Search engines are the most important source of traffic for most SE sites. Understanding how users find the site and which questions they land on can be very useful in order to promote the site or improve the experience for new users.
My subjective impression is that on many sites is that a small percentage of the questions are responsible for a very large part of the total traffic. Being able to identify those questions would e.g. allow us to polish them up a bit and thereby improve the first impression of a lot of new visitors with a minimum of effort.
Moderators already have access to a tool that lists the most popular search terms for a specific time frame. But there are several problems with that tool that make it difficult to impossible to gain useful information from it:
- There are results from the wrong SE site in that list, and there seems to be no obvious way to fix that.
- Google defaults to SSL-encryption for logged in users now which means that the referer containing the search terms is not transmitted anymore. This leads to a large percentage of
(not provided)
entries in that list. - Search terms are not aggregated in any way, so you get usually multiple very similar entries for one popular question. This makes it very hard to compare how much traffic from search engines specific questions actually get.
- You cannot easily identify which question the users found by a specific search term (except for performing the search using the same keywords yourself).
Another possible way to identify questions that drive traffic to the site is the number of views. But this method has several limitations, it doesn't distinguish between referred traffic (e.g. from Hackernews or Reddit) and it can't be restricted to specific timeframes.
What I think would be a more useful tool would be a list of questions sorted by the amount of search engine traffic they received in the selected timeframe. This would allow us to easily identify what questions drive the most traffic to the site and use that information to guide the promotion of the site and to polish those questions a bit.
The information about which search terms are used is certainly useful, and it should stay available (ideally connected to which question was found by a specific term), but the information which questions are found by search engines is in the end the more useful in my opinion.