Question is do your customers pay to be listed with you? If so are they using you for the dofollow links? If this is the case then you may lose some of your business by changing it to nofollow.

If they are paying there is also a risk of a Google penalty for paid do follow links.

If you are unable to maintain quality and there is no good reason to have a dofollow, then switch to nofollow.

Are the pages hardcoded? Or is all the data in a database? If it is in a database it would take no time at all to run each domain through a loop and check what response status code you get.http://en.wikipedia.org/wiki/List_of_HTTP_status_codes
This would be a very quick way to find broken links. You may even be able to purchase an api on something like majestic or Moz and run the sites through that as well for a better indication of site quality. If the site has very low DA or Trust Flow, you could also make it nofollow or remove etc...

If it is all hardcoded then that would be very hard work all around.

I took one look at the code and was amazed by how much is inline. Most of it should be in js and css files. I noticed the solutions pages many of them don't even have title tags! What's going on there? That needs seo attention and is a big no no for coding. This leads me to believe the code needs very close inspection. It surly cant take that long to strip a whole bunch of that code out of each page or template.

If a page is slow to load or pulls in external js files as well you can have scripts try to execute before the required code has loaded. This is probably what is happening some of the time and would explain the inconsistent nature of it.

Run the site through a bunch of validators. I am sure you will find a bunch of stuff.

OK I think I found out some of your issues!

There are a few major ones. I hope you don't mind but I did some detective work and found out what your site was by the code screen grabs you posted on another reply below.

OK so the first MAJOR ISSUE is that you have TWO BODY tags in your HTML! You need to check your code on all pages / templates and get it fixed ASAP. Archive.com shows that this has been a problem on your site since you did a redesign in OCT 2013. Google is looking at signals like this a lot more now and is not being favourable to them

My concern when you do a site: query in google is that your homepage does not show, it should be the first result. I see that you have 15,100 pages indexed.

When I search for your website "sitename.com.ar" I see in Google there is no title, description even the quick links are broken, just showing just one and saying untitled.

NEXT, All your pages have the same description and keyword tags! Whats going on there? I you cant make them unique then get rid of them, they are doing more damage and Google hates that stuff, I have had pages not show in SERPS because they have duplicate description tags. This happen to five of my pages recently that I somehow missed. But this is on every product Page of yours if not every page on the site.

Also I would remove the keywords meta tag, its pointless and will do you more harm than good.

The paths to your javascript and css files have two slashes (//) should only be one.

I would start with those changes and do a fetch in Google WMT and give it a few days, the homepage should reindex in a very short time frame maybe minutes.

After that we can take another look at the site. These are very basic issues that a large brand site like this should never be doing wrong, as you can see the damage it has caused. Lucky for you that you have only just started there and can be the knight in shining armour to sort it all out

First off if you want to keep the sitemap file, consider breaking it down to multiple files, one for each language/country etc...

Also FYI there are THREE methods, you can also add hreflang to the http header. This might be a good options as well to consider.

https://support.google.com/webmasters/answer/189077?hl=en

if all pages are 100#5 not indexed then I would block it in robots.txt, Google's John Muller confirmed to me that Googlebot will continue to crawl every link to check to see if a nofollow or noindex has changed status.

So as a result we blocked our pages with robots.txt and saw a great increases in index/crawl rates on pages we want Google to pay attention to. It also reduces waste in server resources.

However if there are any pages that are index, if you block them in robots.txt then Googlebot will never be able to crawl the link to determine that it should be noindex. This means it could stay in a permanent stage of indexed.

I hope that answers all your questions?

Panda has run recently, does the the drop coincide with that date? Normally there is a shake up in the week leading up to it.

I suspect that someone may have made changes before you go there and they have not given you the full picture.

Was there any issue with the server during that time and the pages speeds were slow or not loading. WMT should give you some indication of the decline, even how dramatic the drops were and what kind of pages were dropped.

FYI you are better to remove the meta descriptions from the pages that have duplication than keep them there, but the best course of action is of course to re-write them.

If you care to share the URL we can all take a look and see if we can spot anything.