Heck Of A Guy

Breathes there a blogger, with soul so dead,
Who never to himself has said,
Is anybody out there reading this stuff?

Who Reads Heck Of A Guy?

And Who Cares?

Most of us who use internet real estate, whether in the form of a traditional website or blog, to sell ferrets, support or oppose a politician, promote or criticize a point of view, sell advertising space, or entertain & enlighten the populace, want to know if anyone is paying attention.

Often, the number and type of readers has an impact on monies earned or success of a campaign. Those catering to a target population can orient their content and format to match the preferences of those viewers. Connection-minded bloggers want to know who is linking to their site. Businesses need to optimize their sites for search engines. If nothing else, almost everyone wants to keep score.

Certain readers may declare themselves by placing orders for truckloads of ferrets, sending huge checks for ad clicks, forwarding an adulatory e-mail or comment, sliding envelopes stuffed with small, unmarked bills under the door, forming a cult, purchasing gifts either spontaneously or from a blogger’s wishlist, or proffering explicitly described sexual favors.

To capture a picture of an audience beyond such rabid fans, however, one must typically rely on statistical devices that count, collate, and catalog whatever enters a given internet entity.

One can, in fact, usefully think of these website counters working much like home alarm systems. The simplest models are indiscriminate, setting off sirens when the family cat walks across a motion sensor or counting a spambot as a loyal reader. More sophisticated systems provide information that may be so complex as to be misleading or difficult to decipher.

But in the ordinary course of events, stats are all we have – and when your only tool is a screwdriver, you can be sure that somebody is going to get screwed.

At Heck Of A Guy, the stats show that …

The Hits Just Keep On Coming

The previous Heck Of A Guy 2006 post, Heck Of A Guy 2006: Reflections, laid claim to certain findings, including “the Heck Of A Guy blog had 349662 hits in December 2006.” That statement, while accurate, would have been more complete written as

The Heck Of A Guy blog had X hits in December 2006, where X = any of the numbers provided by a variety of statistical services. One such X is 349662, supplied through the courtesy of Webalizer.

Another X, no doubt quite nice in its own way, is 174307, which is the number of hits recorded by Awstats for that same time period.

Don’t like those numbers? Not to worry; several other choices are available from other services using other methodologies.

Oh, and there is one additional line that would have made that paragraph about Heck Of A Guy blog statistics even more complete:

Of course, precise hit counts are moot since hits are nearly useless as a website statistic

The Quick & Dirty Web Stats Tutorial

Consider the following diagram:

If you determined that this diagram has nothing to do with website statistics, you would be almost correct. This diagram has nothing to do with website statistics – except this: its usefulness in describing a website’s audience approximates that of the nifty schematics and graphs those snazzy website stats programs produce if the user doesn’t know some basic statistical incantations.

Happily, a few rudimentary concepts and definitions are sufficient for our purposes vis-à-vis the Heck Of A Guy blog and also for exposing some common fallacies about indicators of website popularity in general.

1. Pageviews trump hits. The number of times a page is viewed ( i.e., pageviews) may be relevant. The number of files viewed (i.e., hits) is useless. If, for example, a page has 28 pictures, then viewing that page generates 29 hits (28 for the pictures and 1 for the html file). One page view could produce hundreds of hits.

2. Visits trump pageviews. Visits are the number of times a specific page is viewed by different visitors. A visitor may view the same page multiple times during a visit; each of those viewings would count as a pageview but one visitor can only generate one visit on one page.

3. It is useful to differentiate among visitors. Legitimate visitors are differentiated from spambots, search engines bots and spiders, and administrative traffic. Unique visitors (i.e., different individuals) are differentiated from visitors who return repeatedly. First Time Absolute Unique Visitors are those visitors (counting each visitor only once) who visit a site for the first time during a given date range.

4. The list of details that website statistics can provide is long; life is short. Stats can be compiled from analysis of log files or by tracking a java script embedded in the web page. They can be provided in real time or compiled for any unit of time. Visitors can be described by geographical location, time spent lingering on a given page or at a given website, browser use (e.g., Internet Explorer, Opera, Mozilla), which links are clicked, referring sites, search terms used, purchases made, how long it’s been since the most recent visit, the date of the visit, the time of the visit by time zone of visitor’s origin, etc etc etc. It’s probably useful for anyone using the internet to know that these capabilities exist. For our purposes today, however, most of this data is not essential, in large part because …

5. The trend is your friend. A preeminent principle in dealing with website statistics is that it is advantageous to focus on trends rather than absolute numbers. The idea is that it is more important to know if a website audience is growing over time rather than to know if that site had 10, 100, or 1000 visitors during a specific 24 hour period. While this notion has obvious limitations (e.g., a site that only has one viewer its first month but increases its audience by 1000% in one year still has only 11 viewers to show for a year’s effort), it does seem valid in the long run.¹ One of the benefits derived from this concept is that many of the differences in details and much of the ambiguity wash out when overall trends are spotlighted.

Hit Me With The Digits

Note:
1. The screenshots of the charts shown below were taken during the first few days of January 2007. Consequently, December 2006 is the most recent complete calendar month for which statistics are available. Because those screenshots are already more than a week old, any January 2007 numbers shown are already outdated and should be ignored.

2. Most of the charts are thumbnails and a larger version can be viewed by clicking on them.

3. The earliest available Heck Of A Guy statistics are those from early to mid-August 2006 (depending on which package is being viewed). From March to August 2006, the Heck Of A Guy blog was hosted at another site. Statistics for that period are not available.

The Trend IS Our Friend

All of the statistical packages, in fact, show the same trend: an increase over time in visits, pageviews, and visitors at Heck Of A Guy.

The Webalyzer stats, segmented by months, follow:

And the Awstats charts are below:

The Visitors & Their Visits
For advanced novice bloggers, such as DrHGuy, Google Analytics may well be the coolest and, arguably,² one of the most accurate free statistics package. These are especially helpful in identifying visitors to the site.

The first graphic below is the Google Analytics graph of the December 2006 pageviews of the Heck Of A Guy blog and the second is the Google Analytics graph of visits, both with traffic from spammers, search engines, and my own work filtered out.

No big surprises here. There were 3510 pageviews in December dispersed over 2,055 visits.

Looking at the type of visitors is more revealing. The next graphic shows the number of different visitors each day while the following one represents the percentage of new visitors.

And this is the percentage of absolute unique visitors over the month of December expressed in a pie chart.

Yep, 97% of all the visitors to Heck Of A Guy in December were first-timers – DrHGuy-naïve, virginal, unexposed, unspoiled, and unsullied innocents.

Indeed, the trend has been and continues to be that not only have visits, pageviews, and visitors increased over time at Heck Of A Guy, but the percentage of new visitors has also increased.

I had expected, for reasons that now escape me, that most readers would be from the same group of folks who would check in, if not every day, then every 2nd or 3rd day. That is the case for about 15-20 viewers. The other 1400+ visitors in December, however, were new visitors.

Other Website Stats Tricks

While, as mentioned earlier, much of the information website stats can provide is only tangentially significant to a non-revenue producing blog such as this one, at least one of these non-essential details does impress and amuse my inner eighth-grader sufficiently that I’ve persuaded myself to coercively share it with you, gentle reader – who, I now belatedly realize, is statistically likely to be a first time visitor here, so let me take a moment as host to say …

Howdy, Stranger,

How are you this fine day? Pull up a chair & sit a spell. I’m sorry my blog is such a mess today, but until I checked these stats, I wasn’t expecting so many new visitors. Anyway, if you see anything you like, feel free to click around or leave a comment. You might want to check out some other posts listed under “Favorites” in the sidebar. If you found anything useful, I hope you tell your friends. If not, Godspeed & all that.

Y’all come back, now. Y’hear?

DrHGuy

Anyway, it turns out that it’s pretty simple to determine the geographical location of a viewer’s computer. It happens that, as one might expect, Google Analytics does a nice job of presenting that data in a graphical format.

The map below represents those December 2006 visits to Heck Of A Guy. Each blip indicates where the visit originated. While this appeals primarily to my pen-pal mentality, it is clearly useful, say, if one is marketing air conditioning units, to know if the message is getting across to oil-rich desert countries.

But there’s more.The map is actually a bubble chart. The size of each blip (i.e., bubble) is proportionate to the number of visits from that individual. Check out this blow-up of the same map.

But there’s still more.Hovering the cursor over a bubble provides a legend. Behold, the Durham, North Carolina bubble, for which, one assumes, the Duke of Derm is primarily responsible.

Other Factors That Affect Stats

Nothing is straightforward. For example, …

The number of viewers on a given day fluctuates by the day of the week, regardless of content. Viewers using internet access points at places of business obviously decrease on weekends. Home users also decrease on weekends and around holidays although not as dramatically. Search engines may index some sites several times a day, other sites once a month, and still others not at all. If a wonderful post is written on the 3rd but doesn’t find its way into the Google index until the 17th, its peak readership is likely to be at least two weeks after its publication. The Yahoo and MSN search engines indexed Heck Of A Guy only a couple of weeks after it opened, but Google didn’t include it until two to three months later. The accursed spammers and unrelated technical snafus can make a website inaccessible to some or all potential viewers. And the beat goes on.

Information Not Provided By Website Stats

Disappointingly, I’ve been unable to locate any Statistical Programs that reveal a viewer’s mood, literary, musical, or sexual preferences, disposable income, clothing (or lack thereof) worn during the visit to this site, capacity to concentrate, extent of vocabulary, sense of humor, number of siblings between 5′ 1″ and 6′ who speak at least two languages, or situational context (e.g., visiting this site to complete a school assignment, to improve him- or herself, to avoid doing the work for which he or she is being paid, to thwart boredom, …).

Next

The next Heck Of A Guy 2006 post will focus on content, including how viewers find Heck Of A Guy, what they read once they get here, which posts are hits and which are misses. Following that will be the Heck Of A Guy blog report card and remedial blogging-education plan.

1. Of course, as John Maynard Keynes pointed out, “In the long run, we’re all dead.” [↩]
2. In any discussion of website stats, “arguably” is inevitably a keynote [↩]