Usually Google and Bing do a good job of figuring out when terms are similar unless they have completely separate meanings (see Rand's response on singular vs plural). In your case if you go to google and type in "switchplate" it actually shows the "Did you mean switch plate?".

Also if you do a search on "Switch Plate" you notice Amazon comes up first with the word separated.

So those two notes are some big clues as to what direction to go for that example.

Either way I would use PPC as a playground here and run exact match ad campaigns for both versions of the term, find out which gets more traffic for you, which gets better conversions, and then optimize for that one first. Go with the data.

Anecdotaly I have heard singular is better for e-commerce as plural is usually used in language when talking about an item informationally. So I would probably go for singular and throw in a couple of plural versions on the page like an h2 or h3 link for "Other switch plates".

This could also be handled by doing your high level category pages as plural, and specific product pages as singular. I would probably keep a term like "switch plate" consistently separated in my copy though if I saw this was the big winner in my PPC test, but that is the copywriter/editor in me begging for consistency and not the SEO tweaker going for every version. In the end I still want my site to be user friendly and clean.

Underscores are not good to use in URLs as search engines do not treat it as a space or delimiter. So Snow_Bee would be considered Snowbee. I believe search engines have gotten better at figuring out when a word like this should be two words, but the best practice for this is to use hyphens (e.g. Snow-Bee).

Here is a video from Matt Cutts about it:

http://www.youtube.com/watch?feature=player_embedded&v=AQcSFsQyct8

This wouldn't cause you to drop off though especially if you ranked previously. Is it just for this one term or is it for multiple terms? I would be looking at link backgrounds to be sure there weren't any spammy links getting devalued that maybe didn't get caught before, or perhaps review the content of the page to see if it might have gotten caught by the penguin anti-spamminess updates.

Hands down I use GA much more than AW Stats.

AW Stats has it's place, but the segmentation of users, user defined variables, events, goal tracking, and e-commerce tracking all tied up with adwords data easily trumps AW Stats.

AW Stats is great for getting info on bots, 404 errors, bandwidth usage, and hits, but these are used more for diagnosis and less for business level decisions.

Also, the variations you are talking about between the two rarely matter unless you think the variation is due to large portions of your traffic not being tracked correctly by one or the other.

The reason why I say these variations don't matter is that both methods of tracking have their quirks, and what you are really using them for is to see the trending of the data.

As long as you know the data is as accurate as you can get it, and you know how/why you are getting the data you are getting, then that data becomes actionable based on the trends you see and not based on the explicit value of the number.

As an example lets say I see in GA 1,000 visitors this month, and 1,200 visitors next month, and I see in AW Stats 10,300 visitors and 12,500 visitors. I care less about the differences in the numbers in the systems and more about the 20% growth in that statistic. Considering I can find out the "why" easier and with more clarity in GA I usually don't even look at AW Stats for this type of data.

I also might take a look at AW stats when first setting up analytics to be sure it looks like key indicators are showing similar trends so I know I am not missing a huge chunk of the site or something similar.

Check out articles on the Panda update recently to see a lot about this topic.

I especially like this one from Vanessa Fox

The gist of it is that Google has started to devalue sites with lots of "cloned" pages where the content is very similar from page to page, but the area info, or keyword swaps out.

This is based on the idea that content that is tailored to a page's topic is more useful to the user, and it is more likely to carry nuances than broader content.

So for example I can talk about hotels in Dallas, or New York the same exact way "See Dallas/New York Hotels downtown and be close to all of the hip restaurants and activities"

Or I could go with something more Dallas Specific "See our Downtown Dallas Hotels near the American Airlines center, the Dallas Museum of Art, and direct access to the DART Trains"

The second example is WAY more useful to the user, and so Google's latest updates will tend to favor that over generic topical text.

So if you can afford to get copywriting for each topic I would.

If you can't then I would start picking out the biggest value terms and build content specific to that, use generic content for the other terms for now, but slowly replace those over time as well with more valuable content.

You may be in a situation where there isn't more valuable specific content to outrank you right now so the generic content will do decently in the rankings. In this case you may not see such a horrible ranking, but when someone eventually competes with you and catches on to your scheme it will be better to have the unique content already working for you.

Just my 2 cents

I would get those re-written to dashes for a few different reasons:

1. Dashes are known as the best way to separate terms in URLS
2. IMO dashes are easier to read by the user, and easier to remember thus more user friendly (just my opinion though no data behind this)
3. Plus signs are often used in URL encoding or in query strings, both of which are not great for users, and both of which have been thought of in the past to look somewhat "off putting" to search engines compared to dashes.

So while it might be a pain I would say go to dashes.

Yes this hurts SEO. Mostly in that search engines rank pages based on the URL as a the URL is the only unique identifier they have for a given page.

This means link popularity is split amongst the URLs and the same page is competing with its self to rank so Google usually picks one and ignores the others, and then you lose all of the link juice for links to the others.

So with that said there are multiple ways to fix this. The most useful one is to setup 301 redirects for every option and redirect the user to the one you want to use. This way even if an outside source links to one of the variations google will go through that link, hit the redirect, and see the final URL and use that for ranking. Going the 301 route also keeps your stats programs much cleaner as you won't see stats for various versions of the pages.

The hard part with 301s is catching all variations of a URL.

With that said you might have a 301 in place, but then have parameters passed to the URL often. This is where the rel=canonical tag comes in handy.

rel=canonical is a tag you can place in a page that tells anyone who cares (search bots mostly) that if you see this page at a URL other than the canonical URL you need to treat it as the same page as the canonical URL (that is a simplification, but works for the most part).

So, get 301's in place, and then in your template for each page get a rel=canonical tag setup with the base URL so you are taken care of on both fronts...

This can be rough on huge sites, but there is usually a programmatic way to do this to catch a good portion of your pages.

If it is an HTML 5 page that should be just fine as it is the HTML 5 doctype tag.

Check out the info here: http://www.w3schools.com/html5/tag_doctype.asp

It might cause trouble with really old browsers, but if they have that going I would hope they have already thought about the reasons for and against using it

EDIT: even better link: http://www.w3.org/TR/html5-diff/#doctype

Here is what I posted last time this came up:

I would get those re-written to dashes for a few different reasons:

1. Dashes are known as the best way to separate terms in URLS
2. IMO dashes are easier to read by the user, and easier to remember thus more user friendly (just my opinion though no data behind this)
3. Plus signs are often used in URL encoding or in query strings, both of which are not great for users, and both of which have been thought of in the past to look somewhat "off putting" to search engines compared to dashes.

So while it might be a pain I would say go to dashes.

Also here is link someone posted last time about an article Rands wrote in 2006:

http://www.seomoz.org/blog/11-bes...

I don't have a lot of experience with this, but here is my 2 cents for what it is worth.

Obviously if you are running a multi-lingual site you would want to use the correct characters for the corresponding language on that version of the site.

If you are however running an english site, and writing an article on why it is important to learn to say "Donde esta el baño" then I suggest doing some quick research.

Going to google I can see that the suggested searches it shows as I type in the query don't use the "ñ" character. If you look at the results for the search without the "ñ" character the first set of results do  have that character in them. This suggests to me that Google will recognize either as the same thing for this specific word.

I recommend doing research like this for your specific situation as you might have a conflicting word (e.g. the word changes meaning depending on the use of the character). In which case hopefully the engine can pick up on the context of its use.

Also try multiple engines and match up your strategy to what engines your market tends to use.

Sorry I couldn't be more specific as I haven't had this problem.

This is how I would go about testing, and planning for a solution though