Skip navigation

Testing Search Engine Page Ranking Techniques

I’ve talked a lot about how Google ranks websites with their patented Page Rank program, analyzing the process from a variety of sources, with some hearsay evidence as well as facts. There is still a lot that we don’t know about how Google ranks web pages and how search engines read your web pages, but we are learning more every day. The more we learn, the more we can help Google rank our pages by providing them with clean, valid content and code.

A recent research project into analyzing how Google ranks web pages came from Mike Industries’ The Roundabout SEO Test, and I’m really excited about his findings.

Mike wanted to test how reliable the myths of web page design for search engine optimization were by setting up some specific pages with unique content on it used in a variety of ways, and then see how these published posts do in search engines, specifically Google.

After witnessing the strange results from the Roundabout SEO Test, I set out to discover exactly what effect, if any, HTML code has on Google search results. The first thing I did was make up a word.

That word is “lodefizzle”… a nickname given to the legendary Stephen Lodefink of Finkbuilt.com during our days at the Disney Internet Group.

I created 15 files in a new directory on my site which contained the term “lodefizzle? in different contexts. Some had the word in the title element, some in the H1 element, some in the filename, etc etc etc. The idea was to see how treatment of this word affected search result ranking within Google.

While not a scientific research project, his findings were fascinating and should be evaluated by every web page designer. It isn’t new news, but it reinforces the truth rather than the myth. You can’t say “do it just because you are supposed to” any more. We now have a good idea of why and how it works. Here is a summary of his findings.

Headings Matter: While we are told to put all headings in our posts in heading tags (h1, h2, h3, h4, etc.), Mike found that keywords used within the h1 tag “does indeed assert some dominance” in Google search engine results over h2, h3, and other lessor headings. So use good keywords in your main heading tags. And use heading tags.

But also use them right. In basic Web Design 101, we are taught to structure our headings in an architecture so that h1 comes first, followed by h2, then h3, and so on in consecutive order. You can dance up and down going from h2 to h3 to h2, but you can’t go from h4 to h2 without having come down consecutively from the 1, 2, 3, 4 and so on.

Mike ran a test that checked search results ranking with the order of the headings tags and found that they do matter. Keywords in h1 headings tags rank higher than keywords found in lower headings tags even if they come before the h1 tag in the order of the page. Mike concludes that, “The semantic effect of hierarchical HTML headings is a stronger factor in search rankings than more rudimentary measures such as physical keyword placement.” Clean up your headings order, folks. And put your most important keywords in the highest ranking headings tags.

Tables and Nested Tables Penalizes Search Result Ranking: For years you’ve been told that tables are for data not design, yet do people listen? Do WYSIWYG web page design programs listen? No. Mike found out that indeed a nested table does impact your page rank, but he also found that good use of strong keywords can overcome the “negative effect”, but more research would be needed to determine how much it would take. On a personal note, within a month of leaving table design behind on my website, my page rank soared all over the place. I only changed the design not the content.

Invalid Code Penalizes Search Result Ranking: Mike created a variety of borked table tags and found that yes, invalid codes impacts search result ranking, and in one case, it resulted in the page never being found by Google. He came up with the following conclusions:

1. Somehow grading pages based on how they are rendered as well as how they are coded.

2. Simply counting the rest of the page as an attribute of the invalid table because the attribute is never officially closed off with an end quote.

There may also be other explanations to why this is happening, but this was the most interesting test in the bunch for me.

Conclusion: It’s not clear that validity helps search engine ranking, but it’s definitely true that certain errors in your code can get you completely removed from indexes.

Lesson learned? Make sure you have valid code on your web pages. All of them.

Proper Use of Keywords in Titles, Headings, Filenames, and Incoming Links Matter: Mike then decided to test how keywords placed in the titles, headings, filenames, and links influenced page rank. He found that “Although good semantics are somewhat valuable in optimization, simple things like proper titles, descriptive filenames, and incoming links are dramatically more important.” Use good keywords in all the right places, and use them repeatedly (and judiciously) throughout your web page’s content, using ALT and TITLE tags in images and links, combined with valid and semantic web page architecture, and Google will rank your web page higher than if you don’t do these things.

It all boils down to a good, strong, and valid web page design that meets web standards as well as accessibility standards, and a consistent and strong use of keywords in the right places.

Mike describes in detail how he did all these tests, which you can do yourself to see if your findings match his. And thanks, Mike, for helping us understand even more about how Google search result ranks work. You’ve solved many mysteries that plague students faced with “just because”. Now, they have an idea of why.

Here are more articles that may help you with search engine optimization and understanding search engine page ranking:


Site Search Tags: , , , , , , , , , , , , , , , , , , , , , ,
Copyright Lorelle VanFossen

12 Comments

  1. Posted February 8, 2006 at 8:12 pm | Permalink

    Thanks Lorelle, but I think you’re overstating some of my findings. Specifically:

    1. Tables and nested tables really have very little effect at all on search engine placement. It’s kind of like pulling out one strand of your hair and saying you have less hair now. Yes, you do have less hair, but not in a noticeable way.

    2. Invalid code generally is not punished at all in the world of search engine placement. It is possible, however, to munge your code up *so* badly that not only does it fail to render in a browser but it also fails to get indexed in search engines.

    Thanks… otherwise, good wrap up.

  2. Posted February 8, 2006 at 8:17 pm | Permalink

    Thanks, Mike.

