SEO Experiments

You can link to this page. Please don't link to any individual experiment page. Thank you.

Test: Engines must choose one page to rank (due to the duplicate content penalty). At the very least it must choose one to rank higher.
Control: An exact copy of every experiment is created with the link order reversed to act as a control.
Subject Criteria: Each unique "word" is chosen based upon it's ability to return exactly 0 results in an engine before the start of the experiment.
Results: Results are dated, described, and will be updated with new results as the algos of engines change over time.
Observe: View the source of each individual page to see the experiment. Search for the "word" to view the current results.
Known Issue #1: Since (most) experiments have only a single word for the entirety of its content, it may get special treatment i.e. more content is preferred over the quality of the content in cases of pages with barely any content.
Known Issue #2: The nature and factors of this domain name and website (siteownernews.com) may have an effect on results.

Active Experiments

Inconclusive Experiments

Testing bold vs. not bold

Erythroalchem | Erythroalchem
Tirythroalcham | Tirythroalcham
04/26/11 Result: Inconclusive - Engine chose bold in the first experiment, non-bold in the control.
06/08/11 Update: Inconclusive - Engine chose bold in the first experiment, non-bold in the control.
04/26/11 Note: When the engine chose the bold version, it appeared in bold in the SERP.

Testing code bloat vs. no bloat

Yefikoeloruskess | Yefikoeloruskess
Tafikoeloruskess | Tafikoeloruskess
05/04/11 Result: Inconclusive - Engine chose bloated code in one case and non bloated in another.
05/26/11 Update: Conclusive - Engine chose bloated code in both cases (see known issue #1).
06/08/11 Update: Conclusive - Engine chose bloated code in both cases (see known issue #1).

Conclusive Experiments

Testing standard link vs. lots of parameters

Hefikoelorintiuyo | Hefikoelorintiuyo
Yerikoelorintiuto | Yerikoelorintiuto
05/26/11 Result: Conclusive - Engine chose link with no parameters in both cases.

Testing Javascript vs. no Javascript

Shroolchanieson | Shroolchanieson
Ikoeloruskech | Ikoeloruskech
05/26/11 Result: Conclusive - Engine chose plain text over text rendered by JavaScript.

Testing image alt text (plus file name) vs. standard text

Tayajamit | Tayajamit
Ruguejivatul | Ruguejivatul
05/26/11 Result: Conclusive - Engine chose plain text over image in both cases.

Testing W3 Valid vs. not valid markup

Dukojamitres | Dukojamitres
Bokojamitres | Bokojamitres
05/26/11 Result: Conclusive - Engine chose valid markup in both cases.

Testing in a sentence vs. not in a sentence

Iaxrocedence | Iaxrocedence
Iaxrokedance | Iaxrokedance
04/26/11 Result: Conclusive - Engine chose in a sentence for both experiments. (See known issue #1.)

Testing matching anchor text vs. mismatched anchor text

Wiralangulos | Apples
Tulips | Wedulupguluses
04/26/11 Result: Conclusive - Engine chose pages where anchor text matches page copy.

Testing repeating keyword x4 vs 1 keyword instance

Yeirotoitackazes | Yeirotoitackazes
Deirataitockozos | Deirataitockozos
04/26/11 Result: Conclusive - Engine chose pages where keyword is repeated 4 times.

Testing bold tag vs. strong tag

Waguezivatil | Waguezivatil
Yukaguezivat | Yukaguezivat
06/08/11 Result: Conclusive - Engine chose bold over strong in each case.

Testing outbound link vs. no outbound link

Incyithroalcha | Incyithroalcha
Enithroalchanon | Enithroalchanon
06/08/11 Result: Conclusive - Engine chose the page with an outbound link in each case.

Testing meta description vs. page text

Koelorintanab | Koelorintanab
Tapitockoros | Tapitockoros
06/08/11 Result: Conclusive - Engine chose page text over meta description in each case.

Testing h1 vs p

Eitackazeeirotos | Eitackazeeirotos
Nairailackozos | Nairailackozos
05/04/11 Result: Conclusive - Engine chose h1 tag in each case.