show something for communities < 150 #1316
Comments
|
Getting some sort of incentive to reach 150 would be good. As far as I can tell the only incentive is displaying the "top" something lists. Or am I missing some other benefit of joining a community on Gittip? For me the only reason to join a community currently is to say: "I do / am / use the same as these other people" |
|
@mvdkleijn Yeah, we need to add value to communites >= 150 as well. See #967, etc. |
|
Yes please :) |
|
Based on @pjc's comments at #266 (comment), I think it makes sense to provide something at 50. I would expect this to be less than what we provide at 150. I don't want to lose the incentive to reach 150. |
|
What the heck do we provide at 150? A list of people in it as far as I can tell an nothing else. The communities "feature" sucks. It is near impossible to "self-organize" since you can't see who's in your fledgeling community (<150 members) and to make you community grow you'd have to shout into the void on various platforms. (which I loathe to do since I hate spam myself) In essence there's no real reason to form a community right now. The only reason is to abuse it as a sort of "I have this skill" set kinda list. I believe this to be actively hampering the growth of Gittip. If we want Gittip to be "just another payment platform" then fine... but if we want to live up to our stated goals, we're trying to form communities. We need to give people the tools to do that. (also see my comment #967 (comment)) |
|
I have similar concerns as @mvdkleijn, that the existing incentive instead hampers growth by being unnecessarily opaque. @whit537 what was the original reasoning for the 150 person incentive? |
|
+1 perhaps I don't understand the purpose of "teams" and "communities" and how that relates to the limit. Is it just to incentivize people to get to that number? I don't think that'll work. OpenPhoto/Trovebox has a community but it's not at 150. So there's no incentive at all but we have contributors that could benefit from gittip. |
Even so, 23% of all users (5,411 / 23,518) have joined at least one community. Of users who have joined since communities launched, 35% (4,028 / 11,568) have joined at least one community. Humans are social animals, ya know? :-)
We have 2,247 active users, and 23,518 total users. Users who join a Gittip community after joining Gittip are three times as likely to be an active user as those who don't (20% [1,077 / 5,411] vs. 6% [1,170 / 18,107]). My hypothesis is that people are more likely to use Gittip (not just join Gittip) when they see Gittip usage as a practice of a community with which they identify. In other words, I believe this suggests that the communities feature is actively fueling Gittip's growth. Let's make it even better! 💃
The reason to provide an incentive at all was to encourage people to use Gittip. Per the above, I believe it's working. As to why the target is 150: I took that from Dunbar's number. See #266 (comment) for discussion and links re: group size and group dynamics. Per #1316 (comment), I think we should use 50 as the low number for communities. We should show something at 50, and then something more at 150.
Sure, Gittip in general sucks. :-) Let's keep this ticket focused on providing something to communities < 150. As I say in the description, "Seems like this would be good to revisit after we add more value at the 150 tier." We have a number of tickets identifying specific ways to improve the communities feature (e.g., #930 #952 #954 #963 #986 #1240). Want to reticket some of the specific suggestions from #967 (comment)? :-) Here's the sql I used to get the above numbers. |
Teams are for smaller, tight-knit, high-trust groups. Communities are for larger, loose-knit, low-friction groups.
What's your goal, to encourage people who are writing code for Trovebox? Or to build an ecosystem around Trovebox? Or to fund development of Trovebox core? Or ... ? |
|
@whit537 (1) is supported by the data, but (2) is not.
