Welcome to the Q&A Forum

The fact a Caps create a 404 error on LAMP site is a pet peeve of mine - so is the fact Google thinks mix cases on IIS are separate (thus duplicate) URLs.

Too arbitrary to be picky about and cause user frustration.

Thanks goodness at lease DoMaInS can be what ever.

This is a strong indicator something is up and deserves deeper investigation.

Perhaps you have content duplication issues, low value content (Panda), spammy back links (Penguin) or other indexing issue.  See if there is a pattern to the missing pages, perhaps one of the directories is the cause.  How old is the site and how is the domain trust/authority coming along?

Another tactic one client used was a website credits page.  It had full priority content dedicated to the authors/designers and developers of the site.  This link could be treated as higher priority over small footer text.  Bonus, the page would likely be more topical to your business.  Not for everyone however.

If the descriptions are very technical then likely there is a fair amount of repetition in the sentence pattern, diction etc.  I'd recommend playing with regex to help transform content into something original.

For instance, you could search for industry abbreviations CW and replace with long forms _**Clockwise (CW).  **_Maybe they over use an adjective that you could changeto your own voice.

Also, perhaps the stock descriptions have blocks of useless content you could strip out in the mean time?

The DB probably has a few other fields (name, product attributes etc) so be sure to find a unique way of assembling the meta description, title and details.

If you find enough to change, I'd think having the description would be better then having a page that is too light on words.

Be sure to mark up with http://schema.org/Product so SE's understand the nature of the content.

EDIT:  I have used the regex technique to enhance the content of a database by added inline tooltips, diagrams or figures and glossary links.  However with Penguin, I would be careful with automated links.  You would only want to create a handful using the same anchor text.

EDIT2: I forgot - MAKE FREQUENT BACK UPS.  Regex is super powerful and can tank a database really fast.  Make a backup of the original and of every successful iteration - it will take a little longer but it will save your butt when things go bad.

I have had direct experience with this issue.  Here are a few of the things I have done:

1. Make sure the current URLs are rock solid and can be long lived.
2. Ensure all links to the old structure are completely purged from the content.  No good to propping up the old patterns.
3. Get a clear picture of the off-site back links.  No sense worrying about pages that will have no value. If they changed that fast, there won't be many to worry about.
4. For those that have good back links, make a direct 301 redirect to the new page.
5. At the point of low ROI, redirect the rest on pattern matches.  There could be a couple double jumps here but they won't mean that much anyway given #3. (side note, double jumps leak extra link equity so they should be avoided)
6. Ensure your entire site can be fully and easily spidered - before resorting to xml sitemaps.
7. Ensure you have a helpful even compelling 404 page that returns the proper status code.  410 errors are reportedly a bit faster at getting content removed so use if you can.
8. Remove any restrictions you have in the robots.txt file to the old structure until the 404 and 301s take full effect.
9. Submit remove requests to pages and folder. This is particularly important if the site is very large (compared to domain authority) and SEs won't get a full picture of the changes for weeks or months due to low crawl allowance.

Doing these got my sites back on track within a couple weeks.

EDIT: forgot a couple...

• remove any old xml sitemaps
• submit multiple sitemaps for different sections of the site.  This helps narrow down problem spots easier.

If you can use server side programming to change the homepage, this would be better for your users and optimization.  Just be sure to expose enough content on each for Google to spider and index properly.

Yes, 301 redirect the old night home page. In fact, redirect all homepage requests to just http://www.domain.com/

The canonical link can be removed if on the domain root.

My suggestion is to try AJAXing the tabs.  If the outbound links are more of a concern then the keywords of the link, AJAX loading of the tab content would remove them from consideration.  Google won't index content pulled in from an external source.

However, be careful to put a rel="nofollow" on the link that loads the content as you don't want SEs indexing the source.

Do not put a meta nofollow in the head, it will kill all links on the page and seriously mess up your link flow.  Your use of rel="nofollow" is correct in the context of the specific link tags.

