I removed the Clustrmaps control from my blog template this morning, as I’ve been exploring the new Blogger Stats features this week, and was pleasantly surprised to learn that my blog’s daily traffic is about 5x what Clustrmaps had been reporting. I also turned on a control for sharing pages via email, Facebook, etc.
I’m still working on the blog post series to which I threatened to subject you recently, and hope to have a couple of those posts done by early next week.
1 comment:
Hi - thanks for your posting and for your (previous) use of ClustrMaps! We just wanted to point out the reasons for the different numbers you have noticed. In particular, ClustrMaps and Blogger Stats (great service) have not only different aims, but also very different counting methods.
Regarding the counting methods, ClustrMaps gives you a count of 'unique visitors per 24 hour period', whereas Blogger Stats gives you 'Page Views', and the 'counting period to avoid duplicate counts for page views' is typically 30 minutes rather than 24 hours.
To illustrate the crucial difference, imagine that you had 4 different people each reading (say) five different pages on your site (or separately landing at five different blog postings), all on one morning. The counts you would see are as follows:
ClustrMaps: 4 visitors
Blogger Stats: 20 page views
Moreover, if the same people re-visited all those postings later that evening, the counts you would see are as follows:
ClustrMaps: 4 visitors
Blogger Stats: 40 page views
It is crucial to appreciate that in this case there are no "right" or "wrong" numbers, merely different interpretations and counting methods. Our counts are very accurate, and based on industry-standard Apache log files, which provide all the details necessary to compute the results (with no personally identifiable information!). The idea of a restricted time window (normally defined as a 'session') is common in the web analytics industry, but the definition of that window varies for different purposes, and is arbitrary.
There is a great discussion of the problem, along with a nice illustration, in a Wikipedia article on Web Analytics, particularly the section entitled The Hotel Problem.
As mentioned at the beginning, we also have very different aims, namely to provide
a) a 'visual gestalt' overview
b) a 'badge of pride' that you can show off simply on your site (the thumbnail map), with no further explanation
c) city-level detail (all those dots)
d) very rapid map deployment (large maps are all pre-stored for speed)
e) massive scalability (we can show millions of dots instantly, although we appreciate you may not need such numbers, it is nice to know that the capability is there).
Finally, as a gesture of our interest in your blog, we have also upgraded you to ClustrMaps+ at no charge. As a ClustrMaps+ user, you get the following extras:
1. No more ads on the big map page
2. Better 'zoomed in' (continent-level) maps (effective as soon as the next map update takes place, normally within a day or two).
3. More frequent map updates (ther is a lower 'threshold' for the extra dots required before a new map is displayed).
4. Complete maps, including the big world map and the zoomed-in continent maps, are kept permanently in the Maps Archive link (for free users, only the thumbnail map is stored)
5. Greater selection of NICE THUMBNAIL MAP STYLES: check out the much wider array of map styles and languages for your thumbnail map by visiting the 'Map Styles' tab after you login to your ClustrMaps Admin page at http://www.clustrmaps.com/admin/action.php
You can see just a few examples of some new map styles at
http://www.clustrmaps.com/map-styles/examples.html
I hope you'll take the above comments in the constructive spirit in which they are offered, and as evidence in our keen interest in each and every user of ClustrMaps. We hope to see you back on board soon!
All the best,
-CJ on behalf of the ClustrMaps Team
Post a Comment