How can I best handle parameters?

[alovallo]

Thank you for your help in advance! I've read a ton of posts on this forum on this subject and while they've been super helpful I still don't feel entirely confident in what the right approach I should take it. Forgive my very obvious noob questions - I'm still learning!

The problem: I am launching a site (coursereport.com) which will feature a directory of schools. The directory can be filtered by a handful of fields listed below. The URL for the schools directory will be coursereport.com/schools. The directory can be filtered by a number of fields listed here:

• Focus (ex: “Data Science”)
• Cost (ex: “$<5000”)
• City (ex: “Chicago”)
• State/Province (ex: “Illinois”)
• Country (ex: “Canada”)

When a filter is applied to the directories page the CMS produces a new page with URLs like these:

• coursereport.com/schools?focus=datascience&cost=$<5000&city=chicago
• coursereport.com/schools?cost=$>5000&city=buffalo&state=newyork

My questions:

1)   Is the above parameter-based approach appropriate? I’ve seen other directory sites that take a different approach (below) that would transform my examples into more “normal” urls.

coursereport.com/schools?focus=datascience&cost=$<5000&city=chicago

VERSUS

coursereport.com/schools/focus/datascience/cost/$<5000/city/chicago (no params at all)

2)   Assuming I use either approach above isn't it likely that I will have duplicative content issues? Each filter does change on page content but there could be instance where 2 different URLs with different filters applied could produce identical content (ex: focus=datascience&city=chicago OR focus=datascience&state=illinois). Do I need to specify a canonical URL to solve for that case? I understand at a high level how rel=canonical works, but I am having a hard time wrapping my head around what versions of the filtered results ought to be specified as the preferred versions. For example, would I just take all of the /schools?focus=X  combinations and call that the canonical version within any filtered page that contained other additional parameters like cost or city?

3. Should I be changing page titles for the unique filtered URLs?

4. I read through a few google resources to try to better understand the how to best configure url params via webmaster tools. Is my best bet just to follow the advice on the article below and define the rules for each parameter there and not worry about using rel=canonical ?

https://support.google.com/webmasters/answer/1235687

An assortment of the other stuff I’ve read for reference:

http://www.wordtracker.com/academy/seo-clean-urls

http://www.practicalecommerce.com/articles/3857-SEO-When-Product-Facets-and-Filters-Fail

http://www.searchenginejournal.com/five-steps-to-seo-friendly-site-url-structure/59813/

http://googlewebmastercentral.blogspot.com/2011/07/improved-handling-of-urls-with.html

[CleverPhD]

I think you have your answer then on how you want to focus your URLs and your site!

[alovallo]

Absolutely helpful. I really appreciate it. I think one real use case that I may want to solve for is the "focus" plus "city" combo. Ie: "data science schools in chicago". Based on the research I've done thus far I think that may be the only permutation really worth worrying about. Again - thanks a lot!

[CleverPhD]

I am not going to be very helpful here.

Looking at those parameters and all the options you would have for URLs, yes you are ripe for duplicate content issue and a whole mess of search engine problems/confusion.

I read this the other day in the QNA forum here at Moz and I wish I could remember to give them credit for the quote, they said  "Don't submit search results to the search engines" - so true - so true ....

Why? You end up with an almost infinite number of thin, duplicate pages that Google then does not know which ones to rank.  Even if you put all the parameters into a static URL you still have the same problem.

I think you need to step back a sec

Are people searching for "data science schools in Chicago Illinois that cost less than $5000"?

Why would you even want to attempt to setup pages that could potentially rank for those terms based on the URL?

Launch the search function on the site, but hide all the search URLs behind robots.txt

Just setup things like

/search/?focus=datascience&cost=$<5000&city=chicago

/search/focus/datascience/cost/$<5000/city/chicago

put /search/ in robots and you are set

Another option (from one of my favorite WBF http://moz.com/blog/whiteboard-friday-using-the-hash)

Hide all the parameters behind the hash and they stay hidden from the search engines

/schools#?focus=datascience&cost=$<5000&city=chicago

Then go back, do your keyword research and build helpful static URL pages around what your users are searching for and then get those pages to rank.  If that ultimately is the type of page above, I would bet you $3,141 plus an apple pie that you need to setup a simpler organization of pages and urls around location  say /il/chicago/school-name or type /data-science/school-name  and then all the other iterations, you would hide behind a hash etc.

Maybe this did help - I hope so.