(2) implies that joining a community was the cause of their activity, when that's not necessarily the case. |
As someone who joined a while back and +1ed this change, I can tell you that the reason I'm not active on Gittip is directly related to the fact that I can't tell if my community has a presence. Even if I were to bring people to Gittip, I can't tell how that's impacting the community's growth generally on Gittip until 150. I also don't see any value in hiding this information. It's an artificial incentive, that isn't even working. You say that perhaps given the numbers above it is working? I say, the numbers above are in spite of this, and they'd be better after a change. |
Is it fair to say that it's not working for you because you don't identify with any of the larger communities? My take is that we should add additional value > 150, and then add something at 50. |
|
I think it's fairer to say that you don't have more larger communities because of this limitation. I'm heavily involved in the WordPress community generally, which isn't heavily engaged with Gittip at the moment. I saw increased interest when I mentioned Gittip, but when you can't see who is / isn't already using it easily, why join? If I don't think there's a presence from my community on Gittip, what incentive is there to stick around? If I can see, and have an effect on, the growth of my community on Gittip, then I'm happy to stay. Does that make sense? |
Thanks for the push-back. I've changed this sentence above:
To read instead:
I'm not sure how the observation that "users who join a community are three times as likely to be active" implies causality any more than the observation that "active users are three times more likely to be part of a community." It seems to me that if the one implies causality, so does the other. I've also dug a bit into how one can determine causation from data, finding it suggested that "the question of 'true causality' is deeply philosophical," with "predictive causality" being one way of scoping the question so as to be more easily answerable. I don't have any experience in determining causality. Do you? How do we do this? My hypothesis is that using Gittip's community feature causes active use of Gittip because humans follow the pack (so to speak). When users see a community, with which they identify, using Gittip, they are more likely to use Gittip themselves, because it reifies their group membership. Clicking "join" on a community page triggers some reflex of group identity, and causes people to take further action on Gittip related to their community. I suppose down this rabbit hole we would look at research on group belonging. An alternative explanation might be that Gittip users become active first, become invested in being a Gittip user qua Gittip user, and this leads them to explore the site and experiment with its features, such as the communities feature. This seems implausible to me, but of course that doesn't mean it isn't true. Perhaps one way of getting at the truth of it is to see whether people join a community first and then become active, or whether they become active first and then join a community. "Becoming active" means |
@Japh Yes, and it's helpful to have the specific background experience. Thanks! :-) Does it sound right to you to have increasing incentives at 50 and 150 users? Or do think having only one gate at 50(?) is what we should do? |
Certainly the WordPress community is larger than 150 people. ;-) In fact, it's currently at 256 on Gittip. Maybe the problem is that existing communities don't offer much value? Like, people show up at Gittip, look at a parallel community like Drupal (or whatever), and nothing really grabs them, so they don't join the WordPress community? I hypothesize that making the 150+ experience more compelling will make it much easier to reach that threshold on Gittip for new communities. I don't think that making the existing experience available to 50+ will help us much. Honestly? 150 is tiny for an Internet community. I bet there are millions of Internet communities with 150+ members. I think we should focus on adding value at 150+, and then reevaluate this ticket. |
I think this works better when you can see that others are also a part of your community.
There were less than 150 when I joined, it's great to see it's grown since then, perhaps now I'll also become active ;) I agree that it'd be great to allow benefits once a community reaches a certain size to incentivise growth. I don't agree that hiding members for communities < 150 (or < 50) does this. |
+1, probably talking in circles until this happens. Can't wait to see where this goes! |
Sure, okay. So maybe we even show a basic list of members right from the get-go and frame it as "Here are your co-conspirators! Work with them to recruit more people!" Is that the sort of thing you're thinking of? |
|
(Frame it that way for some initial small value, < 10 or < 50 or something.) |
|
Yeah, basically. I think that would definitely help engage new community members, which is the real goal, right? I mean, yes, you want incentives to keep going, but you also need to complete that "group feeling" to begin with to get initial engagement :) |
|
@Japh I think the circling has happily converged. :-) Okay! Enough talk! More code! 💃 |
Community pages are currently just the main page, scoped to a subset of people. If we're to limit that already limited feature set further for < 150 people, what would that look like? There's also been talk of adding value for > 150 people; what would that look like? |
|
I see my comment made a dent... but how the hell did this discussion go from "the communities feature isn't working" to "I have a hypothesis on active vs inactive users"? It doesn't matter if the user is active or not (for the concerns I voiced earlier)... The communities "feature" is in essence just a list of people in a community after the community already formed on different platforms. The only reason why for example the Drupal community has grown on Gittip is because it was a pre-existing, highly connected social group outside of Gittip. Small communities are being totally ignored, maybe even actively hampered on Gittip by the blind insistence that we should be using some sort of magic number of 150, and yes I know the theories behind it but I also know its not the only number mentioned. In effect, Gittip is not forming communities at all, its not even growing existing communities. Instead it is nothing more than yet-another-list-of-pre-existing-communities. The only reason why the whole Drupal thing worked was because a single brave individual had enough trust in his existing community to live full time on Gittip profits. Gittip communities are supposed to be about creating and growing new and existing groups by allowing them to self-organize... which is being actively hampered by the 150 people barrier in my opinion, causing people to be unable to talk to each other as well as the total lack of a real community page. My suggestion is to drop the 150 limit (and all other limits) completely and implement some community supporting features like:
