One of the goodies in our PubCon bags was the November 2006 Website Services magazine. The magazine got buried, and it was mid-December before I had a chance to page through it. I found one article, Website Design, Search Engines and Organic Keywords, that gave me pause and made me wonder about the fact-checking and editorial overview of the rest of the magazine, and the advice being given to those new to the SEO field. Before writing this, I emailed the author, Paul Bruemmer, in December to understand the methods used in his article. I have not received a reply other than the auto-responder saying he would be back in the office on January 2.
I received a phone call from Paul Bruemmer this morning (January 15th) to clarify some points in the article. He (rightfully) points out this is a print publication, and the material he submitted was considerably more detailed than what ended up in print, due to the pruning and rearranging that happens in the editing and publishing phases. The issues covered in the article can be complex, and hard to capture in a short piece, especially in a publication that is not dedicated to search engine optimization. The industry is still new, and is less than transparent in its functioning (understandably so in some cases, as the search engines do not want publish an exact blueprint of how they operate). The article was not intended to be an in-depth analysis to be handed to a company, but an article to highlight that even some of the top-name companies have room for improvement. By doing so, the article (and my conversation with Bruemmer) highlighted some of the issues and variables facing the search engine optimization industry, including inbound link value and calculation (especially deep links versus homepage links), Page Rank and canonical issues (www.nytimes.com has a PR of 10, nytimes.com has a PR of 8), and issues of how Page Rank follows 301 redirects.
The Wall Street Journal and The New York Times were compared, with Bruemmer pointing out the numerous ways he felt the WSJ had bad search engine optimization techniques. When I did my own queries in Google, I found considerably different information. My query results are from late December and early January, Bruemmer’s appear to be from early October (judging from the screenshot in the article). Bruemmer’s original premise that the New York Times does a better job at SEO than the Wall Street Journal, with more pages indexed and more incoming links, is still valid, but there is a bit less of a gap between the two than is initially shown. Alexa’s traffic rankings of the two sites also show the NYT considerably ahead of the WSJ.
Pages indexed comparison
Bruemmer includes a screenshot of a query used to show how few of the Wall Street Journal’s pages are in Google, but what it really shows (if you do your own queries) is canonical issues with search engines (searches showing pages indexed in Google)
- 394 pages indexed site:www.wsj.com (Bruemmer)
- 181,000 pages indexed site:wsj.com
The next query shown is the failure of the landing page to be indexed in Google. When you go to www.wsj.com from a US-based IP, you are redirected to online.wsj.com/public/us. Bruemmer uses a query on this exact page to state that Google has very few of these pages indexed. The problem is that none of the articles on this page use the public/us in their URL (go to the “Free Today” section and view an article).
- 29 pages indexed site:online.wsj.com/public/us (Bruemmer)
- 395,000 pages indexed site:online.wsj.com
Next, the Wall Street Journal is compared to the New York Times, yet the comparison is possibly unfair and is difficult to replicate, given the lack of screenshots or explanation of the query used.
- 2,960,000 pages indexed (Bruemmer)
- 244,000 pages indexed site:www.nytimes.com
- 4,840,000 pages indexed site:nytimes.com
My questions for Bruemmer:
- Why was the www not omitted from the site: query for the Wall Street Journal?
- Why wasn’t the query for the New York Times shown, so we could see how it was calculated?
Bruemmer answered that due to publishing space, not all of the screenshots and query information he submitted with the article were included in the print version, nor were canonical issues included.
Inbound Links
I am concerned that those new to Search Engine Optimization will follow Bruemmer’s lead and use Google to calculate inbound links. Google is fine for many things, but is well known to not list all of your inbound links (Rand from SEOmoz states that Google themselves admit the numbers are bad). Why not use Yahoo Site Explorer or even Microsoft’s Live search?
New York Times inbound links to home page:
- Google: 196,000 Exact link: query unknown. (Bruemmer)
- Google: 590,000 link:www.nytimes.com
- Google: 12,000 link:nytimes.com
- Yahoo: 1,780,451 for www.nytimes.com excluding inlinks from this domain
- Yahoo: 257,253 for nytimes.com excluding inlinks from this domain
- Live: 1,155,272 link:www.nytimes.com -site:www.nytimes.com
- Live: 1,155,272 link:nytimes.com -site:nytimes.com
Wall Street Journal inbound links to home page:
- Google: 23,000 Exact link: query unknown (Bruemmer)
- Google: 66,000 link:www.wsj.com
- Google: 66,000 link:wsj.com
- Yahoo: 2,788,196 for www.wsj.com excluding inlinks from this domain
- Yahoo: 20,649 for wsj.com excluding inlinks from this domain
- Live: 272,587 link:www.wsj.com -site:www.wsj.com
- Live: 270,695 link:wsj.com -site:wsj.com
If you believe Yahoo Site Explorer, the Wall Street Journal has twice as many inbound links as the New York Times. I know that no queries are perfect, but I take exception to using a known bad query, and not giving the exact method so that the methods can be duplicated. This, after all, is an article designed to help people use these techniques. Bruemmer points out the lack of space both for the multitude of queries and for an explanation of the pros and cons of each query. The Google link queries are able to give a rough order of magnitude indication of incoming links.
Editorial Inconsistencies
There are two places in the magazine that caught my attention and made me wondering if an editor was napping for the final check of this publication.
In the Bruemmer article, the redirect from www.wsj.com to online.wsj.com/public/us is called a 301 redirect twice, and a 302 redirect three times, with no explanation given as to what the redirect means. The two redirects are different, Matt Cutts gives a good explanation about the difference between a 301 and a 302 redirect.
Later in the magazine, there is an article titled Is Your Website a Virtual Money Pit? 10 Ways to Keep Website Costs Low. The second way to save money is to “Choose a shared Web host”, as it is “typically much less expensive than a dedicated host”. Issues regarding security and space regarding other sites on the same host are briefly mentioned. Back to Brummer’s article about SEO, he devotes a section to the server and the choice between a dedicated server and a virtual hosted. He writes “A dedicate server is the better choice, because it allows for much more flexibility and provides the ability to keep your IP address clean — this means a clear history devoid of spamming. It’s also important to have a dedicated IP address, rather than a shared IP address.”
Summary
The magazine has good intentions, but this article was disappointing with its internal contradictions, lack of using the same methods for comparing information, and contradictions between articles. Unfortunately, when I see information that I know is wrong, I view with some suspicion the information in the other parts of the publication. I appreciate Brummer’s prompt response when my post was brought to his attention (in the middle of a busy time in his life), and his concern about my post and the industry of search engine optimization as a whole.
Comments on this entry are closed.