33

I mentioned this in TL before, and it got a positive response. It has been discussed internally, but I'd like to see what the commuinty in general thinks about it.

Pre-ramble

In the past, sites used to graduate within 90 days if they showed sufficient potential, or be closed down. It seems to have been a pretty straightforward process, with not much uncertainty.

That process had its flaws (one of them was probably that 90 days is not enough to tell if a site will succeed), and was replaced with the current process where sites stay indefinitely in beta till graduation. Sites get closed if they have a consistent level of low activity or other crippling issues.

The problem with this process is that for many sites, you can't really see the light at the end of the tunnel — it's obvious that they're not in much danger of failure, but graduation is something too distant to imagine.

This goal-that-will-take-years-to-reach is a bit of a turn-off. For example, with Chemistry, I see a slow, steady growth. I can't think of any reason why it would fail in the next few years (unless all the chemists dropped everything and started working on nitrogen compounds), but graduation seems very far off too. I do have some plans to help speed it up, but not enough to make graduation come visibly closer.

This seems to be true for many sites out there. But the reasons for each are different.

For example, Bitcoin has a very high level of questions pouring in, and many, many visitors, though the community isn't keeping up as well as other sites. Music has the opposite problem: a community that is handling the incoming questions very well, but there aren't as many questions incoming. Chemistry has a bit of both worlds, though the community moderation is excellent.

(Mind you, these are conclusions I've made from looking at the numbers, but either way I doubt that all beta sites have the same problems that need to be overcome)

There are also sites that are still in beta after 3 years, and many more sites that seem to be going down the same road.

Rationale

All of this makes me think — why graduation in the first place? The original reason behind the 90d beta is to test the site out and see if it has the potential to succeed. The current beta system seems to be supposed to bring the site to a sustainable pace and then graduate. But when you look at it, many of the sites currently in beta are in no danger of being closed, ever, it's just a matter of sitting it out until it graduates. If the only road a site can go down is graduation, why wait? If success is the only possible event, hasn't success already been achieved?

Of course, graduation needs resources1. It's understandable that the site design may have to wait and be handed out on a priority basis because we don't have that many designers.

But why should the rest of the graduation goodies wait because of the design? For that matter, why should they have to wait just because a site is not even on the sure track to graduation due to a single aspect?

There are multiple goodies that a site gets on graduation:

  • Site design
  • New rep levels
  • Moderator elections
  • Community ads
  • Site footer
  • Migration paths
  • Less pressure on other sites not to migrate to the site

They all contribute to the requirements for graduation:

  • Site design (Jin must have time, and the site must be pretty well established)
  • New rep levels (Effective community moderation that won't go away on the rep level increase)
  • Moderator elections (Enough "permanent" members, a well knit community, an active-ish meta, good community moderation, see also)
  • Community ads (Views, ?)
  • Site footer (Views, ?)
  • Migration paths (Enough questions)
  • Less pressure on other sites not to migrate to the site (Established scope, no danger of "inorganic growth" due to migration)

The blocking ones here are the requirements for design, rep levels, and mod elections. But they don't all block the site graduation for some sites.

Why should any one of these hamper the others? To graduate, a site must reach a pretty nebulous set of goals, which span all aspects of the site.

But what if graduation was replaced with incremental awarding of "graduation goodies"?

The proposal

I propose that each of the graduation goodies that a site can get is awarded separately, when the site reaches a level where it can be said to "deserve" that particular goody.

For example, the site gets community ads when it reaches a high level of views (eg Bitcoin). It gets elections when the community is in need of mods and is large/stable enough to support elections (Code Review might be a candidate for this). It gets a design when it has high views, is in zero danger of failing, and Jin has time. The privilege levels go up when the community moderation can continue effectively with the new privilege levels (eg Academia). "Don't migrate to beta sites" can be replaced with "Don't migrate to sites with low activity".

This can have at least the following benefits:

  • One can set down more objective thresholds for almost each goody, as opposed to the current nebulous threshold for graduation.
  • By breaking down one distant goal into multiple goals, some of which are not so far off, the "beta"ness of a site is less demotivating.
  • Sites don't have to lose out on a feature that they can support just because an unrelated aspect of the site is not up to scratch.

Thoughts?

1. i.e. bacon for Jin. Lots of it.

