What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

One of the interesting things about Google is that they tend to have products in beta for a very long period of time, despite the fact that they for all purposes ceased to be a beta product some time earlier. The single best example is Gmail, which was in "beta" for years, despite it feeling far more like a finished product.

I'm beginning to wonder if we are starting to follow the same track. It seems like the site betas are getting longer and longer, to the point that right now we have 2 sites that are in beta after over a year of work. Even popular sites, like Skeptics, Personal Money, and On Startups, are looking to extend this period. Compared to when I first joined Stack Exchange, the amount of time in beta seems to have doubled or tripled.

We've had 3 sites achieve full status recently (Database Administrators, Bicycles, and IT Security) but aside from that, it's been 4 months since a site has been fully released. I know in the end it doesn't really matter all that much, but it would be nice in a lot of ways to have more sites be promoted.

In my understanding, there is several benefits which promoted sites receive. Perhaps the largest is the ability to elect their moderators. Others include a custom theme, restricting some of the privileges to the core group of users, being subject to the Stack Exchange Data Explorer (And thus being able to run cool analysis scripts on the site), and just being recognized as an official site.

So, I guess all of this comes down to, are betas becoming more difficult to finish, is that a good thing, and why?

share|improve this question
    
Beta doesn't mean "unfinished product". It means "not ready to launch as a full site". Related: Does this site have a chance of succeeding? –  Cody Gray Aug 17 '11 at 13:50
2  
...in beta after over a year of work. You should check out Board and Card Games. We've been in Beta for 2.4 years! –  ire_and_curses Feb 28 '13 at 20:55

4 Answers 4

Sites are probably taking longer to launch because:

  1. In addition to the technology overhead, there's also human overhead in maintaining them all, and SE needs to hire actual paid employees to manage the ever-growing network. The growth of new sites has to keep pace with the growth of the SE team.

  2. Some betas just do not have enough traffic (or other metrics) to prove themselves viable yet. SE's policy is not to launch a site unless it has a reasonable chance at success.

  3. A lot of betas that do have traffic are still suffering from quality problems. One of the things we all started realizing a few months ago is that betas actually need to be watched very carefully so that they don't go off the rails. When a site with poorly-defined/ambiguous/disputed scope goes live (*cough*Programmers*cough*), you get... problems.

I'd rather see a steady trickle of really good sites go hot and survive than dozens upon dozens of sites spring into action for 30 days and then wither away.

I think that some dampers have been placed on the Area 51 staging process as well, because part of this "perpetual beta" problem is that way too many sites were going into beta, way too fast, without any real commitment or interest from the community, but rather propped up by Area 51 rep junkies.

share|improve this answer
    
I think you might well be right, the problem seems to be in the Area 51 staging ground. Thanks for your answer! meta.stackexchange.com/questions/80106/… –  PearsonArtPhoto Aug 17 '11 at 14:04
3  
#1 does not factor at all, #2 is true, #3 only applies a little and mostly to the SE 1.0 imported sites (money, startups) which have inherited diseases from the broken home they came out of. Also no "dampers" have been placed on Area 51, except for a mild decay in commits over time so zombie proposals don't stumble across the finish line. And that was, geez, at least 5 months ago? –  Jeff Atwood Aug 18 '11 at 9:54
up vote 14 down vote accepted
+50

I asked this question 3 months ago, and I thought I'd post a few tidbits of some updated data.

Since that post, there have been 2 sites that have graduated, Drupal Answers and Sharepoint, each of which was in beta for about 6 months. There are currently 13 sites which have been added in the last 90 days.

Travel, launched in Jan 2013, has the following statistics:

  • 5.4 questions/day
  • 100% answered
  • 427 with 200+
  • 45 with 2000+
  • 32 with 3000+
  • 2.1 answer ratio
  • 4306 visits/day

The statistics for the two sites promoted at the end of 2011:

Drupal Answers

  • 24.9 questions/day
  • 87% answered
  • 191 with 200+
  • 12 with 2000+
  • 5 with 3000+
  • 1.7 answer ratio
  • 3524 visits/day

Sharepoint

  • 20.9 questions/day
  • 91% answered
  • 171 users with 200+
  • 24 users with 2000+
  • 12 users with 1000+
  • 1.8 answers/question
  • 5071 visits/day

Let's compare this to a few sites released one year prior, at the end of 2010. Remember, the Area 51 statistics stop after the site is fully released. I'm going to compare TeX and Statistical Analysis, chosen at random from the list.

