What’s the Best Way to Measure the Rank of a Website?
In the past couple months my page rank has dropped from 4 to 1 and my Alexa ranking has gone from 150k to 75k.
In terms of selling links the page rank drop is a negative factor and the Alexa increase [with Alexa the lower the number the better] is a positive factor.
It has lead me to think about what is the best way to value a link on a website.
Page Rank
Since my PR downgrade the amount of traffic that is being referred to this site from Google has gone steadily up. I can rank for relatively difficult keywords just by writing one well optimized article on that topic.
If you Google the recent IM product Pipeline Profits, you’ll find my site on the front page of Google right behind a bunch of PR5 or 6 sites.
My conclusion is that Page Rank has not much to do with the actual ability of a website to pull in traffic. It is an indicator but certainly far from having a direct correlation to the search engine or traffic value of a site.
Therefore for the sake of measuring a site’s link value, Page Rank is a very poor indicator.
Alexa
The second most popular measure of a site’s link value these days. What does is really measure?
It is sold as an indicator of overall traffic but that is far from the truth. The Alexa ranking is a relative measure of the number of Alexa toolbar users that visit the site. If your visitors include a high percentage of people using the toolbar, you’ll have a disproportionally low Alexa number.
I have a test site that I put up to try out new plugins and other ideas. It isn’t indexed in any search engines and I’m the only one who ever visits that site. In two months is went from an Alexa rating of ~5,000,000 [the default], down to ~440,000 which a lot of people would consider a pretty good rating; just from my own visits using a browser with the Alexa toolbar!
There are even automated bots that can game the Alexa ranking without you even having to visit your own site.
Are you sure you want to buy links using the Alexa toolbar or Page Rank as the valuations?
In business, the president of a company should always have one key metric that accurately reflects what is going on in a company. In a blog network it could be something like, “earnings per page view” for a drugstore it could be, “earnings per customer visit.”
What should that be for the valuation of a link, “cost per ???”
Like I outlined above I think that putting Page Rank or Alexa rank in for the “???” is a big mistake.
The Alternatives
The best would be if the webmaster make his server logs available and we could derive a formula based on overall traffic, referrals and search engine traffic, but that is an unrealistic goal. It just isn’t feasible to gather and standardize that information on any large scale.
What about outbound clicks? Isn’t the number of outbound clicks a very good indicator of a sites prowess? In order to have a high number of outbound clicks a site must have:
- Lot’s of quality traffic – low quality traffic won’t click on links
- A writer that engages his readers enough to get them to take action
It is easy to measure outbound clicks, now all we need is an Alexa or PageRank style system that aggregates them and puts them into a scale so we can see where our site stacks up against others in our niche or on the web in general.
Ways to count outbound clicks: MyBlogLog & Google Analytics.
Do you agree that number of outbound clicks is the best determination of a site’s value?
Other suggestions?
Did you know that PR and Alexa were such bad indicators?
Browse Timeline
- « Pitch4Profit Contest – Fast Track to the Dragon’s Den
- » External Wikipedia Links Now Marked nofollow
Comments ( 9 )
[...] The answer is NO. You may want to scroll down to Alexa and read for yourself. Explore posts in the same categories: Search Engine and Directory stuff [...]
FrogEngine.com - Directory Blog » Blog Archive » Should I use Alexa as an indicator of how much traffic a site gets? added these pithy words on Jan 23 07 at 2:56 pmTyler added these pithy words on Jan 23 07 at 6:28 am“now all we need is an Alexa or PageRank style system that aggregates them and puts them into a scale so we can see where our site stacks up against others in our niche or on the web in general.”
Sounds like a job for a former government employed programmer??? n’est pas?
Kimber added these pithy words on Jan 23 07 at 7:23 amWell, I may not be up on tech issues but looking at my two blogs, RTF’s and NLL’s, I see that according to Alexa, NLL’s blows RTF’s out of the water. That despite RTF’s having twice the daily traffic.
Seth Godin puts a lot of faith in the number of RSS subscribers.
As a marketer, I of course prefer the Amazon model. Payment for sales.
Jon added these pithy words on Jan 23 07 at 8:29 am@Tyler
“Sounds like a job for a former government employed programmer?”
If you know any who aren’t working 14 hour days trying to build a blog network then let me know
@Kimber- I agree that sales is the only thing that really matters, but us link sellers need a publicly available way to compare sites…RSS subscribers is a good measure too, similar to my outbound clicks because it means that the readers took some action.
Mike Dammann added these pithy words on Jan 23 07 at 2:59 pmLove the info. It needs to be known by more people, especially since many webmasters help each others by visiting their sites using the alexa toolbar. People get cheated left and right.
Know.to_alex.duggan added these pithy words on Jan 25 07 at 7:21 pmSearch engine traffic still seems like such a misterious black art to me. However I’m starting to get the feeling that it’s something that can only be understood by experience and therefore will never make for a good subject for study at school. Whatever makes it into the curriculum will always be out of date since the rules are always changing.
Jon added these pithy words on Jan 25 07 at 10:48 pmHey Alex. My approach to getting traffic from search engines or SEO in general is to mostly forget about it. As the search engines get better and better, the results become more and more accurate; meaning that the best resources more and more are at the top of the results. Therefore forget all that stuff and focus on two things: create fabulous content that resonates with what people are searching for, and tell people about your great content.
Just do those two things and you’ll have no need for SEO. The search engines will find it’s in their best interest to send traffic to your site.
Fastnoc added these pithy words on Jan 29 07 at 11:01 pmI think it’s relatively poor advice to recommend google analytics.
I’ve done multiple tests of traffic numbers between Google, and a locally installed log parser.
it should be stated first that there is no better way to analyze data than an actual log parser, as using these ‘drop your code in a page’ places introduce two more points of failure and are inherently bad for statistics because of it.
Over 2 years I’ve had differences of up to 30% depending on the site. Google Analytics also becomes less and less accurate the more traffic your site gets. The only thing I can figure is they throttle your accounts. But it’s definitely true.
So for all the great looking stuff that Google did. They destroyed Urchin by taking it over and making it an online stats analyzer. Urchin use to be a log parser, and you can still get it. but it costs a fortune now.
Awstats has been far more accurate for my tests than anything out there in the online analyzer stuff. But once you get to log parsers, while Awstats is good, there are MANY nicer programs that get much more detailed reports and make it easy to drill down to get more data.
I would imagine most log parsers have the same data. But it’s the way it’s presented and how easy it is for comparison’s etc that comes down to preference.
Jon added these pithy words on Jan 29 07 at 11:23 pm@Fastnoc, don’t know if you scanned the post too quickly but I wasn’t talking about which stats package offered the most accurate stats, it is obvious that server stats are better than anything that is browser or javascript based. However it is extremely unlikely that a system could be devised that relies on server stats to accurately compare websites to get a sense of a site’s relative popularity; which is what the article is about.