  • 4
    +1 for "dropped everything and started working on nitrogen compounds". :D Oh, and because I like the proposal. – Ilmari Karonen Dec 15 '13 at 13:48
  • Especially decoupling it from the design would already benefit sites that have effectively graduated a year ago but are only waiting for design. But nice approach in general. – Christian Rau Feb 10 '15 at 16:41
  • How about some kind of somewhat more incremental tiered system. Like beta is the trial period, then there's silver, which removes the beta label, adds trivial css tweaks and other low-hanging fruit customizations, and provides all the features necessary for long-term sustained operation of the site (like mod elections), then gold which provides all the benefits of existing "graduated" sites. (Or something along these lines.) – Ajedi32 Jun 3 '15 at 17:15
  • Is this status-completed? – Raphael Jan 15 '16 at 16:42
19

I think this would just overcomplicate the whole process, there is a lot to improve about the graduation process as it is right now, but you're attacking it from the wrong side.

Some changes would be confusing, changing reputation levels independent from the design for example could easily confuse users as they can't easily tell if a site has beta reputation levels or graduated ones.

Other changes are inconsequential, the community ads or migration paths are minor issues.

Splitting the graduation would not solve the actual issue, I think it would be a distraction with not enough benefit.

The significant problems I see with the graduation process are:

  • Criteria for graduation are unclear and there is not a lot of feedback from SE unless you seriously poke them
  • The current severe bottleneck in available designer time

Splitting up the graduation into multiple parts won't help with the first point. There are some criteria that are rather subjective and can't be put into a convenient number. There is also far too much variability between sites to have any absolute criteria. A site about a niche topic has fewer questions and fewer answers than a site about a broader and more popular topic. For some sites 3-5 questions a day is a sign of trouble, for others this is good enough and not a barrier to graduation.

The designer bottleneck can only be solved by hiring another one or two designers, which SE is currently trying to do. I wouldn't want to change the rules to circumvent what is hopefully a temporary issue (though it has been a temporary issue for a while now).

What SE should do in my opinion to improve the transparency of the graduation process and to clear up some common misconceptions about it is the following:

  • Remove those stupid target numbers on Area 51. They are meaningless at best, and outright misleading at worst.
  • Make a subset of the moderator analytics graphs available instead. Especially for traffic and number of questions the trend is more important than the absolute number, so the community should be able to easily see the trend.
  • Communicate more about the regular reviews of the beta sites with the community. Just tell the community which parts are holding it back from graduation, and which parts are going well
  • 6
    We are looking at ways to improve the communication. Unfortunately when we tried it in the past, it basically came down to repeating the same things over and over on most sites and that becomes supremely unhelpful, verging on condescending, very quickly. I'm not sure that's any better than no communication at all - especially on sites that are doing well as betas. The community-driven self-evaluations were meant to try and give the community a window into how they're doing quality-wise. It certainly has had mixed results, so we're thinking of other things to try now. – Adam Lear Dec 15 '13 at 16:02
  • 7
    @AnnaLear It seems that half the time is spent on correcting misconceptions the community has about which statistics are important for graduation, especially users worrying about too few questions per day. And most evaluations boil down to "site still needs to grow, no major problems" which doesn't really fulfill the expectation of the community. I was more thinking about a boilerplate post on the meta that just explains the process if nothing is wrong, and that is updated if there is anything specific SE can say. – Mad Scientist Dec 15 '13 at 16:12
  • @AnnaLear it would also be helpful to warn sites that are doing not so well and getting endangered of being closed; telling them what the problems are, and encouraging the community to do something about it. For example I lurked on Theoretical Physics SE all the time since it was in public beta, but I never saw anybody from Stack Exchange giving them a timely word of warning before it was decided to close it. And indeed many users were negatively surprised by the closing announcement stated in a Stack Exchange blog post and finally on the meta of the site. – Dilaton Dec 15 '13 at 23:51
  • Maybe timely words of warning could have saved some sites in the past, and can generally encourage communities that are somewhat struggling to come together and put some effort into the things that are working not so great and holding them back? – Dilaton Dec 15 '13 at 23:54
  • This is being asked on Crypto, and it seems that none of us really know which criteria is holding us back in beta. Clearly I can't speak for everyone, but I would love to which aspects of the site The Powers felt 'wasn't ready' - especially if it was something we could work on. – figlesquidge Dec 24 '13 at 19:02
  • 2
12

This needs to happen, in one form or the other.

Due to the big design backlog, sites have been held in the graduation queue for over a year (maybe years in some cases?). We are too big to be beta sites, we miss the features of graduated sites, but are held back by a lack of graphics.

More concretely, what we (speaking as a moderator on Computer Science) need to have are

  • moderator elections,
  • migration paths and
  • a brand to advertise with.

Note how only the latter depends on the site having a pretty and distinctive look. I can see no reason why we can't get some features before others.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .