In the first three parts of this series I used the example of Qype to show how you can analyze the loss of Google-traffic. I tried to show and comment my progression from isolating the afflicted parts of a website and finding out what the problem is, to listing possible causes for said problem. This fourth part of the series will deal with the implications of the former steps. I want to note that I was not in contact with Qype at no time during this evaluation or since – these insights are completely derived from publicly available data-sources. It is possible to get a better understanding of the reasons for changes when you are in direct communication with a website-operator, which will then also make the recommendations more fitting.

To keep it short: I think that duplicate-content caused the ranking-penalty of the /place/-directory. As far as I am concerned, I believe that this was also done automatically through the Google-algorithm and had nothing to do with Google intervening manually. Why do I think so?

Lets take another look at the VisibilityIndex for the domain. We can split this story into three phases: the first phase (sensibly labeled with “1”) shows a decrease in visibility over a few weeks, though it is rather constant and not just a break. It is interesting to take a look at the ranking-distribution for this timeframe (as can be seen here) – the number of results on the first and second results-page are becoming fewer, but this is happening as part of a normal progression. At this time, there is not yet a penalty on the directory, but Google is just throwing Qype pages out of the index. In the second phase, the whole progress stabilizes: the amount of returned keywords stays mostly constant and there is no change in the VisibilityIndex. The only fluctuations in the ranking-distribution are in the decimal places. Then the third phase comes around and things get interesting: the VisibilityIndex takes a leap upwards – and this is not because the available keywords are ranking better once again, but because, once again, more pages are put in the index, which also adds additional keywords (+20.000). And that seems to be what triggered the penalty that hit one week later.

To get your head around this, you need to get inside Google's mindset: even though it would be great, no site is without its problems. This especially true for naturally grown pages, where you will often have some kind of problem show up. Google found a way to deal with this, though. For this, Google balances the problems with the trust they have into the domain. Problems are things like duplicate content and so on, while the trust is measured in links. If the balance tips towards the problem side, Google will instate a penalty and the site goes to the back of the searchengines result-pages. This can be seen nicely for wissen.spiegel.de, a wikipedia-clone: sometimes the domains massive linkpower prevails, at other times Google decided that the duplicate content problem is just too vast.

And I think that this is exactly what also happened to Qype: it seems that the amount and quality of the incoming links was not enough anymore to balance out the onpage-problems. Massive internal duplicate content in different facets is to blame for this and having these badly done purchased links has probably not helped either. As far as the other two possible problems are concerned (rankings for sale and conflict of interest with Google), at the moment I do not think they had anything to do with this, seeing how I can not make out the necessary manual changes in the Google-SERPs that would show up for something like this.

What should Qype change?
As short as this answer is, it will be complicated and time-consuming to implement the needed changes: Qype needs to get a grip on their pages. Many of the seemingly naturall URL- and navigation-structures that grew over time need to be unified and improved. The product-team has to work together with the (existing?) SEO-team to decide what should show up in the URL and all of this needs to be incorporated correctly and monitored by useful processes. Once these basic structures are in place, Qype will be able to get rid of this penalty – which was implemented by the algorithm – as quickly as they caught it. So if they put in a night-shift in Hamburg and make the right decisions, they might be able to have a very different picture as far as SEO-traffic is concerned at the end of May. I wish you good luck.

There is something rotten in the state of Qype
There is something rotten in the state of Qype - part II
There is something rotten in the state of Qype - part III
There is something rotten in the state of Qype - part IV