Can some keywords get penality? - all situation

[bele]

Last 3 years we created backlinks with 3 main anchors for our website.
Domain name example is www.jackusedcars.com , keywords: bmw , audi , mercedes. We have chosen some big keywords as our main keywords.

And some more small search volume words: buy used cars, used cars sell off. 90% of backlinks are with main keywords.

OSE:
 BMW          2,505      162,638
 audi            1,111      209,542
 mercedes  735         64,649
 used cars  382         28,368
 car sale      136         8,517
 toyota          108        13,106
 buy used car 34       820
 car sell off     28        710
 usedcars.com 26    45
 sold cars       23        472

Website title example is: BMW Shop, buy Audi and Mercedes used cars

90% of backlinks are to index page. (Now we have Linking Root Domains 5,158; Total Links: 512k) all backlinks are related, we never used any auto spam tool etc.
in 2011, November 16-30th "audi" keyword traffic dropped, around -80%. Other keywords were ok.
We haven't been kicked by penguin in august 24. Graph was the same. Since November, our index page traffic dropped by 70%...

1st question: We got penalized for overoptimizing with "AUDI" keyword? Or its just another reason it stopped driving traffic?

I know that for such linkbuilding we could get kicked by penguin on next update. So now we are de-optimizing the website, changing our old backlinks to different anchors and different urls (car pages - with car name anchors). We are quite good in the search results with product pages now. 1st page always - serp depends on the competition. For example "Used BMW 530 car for sale".

We are creating new backlinks like this:
10% to index with different anchors - not the old big ones
20% to http://italian.jackusedcars.com with italian anchors
20% to http://www.jackusedcars.com/search/bmw530 with "buy bmw530", "bmw530 sale" and so on
50% to product names http://www.jackusedcars.com/BMW-530-i-x-2009-full-options.html with product name anchor

We still want to get our traffic back with popular keywords. We have them written in title and keyword density on-page is 0.99% (previously was 1.37%).
Every month we lose around 10% of traffic to index page. We were with these keywords in top3, and now only 1 keyword is somewhere in top10, others are not even in top50.

2nd question: Should we remove "BMW, AUDI, Mercedes" from title? (they are still driving us around 20% traffic + 20% with bmw sub-keywords) We could lose almost 50% of total traffic. Only sub-pages will drive traffic with non-popular keywords.

We have plans of making page www.jackusedcars.com/bmw and optimize it with "BMW" keyword. Could it go through?
Some old backlinks would be changed to this page.
Our best conversion is with these main keywords, so we really need to get them back.

All comments are welcome. Graph attached.

visitors_yearly.jpg

[Zoolander]

I don't work in SEO and I'm not an expert so I grant you don't need to take me as one in what I say.  I want to participate here, hopefully won't embarras myself!  Don't know how much you'd want to derail from your already established words but I notice you have no keywords of  "used (insert variant of car)" which would be more specific hence closer to purchase intention?

...such as "used BMW" which for google USA gets a local exact match of 9,900 and global exact of 27,000 or "used audi" 4,400 LEM and 18,100 GEM.  "BMW" is 550,000 LEM and 4,090,000 GEM but 84% competative and not as specific to what you do whereas "used BMW" is 54% competative, more specific and closer to purchase intention.

"used cars" and their makes such as "mercedes" are two terms that don't seem to be brought together.  Might you be capturing market of people not looking specifically for a used mercedes?

[bele]

Dear Arpeggio.

thank you for a comment

BMW AUDI Mercedes, URL and Title were just as example. Real keywords are different.

[kchandler]

I am actually going through this exact issue with a client right now as well. My hypothesis is that Google is not determining/penalizing us based on the keyword that is used in the search query but where the large amount of exact match anchor text is pointing.

Essentially i believe the penalty is at the page level. So a few weeks ago i began running the following test that is something kind of like this:

1. Create a new page on your site targeted for one of your keywords that are dropping, for example "used BMW" or "used audi"
2. Make sure to get your on-page optimizations just right with the term target tool.
3. Then begin building new links to your new page with a large variation of anchor text from sites with a Moztrust over 2.

If my hypothesis is true then your new page will begin to rank as it's authority rises. This is at least what i am currently testing (only a few weeks into it). Hope this helps out!

Cheers - Kyle Chandler

[MarieHaynes]

I think a main key to improving here is to figure out why you dropped in the first place.  You're right, it's not Penguin if your drops started in November of 2011.   There was a Panda update on November 18th so you may want to look in this direction first.  Evaluate the pages that have dropped for duplicate content and thin content (i.e. only a few unique sentences on the page.)

Google did give out some unnatural links warnings at that time (although the majority of them started in Jan/Feb 2012).  I'm sure you've checked your WMT for unnatural links messages, right?

Another thought is to look at where you were building links.  It's possible that the sites that were linking to you got penalized or deindexed.  Lots of blog networks got devalued (mind you this was in Februrary 2012) and many directories got deindexed in May of 2012 or just prior to that.

The other possibilities are if something major happened in November such as accidental noindexing, a robots.txt problem, malware, a change in url structure or other things.

[bele]

Thanks for the answer.

If the penality is on the page level, i assume it should had a drop on BMW keyword also. But as you see the decrease is stable. Also with other smaller keywords we are in the 1st place. Also in some countries our index is in 1st place with the popular keyword.

[bele]

Thanks for the answer, but no, nothing mentioned happened.