For years, enterprising webmasters have attempted to build a reliable third-party gauge for measuring the strength of a specific web page. The rationale is simple. Currently, toolbar PR serves as the default gauge for the value of an inbound link from a specific web page. Sophisticated webmasters have devised more effective methods, but for the most part, the valuation of an inbound links (reciprocal, paid, or otherwise) has been PageRank.
Clearly, there should be a more effective tool for evaluating the value or “weight” of a specific web page, and coming up with that tool would render Google’s PR useless, once and for all.
Unfortunately, building an alternative to Google’s little green bar has proven to be an elusive undertaking. Currently, the best alternative that I’ve come across is the Page Strength offered up by seomoz.org. Sadly, I feel that this tool falls short of
Rand’s original vision, which he shared with me years ago, and I haven’t seen any better alternatives out there.
Until now!
As many of you already know, Michael Jensen from soloseo.com recently released his IndexRank tool, which analyzes the frequency with which Google indexes a site’s new content. While far from perfect, the tool does provide a fairly systematic way of analyzing the aforementioned “weight” of a specific web page by utilizing one of the most definitive pieces of Google data available. Furthermore, Jensen’s tool relies solely on Google index data and avoids some of the potentially useless information that the seomoz.org “Page Strength” tool relies upon (example: “Page Strength” takes factors such as Alexa ranking into account).
Again, I do see some fairly significant shortcomings that
plague Jensen’s “IndexRank.” For example, it does not seem to have any
reporting/exporting features, which would allow a webmaster to easily transfer
the data to a separate document for further analysis.
Also, the tool is limited to domain queries, so you cannot
gain insight into the weight of specific pages outside of the homepage.
Perhaps the most glaring shortcoming stems from the
IndexRank algorithm itself. Jensen’s algorithm being used seems to favor sites
that have a high ratio of newly indexed content compared to total indexed
content. In laymen’s terms, established sites with an extremely high number of
total indexed pages will get a lower score then sites with less total indexed
pages even if they get more content indexed on a regular basis.
Here’s an example I pulled directly from the tool:
|
|
|
|
Pages
Indexed
|
|
|
|
|
Domain
|
2 Weeks
|
1 Month
|
3 Months
|
6 Months
|
1 Year
|
Total
|
IndexRank
|
|
www.nfl.com
|
859
|
1430
|
9780
|
9790
|
9750
|
50500
|
6
|
|
www.realfootball365.com
|
130
|
246
|
417
|
581
|
1390
|
18500
|
9
|
|
|
|
|
|
|
|
|
|
The fact that the second site, an independent football
portal that has is getting much less content indexed on a regular basis, is
benefiting from the fact that it has way fewer total pages indexed.
Those were the first two shortcomings that popped into my
mind, but I figured that it might be a good idea to ask the guy that turned me
on to this tool – Aaron
Wall from SEOBook.com - if he spotted any other potential bugs.
Here are Aaron’s thoughts on the matter. “I can see Google
going out of their way to obfuscate this data, especially for site:
searches...such that they make it hard to get any value out of such a specific
tool.
There are lots of ways to bolt on data and add features, but
the key is figuring out how to do so without straining your server beyond
its limit or getting so well known that Google ends up blocking the site or
obfuscating the data.”
Aaron definitely had some interesting insights on potential
pitfalls, but it wouldn’t be fair to point out IndexRank’s faults without giving
the creator a chance to respond. Here is an excerpt of my conversation with
Michael Jenson:
Q: Do you plan on adding any reporting and/or exporting features?
A: We plan on integrating IndexRank with our other
tools in SoloSEO.com, such as in our SEO statistics tracking for client
domains and their competitors. We are still working on improving the
algorithm, and depending on the interest from the community we would
take it to the next step with exporting, widgets, etc.
Q: It seems as if IndexRank favors sites that have
a small number of total indexed pages relative to the amount of pages
being indexed within the last two weeks, one month, etc. Are there any
plans to further refine the algorithm, so that more established sites
with an extremely large number of indexed pages don’t suffer?
A: IndexRank doesn't favor sites with a low amount
of indexed pages, but rather favors sites with an increasing amount of
content relative to the amount indexed in previous time periods. We
realize this penalizes other sites without the same momentum or growth
trend as they may have had in the past (such as larger sites), but the
metric is intended to show growth not size. We realize IndexRank isn't
perfect, and we are working on refining and optimizing for improved
accuracy. From the larger sample sets we have looked at, IndexRank
appears to be an accurate indicator of site growth.
Q: Aaron Wall mentioned that Google could
obfuscate the data you are currently using to power the IndexRank tool.
Are you concerned about that possibility?
A: Google is known to obfuscate other data, such
as their backlink reporting through the link: operator, so it wouldn't
surprise us if they did. However, with this type of data being a part
of their overall search features, versus the link: operator being used
primarily by search marketers, it doesn't seem like it would be in
Google's interest to obfuscate their indexing data. Even if they do,
our advantage is that IndexRank doesn't depend on reporting exact
numbers from Google, rather it uses the relative rate of content being
indexed.
P.S. Thank you to both Michael and Aaron for
making themselves available on such short notice to help me pull
together this piece!
