
To unsubscribe send an email to wikimedia-l-leave@lists.wikimedia.org<mailto:wikimedia-l-leave@lists.wikimedia.org>



_______________________________________________
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/63A3HD7UDTQUAB2ALBRPOW5V3IDUSULP/
To unsubscribe send an email to wikimedia-l-leave@lists.wikimedia.org<mailto:wikimedia-l-leave@lists.wikimedia.org>




--

Boodarwun
Gnangarra

'ngany dabakarn koorliny arn boodjera dardoon ngalang Nyungar koortaboodjar'



_______________________________________________
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/OKQ6HNZAJB4XGJSWMBWKOR3HFRLVCT5S/
To unsubscribe send an email to wikimedia-l-leave@lists.wikimedia.org<mailto:wikimedia-l-leave@lists.wikimedia.org>



[https://s-install.avcdn.net/ipm/preview/icons/icon-envelope-tick-green-avg-v1.png]<http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>

Virus-free.www.avg.com<http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>



_______________________________________________
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org<mailto:wikimedia-l@lists.wikimedia.org>, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/NFOQXPW3ZZDDWYLNXQRDSCKWSBYNZQBJ/
To unsubscribe send an email to wikimedia-l-leave@lists.wikimedia.org<mailto:wikimedia-l-leave@lists.wikimedia.org>
--
?lex Hinojo