I wouldn't sweat the shear number of links - the 100 count is a left over from the days when spiders only downloaded 100k from the page.  It has since risen to the point that the practical limitations of over 100 links is more pressing (IE, do you visitors actually value and use that many links?)

If each link is valuable and usable, no need to worry.  If not, perhaps there is a structural way to reduce the count.

Also, load the footer by AJAX  onscroll or on demand.  Assuming all of the pages can be found in the top navigation, the bottom links are just exacerbating  your issues.  Primarily, this section is giving far too much weight to secondary or auxiliary pages.

For instance, your Privacy Policy only needs to be linked to where privacy is a concern (ie the contact form).  Good to put it on the home or about pages too if you have a cookie policy.

How do the back link profiles compare?  One really good link for keyword 1 could be enough to propel it to the top.  Alternatively, perhaps it gets better treatment by your internal link structure.

Another thought: competition is usually a measure of how many pages are returned for a search but it doesn't get into how well optimized those pages are for a given term.  It also doesn't measure the authority and trust of the domains.  It could be the competition for the first keyword is low value domains and unoptimized pages.

If you don't want these indexed, first put a noindex tag on all pages.  Leave the follow alone as the engine still needs to find the pages to change the index status.

Add the domain to GWMT then request a removal all the pages.

Allow this to take effect then add a robots disallow to the entire sub-domain.

Your domain then be cleaned from the index and the duplication won't be an issue.

Before you delete anything, submit a ticket to the help desk.  They may be able to fix it up or point to why it's happening.

http://www.seomoz.org/help

"Domain authority 82"

There's your answer right there - that's a fantastic link.  One to cultivate and track.  Hopefully you can be a resource to the domain again sometime.  Perhaps the topic can be updated year by year etc.

1. Try logging out or using a completely different computer - Google modifies your results according to your past search behavior.
2. Double check you are comparing the same country version of Google.  You may be searching Google.co.uk and tracking Google.com
3. It could be getting a boost from local search - try the search with a friend in a different part of the country.

Page two correctly uses:

rel="prev" href="http://www.buffetdomicilio.com/category/artigos" />

Which indicates to Google it is part of a pagination scheme thus it handles it differently to proper content.

You really don't want Google sending traffic to these pages but instead to the actual articles.   In that case, optimizing the title or duplication reduction isn't such a big a deal. It's just one small part on otherwise different pages so you won't get a bad penalty.

I wouldn't use the canonical link suggestion as this will effective remove the page from your index.  I assume you want these for spidering so anything that removes them could be an issue.

SEOmoz has a number of public tools they can use. After that, a basic membership allows for up to 5 campaigns so you and your friends could share a membership and get all the benefits of Pro.

A word of caution,  don't interlink with your friends sites unless they are of direct benefit to your customers.  Even then,  use sparingly. The absolutely last thing you want to do is anything that feels like cheating. Link exchanges/purchases are frowned upon.

What if they just want to use a pen name to protect their privacy?  Hard to remember with modern life being so exposed and annotated but there are lots of people that prefer to remain anonymous.

So long as the persona is well developed, I wouldn't have a problem with it given the content was high quality.

The Mozbar Page Attributes indicates:

| Internal Followed Links | 276 |
| Internal Nofollow Links | 60 |
| External Followed Links | 14 |
| External Nofollow Links | 14 |

But Link Data says 0 internal links. Curious.

PS, cute robots.txt file.

The phasing made me chuckle...

Holding the obvious impact of normal conversations on marketing aside, one would guess higher positive social interactions would lead to a higher profile.

It wouldn't be a smart business strategy of these sites to not allow people to find hot spots but just continue to order by an arbitrary string of letters. (when you think about it, alphabetical order is really strange UI )

Given scheme.org has clear mark up for ratings,  I would further assume this would impact the SERPs, particularly in that location. This requires that the SE can see the data and that the site has sufficient trust/authority )

Testing this would be tricky however since you are talking about brick  and mortar places. It would be pretty elaborate and potential cruel to try making a fake restaurant out rank a real one. I would expect checkins from a single account would have limited effect or people would be gaming their listings.