Some Web Usage Statistics for
the Kids' Cicada Hunt website

For the last six and a half years or so, I've been keeping track of how
many Web users visit my websites, using data recorded here:
http://www.saltthesandbox.org/webstats/
During the periodical cicada emergence of 2004 in the Eastern United
States, I noticed a surge in visits to Kids' Cicada Hunt: There were more
than 8,000 visits to the home page during one week in mid May. Most
years that page gets less than 200 visits a week at that time of year.

So, I've been wondering what will happen during this year's emergence
over a much smaller area of the Midwestern United States. I made a
graph to show usage patterns over the last five years and maybe help
predict what will happen.

The graphs below track visits to three pages on Kids' Cicada Hunt:
The Kids' Cicada Hunt home page, which turns up in the results when
     users search for "cicada" or "cicadas" on Google or Yahoo.
Our Periodical Cicada page, which is a result when users search for
     "periodical cicada" or "seventeen year locust."
Our page about hunting for Cicada Killer Wasps, which is a result
     when users search for "cicada killers" or "cicada killer wasps."
The top graph is scaled to show all the data. With the bottom graph,
I cut off the top of the 2004 peak in order the highlight patterns during
other years.

The current graphs track data through May 26. We got the first, scattered
reports of 100s of cicadas emerging on the evenings of May 18 and 19.
Cicadas started emerging in really large numbers during week 20 (May 20-26)

Here are some patterns that I consider interesting:

* During years when there is an emergence of periodical cicadas over a smaller
area of the United States (like 2002 and 2003), there was a peak in visits to the
Periodical Cicada page in May (about week 21), but relatively less effect on visits
to the Home Page, which usually peak in August (about week 33). The Periodical
Cicada page also peaked in August, even though it was annual cicadas that emerged
at that time.

* During the big Eastern U.S. emergence year of 2004, visits to both the Home Page
and Periodical Cicada pages started to increase in February, which is when news
sources (on paper and Web) started running stories about the impending emergence.
However, the peak for both pages came in mid May (week 19), once people could
actually see and hear periodical cicadas.

* Visits to the Home Page usually peaked in August, even though annual cicadas
started singing in late June and early July in northern states. August was when
annual cicadas started dying and falling out of the trees. I wonder if "seeing the bug"
was what drove most people to search the Web for cicada information? Cicada
sounds alone (and shed skins of nymphs) may not have been enough to get most
people on the Web.

* As a test of test of my "seeing the bug" hypothesis, cicada killer wasps start
showing up in mid July around here (about week 29), and visits to our Cicada Killer
page peak immediately.

For comparison, we can look at some usage data from the CicadaMania website for
2004. Follow this link to see a graph of "number of people who use the address
www.cicadamania.com to access the site" from April through early June, 2004:
http://www.cicadamania.com/cicada-mania-stats.html#spike
It's interesting to see the peak reached so quickly once the cicadas emerge (even
with all the advanced publicity reached by this site) and then tapering off in an
asymmetrical way. The Kid's Cicada Hunt data is also asymmetrical, but it's smoothed
out because it's displayed week-by-week, and because the peak was reached at the
boundary between two weeks.

For a different sort of comparison, try this graph of blogs using the word "cicadas" per
day for this year, 2007:
http://technorati.com/chart/cicadas
It looks best if you set it for the past 90 days.   Here's a sample downloaded on May 26:


Most of the early peaks for 2007 were associated with news articles in major sources,
like the Chicago Tribune (April and early May peaks). Once the cicadas started emerging
on about May 20, the number of posts on Technorati really went up over the next few days.
However, interest quickly waned.

So, what's going to happen during this spring's emergence? The last week shown on
the graphs is week 20 of 2007. In 2004, visits to the Home Page peaked at 8204
during week 19 and dropped to 6758 during week 20. Now the question for 2007 is,
has the number of visits peaked as of week 20 (May 20-26), or will visitation be
maintained as cicadas continue to emerge at different locations over the next week
or so?

Complicating any predictions, the Google page ranks for the two pages have
changed since 2004. Kids' Cicada Hunt used to be on the first page of results
when searching for "cicada" or "cicadas;" now it's on the second or third page.
Our Periodical Cicada page sometimes shows up on Google's first page of results
for "periodical cicada" or "seventeen year locust." Of course, lots of other things
have changed, too. More users have broadband and are on the web more often,
and "old web" pages, like Kids' Cicada Hunt, have potential competition from
"new web" blogs, social networking sites, and so on. Or maybe the "new web" will
drive visits to our old web pages.

I guess we'll just have to wait and see how many visitors we get this spring on
Kids' Cicada Hunt.

Click here to return to our main Periodical Cicada page.

The following links take you to pages about Annual Cicadas:

Hunt

Find

Solve

Link

Home

Cicada Hunt! is part of the Salt the Sandbox Web.
For more information visit the Salt the Sandbox home page.

This page was created on March 31, 2007, and it was last updated on May 27, 2007.

Some Web Usage Statistics for the Kids' Cicada Hunt website

Some Web Usage Statistics for
the Kids' Cicada Hunt website