While working on a number of classifieds websites I have encountered an interesting phenomenon, which I would like to share with you. I have called it "landing page lifecycle", and in this case it concerns ad details pages on classifieds websites. However, I think that it applies to any landing page of temporary character (like news, articles, discussions).

Traffic patterns for date-sensitive landing pages

During the advanced analysis of the traffic sources and landing pages for one of the classifieds websites (Morusek.pl) I have noticed that the traffic once gathered around a single ad details page eventually decreases in time. It is even more visible, if you gather the traffic for a number of pages that were created in a short period of time. Below, you'll find some examples:

Non-paid search traffic through pages (ads) created between June the 13th and June the 21st , 2011:

Non-paid search traffic through pages (ads) created between May the 29th and June the 6st , 2011:

Non-paid search traffic through pages (ads) created between May the 13th and May the 21st, 2011:

As you can see, the behavior is very similar for different date intervals. In all the above-mentioned examples the traffic first sharply increases, reaches its maximum, and then gradually decreases to 15%-20% of the initial weekly numbers.

I have also combined the data for 10 different sets of dates to obtain more objective general trends:

And again, there's almost instantaneous traffic increase and just after a week it starts decreasing. Firstly very fast, then it slows down. After one month there's only 30%-40% traffic left. After two months it is only 20%. After 3-4 months the traffic decreases to ca. 10%-15% of the initial values.

Behavior analysis

Of course there is no mystery behind such a rapid traffic increase at the beginning. It takes Google up to 24 hours until an ad gets indexed and just after that time it can be ranked and gain the traffic. But why is the traffic decreasing?

If we analyse the traffic page by page and phrase by phrase, we can find the three following types of behavior for every time-sensitive landing page.

The survivors

As we can see from the above graphs, the traffic decreases down only to reach some value. Usually it is 10-15% within 4-6 months. It means that there must be some landing pages that somehow resist the decrease. Those are the pages that are either linked from external sources and manage to gain good link popularity, or those that are optimized for keywords of very low competition. Below there is an example of the traffic for 5 landing pages (above) and 5 corresponding groups of key phrases (below).

External competition

Some landing pages are not so fortunate. They are optimized for phrases that are quite competitive and it may appear that just after several days there are similar ads created on other websites. As a result old pages loose rankings because the other are more up-to-date. In this case both the traffic for selected landing pages and the number of visits for related keywords decrease to 0:

Internal keyword cannibalization

There is also a third type of landing pages. These ads are optimized for phrases that are competitive also in case of other pages within the website. If Google finds a new page that is optimized for the same phrase, it selects the more up-to-date page for SERPs. As a result the traffic for the old landing pages decreases, however the traffic for related phrases remains on the same level:

Conclusions

The above-mentioned data is a good example of how Google promotes fresh pages on SERPs and how that influences the turnover of landing pages and the traffic. However, it also shows how "fragile" the time-sensitive landing pages are. They can drive the traffic for 2-3 weeks, and then they are gone, substituted by fresh ones.