TeX

  • 12.7 questions/day
  • 99% answered
  • 162 users with 200+
  • 21 users with 2000+
  • 10 users with 1000+
  • 2.4 answers/question
  • 1493 visits/day

Statistical

  • 7.6 questions/day
  • 96% answered
  • 138 users with 200+
  • 10 users with 2000+
  • 8 users with 1000+
  • 3.0 answers/question
  • 1159 visits/day

The statistics required to launch is growing significantly, year over year.

So, one might ask, what affect does this all have? Well, let's list what the differences are between a beta site and a fully launched site.

  • Higher Reputation requirements for just about everything
  • A custom theme
  • Periodic moderator elections, vs moderators appointed to a site within the first month.

Of these, the principal thing that I see is the last, namely the ability to elect moderators. Being able to choose who will lead the community is a big deal. Also, the people who will be nominated are likely to have stuck with the site for many months, greatly increasing their activity as a whole.

So, I'm going to say, yes, sites are remaining in beta too long.

share|improve this answer
8  
+1 and agreed on your last point about moderator elections. But I wouldn't overlook the second-to-last point: Removing the beta label and having a custom theme would also make a site feel more like a home, less like a house. People want a sense of permanence about the place. –  Chris W. Rea Nov 19 '11 at 0:16

As far as launching sites from public beta to full "statehood" -- assuming baseline quality, which we always have -- it's really just a matter of getting the sites to a point where they are growing at a reasonable rate.

Here's a sparkline of visits for twelve current Stack Exchange 2.0 sites that have been in beta for at least 90 days, have at least 10 users with 2000+ rep, and at least 5 users with 3000+ rep.

enter image description here

Can you tell which site is which? Which one would you bless with "statehood"?

share|improve this answer
4  
Can you make your images clickable ? ;) –  user150926 Aug 18 '11 at 12:10
20  
Kind of hard to know what to make of these when they don't have a scale... –  Aarobot Aug 18 '11 at 13:59
2  
I'm assuming it's showing 1k/visit, but what's the value of the highest peak? (for scale). Two look almost completely dead unless the peaks in them are significant. –  Tim Post Aug 18 '11 at 14:02
2  
Middle and right second row looks like it's getting consistent traffic. Launch those next. –  random Aug 18 '11 at 14:09
2  
I'd say the bottom full row left 2 should be soon, as well as the 3rd row left and right. It's hard to tell with the spike throwing things off for the two in the second row. Also, the top left seems to be good as well. –  PearsonArtPhoto Aug 18 '11 at 14:59
6  
Doesn't this methodology penalize sites that had abnormally-high traffic peaks due to a few well-ranking but time-sensitive questions? As the public ceases to search about a time-sensitive topic, visits to well-ranked pages drop off. I'm referring to questions like these which would have peaked in visits at three specific times (around June 2010, December 2010, June 2011) owing to changes in legislation and the dates stimulus cheques were mailed out. 179K views on one question was nice, but not sustainable. –  Chris W. Rea Nov 18 '11 at 21:10

There are parameters on releasing a site and a direct measure of time is not one of them. They all relate to the size and activity of the community.

share|improve this answer
1  
Right, but at least in my view, there are an increasing number of communities with a large size, active group, but aren't getting pushed into beta. The 3 examples I linked are examples of this, they all have large communities, are doing well, have been keeping it up for months, but aren't getting pushed beyond beta. –  PearsonArtPhoto Aug 17 '11 at 13:56
2  
@Pearsonartphoto +1. I agree. –  Chris W. Rea Nov 18 '11 at 21:28

You must log in to answer this question.

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