Gittip doesn't need to provide forums or mailing lists or repositories, etc.. Other sites already do that. However, Gittip could and should provide a central profile / hub page that has links to various resources. If you want communities to be self-organized, let them organize themselves! If you don't want people to be able to do things like this because you feel its already done better by different sites, and maybe you'd be right, then don't pretend we have communities or are a social platform. Gittip would in this case be nothing more than yet-another-payment-platform. Disclaimer: I do love the Gittip concept to bits and don't want to offend or annoy with my comments but I wanted to be clear even if it might offend or annoy. I hope I am clear.. By the way... I personally feel the blog post that was formed in #1358 was a little premature... at best. |
|
@mvdkleijn I agree with you. If the 150 person mark was intended to gamify community building, it's very likely to have had the opposite effect and what we see as success now is likely no more than confirmation bias at play. Exclusivity hurts fledgling communities, especially as it stands now, where small communities are completely opaque. We might be able to properly gamify this by showing comparison stats between communities like the total giving & receiving, plus the number of individuals, teams, and other "plural" accounts. |
|
@mvdkleijn Just a quick note - I like your suggestions as to where to go with the communities feature. |
|
We talked about this at length at the Gittip retreat (video pending). We ended up thinking that we should drop the limit to something low (15 - 50, depending on data -- what's the 80% point, e.g.?), and then stop thinking about communities for a while, because the communities feature is lower down on the priority list from giving and profiles. |
|
I'd suggest to keep the 150 for "fully formed communities" for the browse page but show something also for not fully formed communities. I think without it the page https://www.gittip.com/for/ will loose a lot on usability. |
|
I'd still prefer to drop the limit (for unformed communities at least) all together. The fact that only(!) 18 out of 763 (2.36%) communities on Gittip managed to reach the limit at all should tell you it's not working. Don't interfere with communities at any level. Just give them the tools, remove the artificial limits and give them the possibilities and they'll do it themselves. (like we wanted them to, right??) (heck... giving out badges would be a bigger incentive than a pseudo-random number of 150 or even 10) |
|
@mvdkleijn the reason (still up for debate) that we decided on still having some limit is to prevent you being able to clearly distinguish the exact amount an individual is giving to someone else in a very small group (of say 5 people). After sleeping, I don't think that's going to be a common issue, especially since the stats we show on community pages aren't scoped to that community itself. Regarding @zwn's comment above, he and I were just talking about this. His concern is that a lot of the UI on https://www.gittip.com/for/ will no longer make sense, and would require a lot of developer time to update. We agreed that simply showing the top 20 largest communities is a good substitution for the "viable communities" section. |
I would hope this is going to be addressed at some point; it makes little sense to me why a patron's total giving is listed in every community it is involved in. For example MaxCDN is listed as the top giver in the Python, Node, and Javascript communities, at $292.50. They might be tipping $10 total in Javascript, but they still show up as the most influential patron there. I would really like to see this fixed at some point, because I'd like to see who cares about each community. This is particularly important with smaller communities.
This sounds good. Has anyone skimmed through the list of communities to see if there are redundancies? |
|
The issue to scope the giving and receiving is #963 and we have agreed that it is important to fix (we gave it 3 stars). |
|
I've added this to the "Stabilize Communities Feature" milestone. |
|
@seanlinsley in IRC re-proposes we drop the threshold entirely (this was floated at the retreat as well, as well as above on this ticket). Since there is some concern about inferring who is giving what to whom at low Ns, how about if we show a simple listing below, say, 50 members, and only show the existing three sorted listings above 50? That's basically sticking with some concept of viability but not waiting to add more features above the threshold before we add something below the threshold. |
|
👍 |
|
👍 |
|
Yep, that sounds good to me. And it would work well with #970 @whit537 what about #970 (comment) -- https://www.gittip.com/for/ ?
UPDATE: Ah, and then I actually read the original post for this ticket and see you already covered that:
|
|
I've changed NMEMBERS_THRESHOLD to 50. That gives us 41 viable communities on the |
|
You have an interesting definition of 'too bad' 😓 But at least there's more visibility now 🐙 |
|
Reminds me of http://www.dmoz.org/. <:-) |
|
FAQ updated in #1904. |
|
I think we should close this once the FAQ update lands, and reticket additional changes. |
Simplify FAQs re: communities; #1316
Now that we've dropped the threshold for "viable" communities from 150 to 50 (#1316), that throws a wrench in the "Congratulations!" text we had for the first communities to reach the 150 mark on Gittip. After talking with @hongminhee in [IRC](https://botbot.me/freenode/gittip/msg/9818883/), we're dropping the "Congratulations!" text entirely.
|
Closing. The threshold is now 50. Showing a plainer listing for < 50 is reticketed as #1905. |
I thought we had another ticket for this but I'm not seeing it. Came up with @jmathai on Twitter.
Per discussion at the Gittip retreat and on this ticket, we're going to drop the communities threshold way down, to see if that doesn't encourage greater participation in communities and in Gittip as a whole.
Let's try 12.Let's try 50. Let's try showing a simple listing by date-joined-community for under the threshold, and the current three-column layout for over the threshold. Tasks:[ ] change the community page to show a simple listing by date-joined-community for < THRESHOLDreticketed as change community page to show simple listing for < THRESHOLD #1905Old: The 150 number is working well, in that we now have eleven communities on Gittip and several of those have happened because of an effort to reach the 150 number (Korea being the first such, Drupal the latest). I'm not opposed to adding teasers at 50 and 100 members if it makes the process go faster. Seems like this would be good to revisit after we add more value at the 150 tier.
The text was updated successfully, but these errors were encountered: