Starting a poll for feedback from Bluesky followers about content curation of the official Bluesky account

[joyeecheung]

The Bluesky account has been running for a couple of months. I think it may be a good time to review a bit about what kind of content the followers want to see more/less.

The idea is to use the poll feature of discussions page of this repo (alas, Bluesky still does not support polls, so we'll have to do it here). I tried to categorize the recent posts that are likely to show up again / not one-off, and the polls will look like this:

• Release announcements e.g. https://bsky.app/profile/marcoippolito.dev/post/3lf67w3ehhs2c

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

• Amplification of new official blog posts and tutorials from the official website e.g. https://bsky.app/profile/nodejs.org/post/3loeabn3ccd23

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

• Reposting technical content about project development from Node.js contributors e.g. https://bsky.app/profile/marcoippolito.dev/post/3lfmjtrd4q22b

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

• Recordings of recent working group/team meetings e.g. https://bsky.app/profile/mhdawson.bsky.social/post/3lfazelrdxs2k

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

• Amplification/reposting of Node.js-related conferences, collaboration summits and other events e.g. https://bsky.app/profile/mhdawson.bsky.social/post/3laymhhuchs2g

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

• Request for corporate support from Node.js contributors e.g. https://bsky.app/profile/raisinten.bsky.social/post/3ljpnwt2djn23

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

• Request for community participation in documentation or project development e.g. https://bsky.app/profile/brianmuenzenmeyer.com/post/3ld7vhp3fh22c

  • This is what I follow the nodejs.org account for
  • I don't mind seeing them from the nodejs.org account
  • I want to see less from the nodejs.org account

And then, after we have the content of poll settled, we can open the polls, with a thread asking what kind of content that people want to see more often. When we think the poll is ready, we'll amplify the discussion pages through the nodejs.org account so that the Bluesky followers will have a say in the content curation of the account. When we have the results (maybe after a month of discussions?) volunteers who help managing Bluesky can take the poll results into account when requesting the content.

WDYT? @nodejs/tsc @nodejs/social

[JakobJingleheimer]

Reposing technical content about project development from Node.js contributors e.g

Typo: Reposing → Reposting (or better: Re-posting)

Otherwise LGTM 😀

[mhdawson]

Overall looks good to me.

The only thought I'd have is if we want to include something like the following in the poll to get information if there are other things people are looking for.

In addition to the types of posts listed I would expect the Node.js social handles to also include X. Where X is a open ended write in.

[joyeecheung]

The only thought I'd have is if we want to include something like the following in the poll to get information if there are other things people are looking for.

Yes, the plan is to have a separate thread

with a thread asking what kind of content that people want to see more often

I am thinking people can use the upvotes and we can take the more upvoted ones into consideration (this is what the discussions pages are designed for, essentially).

[joyeecheung]

Looks like we are happy about the content: I am starting to open the polls, see #140 for the mega thread. I am currently being kicked out by a closing coffeeshop, I will come back for a final check and request a Bluesky post to amplify them when I find a place to get back to it..

[joyeecheung]

Opened request for amplification of the survey thread: #141

[joyeecheung]

Tagging it TSC agenda in case I forget, though I don't think we need to discuss it this week, we can wait for a week or two until we have more responses to review the results.

[jasnell]

This is what I follow the nodejs.org account for
I don't mind seeing them from the nodejs.org account
I want to see less from the nodejs.org account

It occurs to me that these response options are quite flawed, in my opinion, and aren't likely to produce a useful signal. Why? Social media surveys tend to have a massive self selection bias (https://pmc.ncbi.nlm.nih.gov/articles/PMC9714525/#:~:text=Social%20media%20is%20increasingly%20used,%E2%80%94%20a%20%E2%80%9Cselection%20bias%E2%80%9D.).

Individuals with negative opinions are far more likely to participate than those with positive opinions.

We don't know how representative the respondents are of the actual subscriber base. We don't even know if the respondents are actually followers of the account. Even if the respondents are assumed to be representative, nothing in this survey would tell us what potential new followers would want to see, what content would attract those new followers. Folks who are not following the social account now are unlikely to even see the survey let alone respond to it.

I'd be extremely hesitant to rely on the results of this survey for anything but generically anecdotal feedback and would not base any of our content decisions on it.

[joyeecheung]

I am aware of the bias of social media polls, but then I don't think there is a better way to understand what followers think of the content, unless you have a more constructive suggestion?

Also this survey is a special case: it's a survey for the content of the account itself. It's like the "I am not interested" button and like buttons on Twitter, of course those who would click the button have a bias, that seems fine to me and still valid to help us decide how to tweak the timeline. I don't think this can be compared with generic surveys that are not about the account itself.

[joyeecheung]

We don't know how representative the respondents are of the actual subscriber base. We don't even know if the respondents are actually followers of the account.

I would doubt anyone who is not following Bluesky would be interested in participating the survey, registering on GitHub and then finding the Bluesky survey probably take way more effort than registering a Bluesky account and follow nodejs.org.

Having respondents who feel strongly enough about the content curation to participate seems enough, we do even less scientific opinion collections in how we develop Node.js core, the software, when we decide whether users want something or not. The Node.js Bluesky account IMO is just an extension of the Node.js software and documentation.

If people don't participate, that means they don't feel strongly enough, I would count that as "abstain".