Context

Our website has a large number of crawl issues stemming from duplicate page content (source: Moz).

According to an SEO firm which recently audited our website, some amount of these crawl issues are due to URL parameter usage. They have recommended that we "make sure every page has a Rel Canonical tag that points to the non-parameter version of that URL…parameters should never appear in Canonical tags."

Here's an example URL where we have parameters in our canonical tag...

http://www.chasing-fireflies.com/costumes-dress-up/womens-costumes/

rel="canonical" href="http://www.chasing-fireflies.com/costumes-dress-up/womens-costumes/?pageSize=0&pageSizeBottom=0" />

Our website runs on IBM WebSphere v 7.

Questions

1. Why it is important that the rel canonical tag points to a non-parameter URL?
2. What is the extent of the negative impact from having rel canonicals pointing to URLs including parameters?
3. Any advice for correcting this?

Thanks for any help!

Very cool, John. Thank you for running the test.

What I'm seeing in the results is that all of the title case title tags are projected to outperform the lowercase title tag I selected as the baseline. Granted, your software warns "Not enough responses & care should be taken when using the predictions", the outcome supports my hypothesis that title case would perform better than all lower case.

From here, I intend to run some before & after tests on our title tags to see if I can measure differences in performance.

Thanks again!