|Help required understanding WMT Search Queries|
| 2:38 pm on May 29, 2014 (gmt 0)|
Having a site that saw a 30% drop in traffic from May 20th I have been looking closer at the Search Queries in WMT.
Looking at the top of the list for May 20-28 (those queries resulting in the most clicks), admittedly the figures show an overal slight drop in position. But the CTR for these queries if anything has risen.
The significant drop in number of clicks comes from a significant drop in impressions.
Does this simply mean that less people are searching?
With the queries I am talking about, I am still shown as being in the top 10 even if I have dropped a little, so it is not due to me now appearing on the second page.
My site is travel related and yes it was a holiday weekend but I have not seen these drops in traffic this time of year before so ofcourse assumed this to due to the update.
| 4:40 pm on May 29, 2014 (gmt 0)|
From what you've described it sounds like fewer people searching is the most likely cause - the other possibilities which come immediately to mind are:
a) Site showing up for a smaller proportion of searches, perhaps due to personalisation, user's IP location etc
b) A temporary ranking change which has since resolved (you mention rank dropped slightly, is it possible it dropped by a bigger amount and then jumped back up?)
Does the level of impressions recover after the holiday weekend? If so, I'd say you've nothing to worry too much about - keep an eye on the data and see what happens. Of course, if there are major site issues which you're aware of, you should probably fix those!
If there's no sign of impressions recovering, is their evidence of wider ranking changes across the SERPs you're looking at?
In that case, a wider investigation to identify what's changed might be needed (including the effects of personalisation) and then maybe look at whether there are any other external factors which might cause a large-scale change in search behaviour.
Hope this helps.
| 9:00 am on May 30, 2014 (gmt 0)|
Another theory - of course those position figures in WMT are averages, so perhaps I was straying onto the second page on occasions. Impressions would be lower if people do not go to the second page. Still doubt that fully explains it though.
| 9:59 am on May 30, 2014 (gmt 0)|
Possible. WMT does give some breakdown of those numbers showing the most common ranking positions - so although it wouldn't show a count where there's no impression, it might give some indication.
| 6:17 pm on May 30, 2014 (gmt 0)|
Another very confusing thing about queries data, even if you've selected the filter "Web" supposedly removing ranking results from Google images searches, if Google chooses to display images at the top of the search, as Google frequently does, this ranking data, typically position one in the results is included in the "web" filtered results.
You can pick these queries out because they tend to have integer "avg. positions" like 1.0 or 4.0, wherever Google chose to mix in the image results with normal SERPS.
I typically see many position 1.0 to 4.0 results that I know are due to my images, again even with the "web" filter on.
Unfortunately even though your image might be shown in position 1.0 in the results, sure doesn't mean it's going to draw a click, clearly depends on the query.
And Google seems to constantly play around with when and where to mix images into the SERPs.
| 9:28 pm on May 30, 2014 (gmt 0)|
|even if you've selected the filter "Web" supposedly removing ranking results from Google images searches |
Excluding image searches doesn't exclude image results.
It just occurred to me to wonder about the opposite, though: what about page visits that originated with image searches?