Website ranks fact and fiction

Recently there has been a list of the "Top 100 Genealogy Sites" making the rounds. You would be led to believe that website rankings were an indication of popularity. But there are some very misleading and confusing conclusions that are drawn from a raw listing of sight rankings based on any of the online ranking sites such as Alexa.com, the one used in this particular ranking. See "Top 100 Most Popular Genealogy Websites - Genea-Musings is #90" by Randy Seaver and the original site from GenealogyInTime Magazine.

I make these observations by examining the listed rankings and comparing them to actual rankings obtained as of the date of this post. I do not want to detract from either the list or the methodology in compiling the list, and I certainly do not want to detract from the popularity of any of the sites listed, especially the blogs. I think it is really amazing that anyone can present information so well as to be that popular and they are all to be congratulated for their hard work.

But if you read any of my posts, you know I like to dig in a little and see what is really going on. There is a commonly quoted statement about statistics from Benjamin Disraeli used to bolster arguments that refers to statistics in an uncomplimentary way. As Mark Twain (Samuel Clemens) is quoted as saying, "Figures often beguile me," he wrote, "particularly when I have the arranging of them myself."

So, let's look at the list and at the website and see what we can learn.

The first thing that jumps out at me is to check the rankings today, rather than whenever they were done. Randy Seaver's site Genea-Musings, was rated #90 with an Alexa site measure of 298,451. I would submit that Randy's actual web traffic is much higher than reflected by what GenealogyInTime said was the measurement:

• Number of visitors to a website.
• Number of page views per visitor to the website.
• Amount of time spent at a website.

The reason is simple, many blogs are read in a reader, such as Google Reader, and there is not an actual click-on visit to the original blog. Because of this simple issue, Alexa's ranking are heavily skewed towards more traditional websites. The difference can be significant. Think about it, when was the last time you opened a link from your Reader, even for one of your favorite sites? But how many times have you read Ancestry.com in a Reader? The point is obvious.

Next, there is another simple problem with any list based on websites vs. blogs, money! If you pay for an Ancestry.com connection, you are probably going to use it as frequently as you possibly can. That fact does not diminish the value or popularity of Ancestry.com, but it does show how they can get higher numbers than a site that needs to attract readers every day through content.

Also, look closely at the list, #1 Ancestry.com is also #8, #14, #18, #31 and #77. These sites are all interrelated, owned by the same company and backed by the same advertising effort. Is this fair to count them as six different websites for ranking purposes? Maybe it is, maybe it isn't. But remember to add in Genealogy.com at #7, MyFamily.com at #10, Fold3.com at #17, Family Tree Maker #53 and others I might have missed, all Ancestry.com companies.

Now let's look at the list. This is an international list, so we are going to have companies and websites that we, individually, no matter where we live, will not be familiar with. But there is a problem with what GenealogyInTime Magazine says that it did and what is on the list. For example, they say "We focused on websites whose primary function is genealogy." However, several of the sites on the list are not primarily genealogy based sites. Look at NewsPaperArchive.com #12 on the list. Arguably, this subscription site is useful to genealogists, but they are only mentioned in small print at the very bottom of their startup page. GenealogyBank.com another newspaper archive is listed but its parent company, NewsBank.com which is one of the largest online repository of newspapers is missing.

The point is that how many sites and what order they are shown in is determined by the way you count sites and whether you consider them to be "genealogy" sites or not.

What about the numbers. If I look at Genea-Musings rank in Alexa.com today, Randy shows the Alexa Traffic Rank of 271,555 which, using the numbers from the older list (not perfect but interesting) would put him at #80 on the list, ten steps higher. If all of the Ancestry.com sites were counted as one (at #1), then he would be ranked somewhere around #70!

What about those at the bottom of the list? The last one listed on the Top 100 list was a site called, "Immigrant Ship Transcribers Guild" list as having an Alexa.com Traffic Rank of 380,841. Today, they are listed at 401,336. Where did the higher numbers come from?

I could go on and on and on, but don't get too excited about web statistics.