    In the tests I did, as mentioned, page rank went through the roof as soon as the table design was removed, so from my personal, also unscientific tests, table layouts had an impact. And this was long before search engines crawlers and Page Rank are as sophisticated as they are now. Back in the “old days” of web page design. ;-)

    And thanks again for the great research project. It will seriously help so many understand better how this all works, and what will and won’t work. You are a star!

  3. Posted February 8, 2006 at 11:45 pm | Permalink

    Valid code on your blog is important from a user perspective. So many, (too damn many) WordPress themes are full of broken code that they bork IE and for most blogs, we cannot afford to ignore IE users.

    Other than the default Kubrick theme I have only found two WordPress themes that validate.

    WordPress Theme designers – do not use outdated paragrapgh and li classes and you would probably have a theme that would validate.

  4. Posted February 9, 2006 at 8:53 am | Permalink

    remotecontrolceo:

    If you look closely through the CSS of the Kubrick/Default Theme, you will find a ton of hacks to force that layout to work. Many of us were heart broken when so many WordPress Theme designers based their WordPress Theme designs on that Theme. It’s a great Theme, but it has many, many problems, hacks, and odd ball features.

    WordPress Support volunteers and the WordPress Codex do all we can to promote validation of Themes before releasing them to the public. There currently is no “filtration” process to “certify” WordPress Themes before they are listed publicly. It’s self monitoring. I was hoping that at least the Themes that would be used with WordPress.com would have higher standards, but they are still working on that.

    People have to remember that much of what is on the web right now is self monitored. If it needs to be fixed, people need to be told and they have to fix it. They aren’t going to get punished in any way, unless they falter in search engine page ranks. So we have to keep promoting web standards, seo standards, and validation.

  5. Posted February 9, 2006 at 10:43 am | Permalink

    Lorelle: Hmm, well are you saying “PageRank went through the roof” or “My search engine placement went through the roof”? Both are entirely different things and, as detailed in my tests, have very little (if any) effect on each other. PageRank is merely an indicator of the magnitude of people who link to you (quantity combined with quality). “Search engine placement” is how high you show up in searches for particular things. As shown in my tests, a PageRank 5 person can show up higher than a PageRank 8 person, even though the 8 is supposed about 1000 times greater in terms of PR.

    I think what you are probably talking about is “search engine placement” and it likely wasn’t the removal of tables which helped you… it was the addition of semantic code. Semantic code + tables would have probably had the same effect. Don’t get me wrong… I don’t like tables. I just don’t think Google cares about them.

  6. Posted February 9, 2006 at 2:48 pm | Permalink

    You’re right, Mike. My “search engine placement” went through the roof, though that isn’t a very scientific finding. ;-) Excellent point.

  7. Posted September 15, 2006 at 4:32 am | Permalink

    I have recently been looking into seo and I was hoping you could tell me the best way to test my sites ranking and position. It would be easy if we turned up in even the top 100 but at the moment we don’t. I am very keen on improving our rankings and will certainly be trying the techniques listed on this page, but I would like a dependable method of testing my results. I don’t see the point of starting to add or change things if the results of these changes cannot be reliably tested.

  8. Posted September 15, 2006 at 5:14 am | Permalink

    I should add that I am only really interested in results from the UK as we are a UK based charity and cannot help anyone outside of Scotland

  9. Posted September 15, 2006 at 7:40 am | Permalink

    Besides the unreliable but “good enough” test from above, I recommend you read Do-It-Yourself Search Engine Optimization Guide and use the tools provided in that article to help test your site.

    Honestly, if you are serious about SEO in the UK, then hire an expert in UK search engines, SEO, and web development. Don’t do it yourself because that is highly specialized. If you want it done right, then pay for someone whose mortgage relies on getting it right the first time. Otherwise, write keyword rich content and let web nature take care of itself.

  10. Posted September 15, 2006 at 10:14 am | Permalink

    Thanks Lorelle I will try and do it myself because I would like the experience but also our work is not reliant on the net We use it more to keep in contact with current clients rather than to attract more, Thanks for your help.

  11. Posted September 15, 2006 at 10:30 am | Permalink

    Then you are on the right track. Focus on what your clients need on your website, and pack it with the keywords that service that need, that people will search for, and you will do fine. After all, it isn’t UK page rank you need to monitor, it’s your customer’s needs. Ask them.

    Check out The Question: What Do You Love and Hate About a Website?. A UK news show asked this question and got amazing responses. That will help you better than monitoring UK search engines.

    Good luck and let us know how it turns out and what you learn.

  12. Scott Million
    Posted May 13, 2009 at 5:24 am | Permalink

    Thanks for the heads up, I’m optimizing my search engine marketing campaign and I also want to try these few more steps you have mentioned as to have tested it myself.


5 Trackbacks/Pingbacks

  1. [...] I recently wrote about some simple techniques to help your search engine results ranking, and these techniques can easily be done by WordPress.com users, so let’s start there. [...]

  2. [...] Testing Search Engine Page Ranking Techniques [...]

  3. […] Testing Search Engine Page Ranking Techniques […]

  4. […] Testing Search Engine Page Ranking Techniques […]

  5. […] Testing Search Engine Page Ranking Techniques […]

Post a Comment

Follow

Get every new post delivered to your Inbox.

Join 20,278 other followers

%d bloggers like this: