15

This Question is inspired by the now Closed Question on Meta Stack Overflow of Why is Stack Overflow so negative of late? and the write-up of Podcast #59 – The Decline and Fall of Stack Overflow.

If you visit https://www.quantcast.com/gis.stackexchange.com and click VISITS at the top, then All under the graph, you should see a graphic like below.

enter image description here

What is very evident is that in the past 3-4 weeks our number of visits has jumped from around 28,000 to around 33,000 per day.

Why has GIS Stack Exchange become so popular of late?

  • 1
    Maybe the Google Panda Update (4.0) ? searchengineland.com/google-begins-rolling-panda-4-0-now-192043 was on 19/20th May 2014). It likes recent changes on updated websites like Stack Exchange. – Mapperz Jun 6 '14 at 17:10
  • 3
    A small impact the is a weekly Twitter discussion called #gistribe twitter.com/search?src=typd&q=%23gistribe each re-tweet is going to approx 10,000 GIS users. I have been joining in with some hints to GIS-SE. – Mapperz Jun 6 '14 at 17:15
  • +1. As a side discussion: one interesting thing is that until half of first semester of 2013 GIS SE was growing very fast in visits and from one year behind to nowadays it decelerated (exception for the 3 last weeks). What changed after April, 2013? – Andre Silva Jun 6 '14 at 18:37
  • @Mapperz Intuitively it would make sense we are amongst the "winners" from the Panda update - pity no Stack sites are high enough (yet) to figure at blog.searchmetrics.com/us/2014/05/21/… – PolyGeo Jun 6 '14 at 23:54
  • @AndreSilva That side discussion could be worthy of a separate Q&A but when the stackoverflow.com graph equivalent to the one on my Question is examined a similar although shallower dip post April 2013 is seen, but no recent spike. – PolyGeo Jun 6 '14 at 23:58
  • 4
    I started using arcpy.da a few weeks ago, which would account for at least a few of those visits... – Stephen Lead Jun 7 '14 at 9:20
  • Awww, GIS.SE gets time off for Christmas, too! – nmpeterson Jun 17 '14 at 0:32
10

I don't know the answer, but I can share some additional data available to mods.

  • SE's internal analytics confirm a sudden jump in visits that began three weeks ago (on Monday, May 19). It amounts to an extra 25000 visits per week--about a 15% increase.

  • There has been a comparable change in numbers of new visits.

  • The amount of feedback by low-rep/anonymous users might have jumped at the same time (but the data are noisier so it's hard to tell).

  • The increase occurs only on weekdays, not on weekends.

  • It has not been accompanied by a detectable increase in the number of posts (neither questions nor answers) nor by any unusual increase in the rate at which new users join.

  • There is no evidence that extra visitors are being delivered by any search site or referring site.

  • The following week (May 26 - 30) there was a large spike in newsletter subscriptions--but it added only 15 to 20 more users than one would have expected based on previous rates.

  • There is a gradual acceleration in the numbers of edits being performed. It was not a sudden jump, but nevertheless has increased over the last three weeks to a record high.

It looks like a group of people may be passively using our site on a regular basis and perhaps are starting to become more active. We should be proud that they find ongoing value in it.

  • 1
    +1 I may be cherry picking your dot points but I have a suspicion that no "detectable increase in the number of posts (neither questions nor answers)" and a "gradual acceleration in the numbers of edits being performed" may mean that the overall quality of Q&As (including their titles) here is lifting and becoming more attractive to potential visitors when spotted amongst their search results. – PolyGeo Jun 6 '14 at 23:37
  • Is it just me or has the number of bot-induced vandalism reports also increased? Might account for some visitors as well... – Curlew Jun 8 '14 at 18:17
9

My hypothesis: May is the beginning of the field season for many GIS professionals in the northern hemisphere. The highlighted peaks show times corresponding to instances where individuals are finalizing their methods, techniques, etc they will be using throughout the summer. As a result, there is much more activity on GIS.SE. These peaks are simply outliers on a linear trendline of growth.

enter image description here

2

The reason that stackexchange GIS have a steep rise in visitors at last few weeks is that IBM is using search engine with Watson on all Stack Exchange sites, they are using it for internal needs of daily work, when a IBM worker need help he ask Watson that search answers at S.E - choose the best answer and even change it to be better

enter image description here

  • 2
    How do you know that Watson started visiting GIS SE and DBA SE in the last 3-4 weeks? If you can include a link that would be very helpful. Why might SO not have a similar spike? – PolyGeo Jun 10 '14 at 4:30
  • From the Answer of @whuber - "The [15%] increase occurs only on weekdays, not on weekends." - I think it unlikely (but not impossible) that Watson works only weekdays. – PolyGeo Jun 13 '14 at 7:02
  • Hi PolyGeo - The reason that Watson work only weekdays is that they are using it for internal needs of daily work, when a IBM worker need help he ask Watson that search answers at S.E - choose the best answer and even change it to be better. – Geog Jun 15 '14 at 7:19
  • 1
    @Geog -- You have still failed to give any evidence that this is the case or provide any metrics as to how many searches Watson is allegedly doing on the site or if its search volume is statistically significant. And I'd like to see a link to an answer that was "fixed" by Watson, I have personally not seen any. In your comment you're reasserting what you've already said without giving proof; you haven't made a convincing case for your claims. – Jason Scheirer Jul 26 '14 at 19:35

You must log in to answer this question.

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