SEO

I’m currently working on some SEO tests into onpage optimisation. The following are my working thoughts on what I’m testing, and details of my results. The tests are being carried out for a university research project.

The rankings of each test page will be compared against the other test pages across the major search engines Google, Yahoo and MSN (UK versions).I will be testing the importance of having a keyword (I’ve chosen a unique one to make it easier to see the results) in titles, page heading tags, image alt tags, etc. I’ll also be examining how the various search engines deal with blackhat SEO methods and spam type pages - will these pages be filtered out completely, their rankings decreased, or will they perform best of all?
Rules:

• Each test will have its own page. Each page will be a standard HTML format page, rather than a wordpress blog page. (This is because wordpress pages use php and are automatically generated in parts - my pages will be much simpler to control).

• Tests will all use the same unique keyword I am making up for the sake of these experiments (no other pages currently use this keyword in the search engines - please don’t use it or my tests will be harder to monitor!). Every page also has its test name in the page title tag. The keyword and test name will identify each experiment page and all of them, with luck, will show up on their own Search Engine Results Page (SERP) in Google, MSN and Yahoo. I can then compare rankings.

• Each page shall contain 2 paragraphs of textual content as default. Keyword will be inserted within these according to test specs. I will be altering the wording in paragraphs for each page to prevent any duplicate content filtering or penalities.

• Monitor each page at regular intervals to see how rankings change in each search engine results page.

• No external linking to test pages. All test pages will be linked to from the home page itself and nowhere else. (Please don’t link to my test pages if you are reading this - to the homepage is fine though). This means SE results won’t be affected by offpage factors like different numbers and Page Ranks of incoming links, all pages will be equal.

• All links to test sites have now been made nofollow to prevent spam penalties for the entire site/test pages. Update (11/11/07): links are now followed, obviously the search engine couldn’t pick these up, and they couldn’t be accessed from homepage.

Page Title Tests:

• SEOtest1a: no keyword in title, page content of 2 paragraphs with KW once in each. [Basic Control page]

• SEOTest1b: keyword 1 time in title, page content of 2 paragraphs with KW once in each. [Title only]

• SEOTest1c: keyword 2 times in title, page content of 2 paragraphs with KW once in each. [KW Dense Title]

• SEOTest1d: keyword 3 times in title, page content of 2 paragraphs with KW once in each. Prediction: excessive use of keyword in title, should be seen as a spam page and filtered? [Excessive KW Title]

Heading Tests:

• SEOTest2a: Keyword once in H1, rest as test 1b.[Title & H1]
• SEOTest2b: Keyword once in H2, (not in H1), rest as test 1b. [Title & H2]
• SEOTest2c: Keyword once in H1 and H2, rest as test 1b. [Title, H1 & H2]
• SEOTest2d: Keyword once in H1, not title, rest as test 1a. [H1 only]
• SEOTest2e: Keyword once in H2, not H1 or title, rest as test 1a. [H2 only]
• SEOTest2f: Keyword once in H3, not H1, H2 or title, rest as test 1a. [H3 only - low visibility in SERP?]
• SEOTest2g: Keyword once in H1, H2, H3 not title, rest as test 1a. [H1,H2,H3]

URL Tests

• SEOTest3a: Keyword once in url, rest as 1a. [URL only]
• SEOTest3b: Keyword twice in url, no gap between. Rest as 1a. [URL repeated KW]
• SEOTest3c: Keyword twice in url, with dashes between. Rest as 1a. [URL repeated KW, variation]
• SEOTest3d: Keyword once in url, once in title tag. Rest as 1a. [Title & URL]

Font Weight Tests (Bolded KW’s)

• SEOTest4a:
• SEOTest4b: