4
$\begingroup$

A while ago it was asked

How to increase our traffic?

It seems that we have a steadily growing number of visits recently. At this site I was able to generate a plot of the traffic over time:

enter image description here

I am interested in the explanation of this graph. First of all, how reliable is this data? And second, what might be a reason for the increased traffic?

  • the events at the peeks (whatever happend on Aug 11) ?
  • the start of the winter term in most universities?
  • did we just reached the critical mass?

Can we learn something from this?

$\endgroup$
4
  • 1
    $\begingroup$ The Quantcast data is reliable as SE uses Quantcast on all sites, it is not estimated data like for many other sites. The spikes are likely from posts shared to high-traffic sites like reddit or Hacker News, but without the mod-only analytics it's hard to know the specific cause. $\endgroup$
    – user4342
    Commented Oct 24, 2012 at 12:45
  • $\begingroup$ would like to know the peaks also & there are no answers on that so far. seems possible it is due to just one question or maybe one citation of the site in a high-traffic location... the moderators probably know the answer or can find it.... $\endgroup$
    – vzn
    Commented Oct 31, 2012 at 4:56
  • $\begingroup$ @vzn: See Raphael's answer for the peaks. $\endgroup$
    – A.Schulz
    Commented Oct 31, 2012 at 8:38
  • $\begingroup$ yeah saw that. its not clear if most of the traffic spike went to the home page, or particular questions, or just a few questions etc $\endgroup$
    – vzn
    Commented Oct 31, 2012 at 14:39

3 Answers 3

Reset to default
3
$\begingroup$

I can not disclose details of what data we mods get, but I can tell you that it looks quite similar to what you found. Given that SE likely uses the very same data (probably together with those obtained from Google) this may not be too surprising.

The spikes are due to social media; see also here and compare with the dates on reddit. It is noteworthy that if you consider visitor counts average over seven days, all spikes caused the (before September) mostly flat visitor curve to rise sustainably. Apparently, visibility on social media does bring new regulars.

As for the reason for the latest increase, we can only speculate. Winter term started at universities and many questions seem to be study-related, true, but last summer term did not cause a comparable increase. Therefore, university schedules can't be solely responsible. People with more experience across multiple SE sites may be able to offer more.

Our traffic by search engines is still at only 60%; according to SE folks, that figure should in the nineties eventually.

$\endgroup$
5
$\begingroup$

Moderators have a bit more detailed statistics. We aren't allowed to share the specifics, but I can tell you that the general trends agree with the public observations. In most available measures (number of posts, visits, new users, feedback from anonymous users, …), we have gone through three phases:

  1. The initial excitement of the early beta, with lots of activity.
  2. A lull from mid-April to late August.
  3. A noticeable increase since early September. Our number of visits is especially growing, and the growth of the number of posts is clearly visible.

These phases are pretty much de rigueur in the evolution of a Stack Exchange site. It's likely that the start of the term is helping us grow at the moment as well.

We're still getting a significant part of our traffic by referral rather than by search engines, but not alarmingly so. I won't say which sites because tha tand the referral statistics are known to be buggy besides, but nothing unexpected stands out anyway.

$\endgroup$
2
  • $\begingroup$ Let me point out that on cstheory the search engine traffic grow very slowly and even now it hasn't got to the point that is comparable with SO (>90% IIRC). As long as the rank of the site in search engines is improving we shouldn't need to worry much about this probably. $\endgroup$
    – Kaveh
    Commented Oct 25, 2012 at 5:22
  • $\begingroup$ @Kaveh Is there a way to measure search engine ranks but testing some arbitrary queries? $\endgroup$
    – Raphael Mod
    Commented Oct 25, 2012 at 8:09
5
$\begingroup$

I advertised the site to my students a few weeks ago. There are 60 of them. This accounts for some of the visitors.

$\endgroup$
2
  • 1
    $\begingroup$ So we should wish that your lecture is bad so that many of them come here? ;) $\endgroup$
    – Raphael Mod
    Commented Oct 29, 2012 at 14:55
  • 6
    $\begingroup$ I'll do what I can. $\endgroup$
    – Dave Clarke
    Commented Oct 30, 2012 at 8:17

You must log in to answer this question.

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