NB: This is a guest post by Maria Vasbotten, a search engine advertising intern at EasyToBook.
Almost a month has gone since Google Instant was launched and the first results of its impact are starting to coming in.
To our surprise many SEM/SEO managers have stated that they have not seen any impact of it so far. Here at EasytoBook we are left with a question mark as we do not see a huge impact, but we do see differences.
So we launched an investigation to see what had happened.
Some background
Google Instant is a new search enhancement by Google, which is supposed to reduce the time spent on searching by giving you search results as you type - some more correct then others, but nevertheless it’s supposed to be able to reduce your search time by two to five seconds per search.
Below you can see an example of how Google Instant works. If you search for "london hotels", you only have to start typing your keyword and the results will start showing underneath while you type. The interesting thing is that it will show the same search results for "london ho" as for the keyword "london hotels".
This means that people do not have to type the whole keyword to get the information they want. The only difference is which ads are shown. For a PPC team this means that there are changes to which keywords you need to buy and how much you bid.
Experiment methodology
Firstly, all the information that was needed for the investigation was collected. The information used, was retrieved from Google Analytics, Google Adwords, Google Insight and our own internal systems.
To be able to get a good insight, data from two weeks before Instant was launched was compared with two weeks after the launch.
After collecting, all the information was analyzed. Our accounts structure allows us to analyze our accounts easily by looking at Impressions, Clicks, CPC, CTR and Quality Score.
Findings
After analyzing, we could see that most of the accounts did not change too much, except a slight increase of Impressions and CPC, while CTR and QS suffered a slight decrease.
However when we looked at the impressions on a city level, we could see that there were large differences in the accounts/campaigns related to France/Paris.
As shown on the graph below, Paris had a very noticeable increase on impressions around the time Google Instant got launched and kept a very high curve further on, compared to some of the other cities.
Impressions per city:
As this didn’t give us enough answers, we started looking into where the traffic came from in each account/campaign in relation to France/Paris.
At first we thought it would be the traffic from the USA, which would be the problem as this is one of our larger markets and it is where Google Instant was first launched.
When we looked into it, to our surprise, it was actually the traffic from France itself that had caused the differences.
In this account it clearly shows that the impressions have increased, while CPC increased only slightly and the CTR dropped dramatically and therefore our Quality Score decreased drastically, which you can see in the following graphs.
Average Quality Score VS CTR Paris:
Impressions VS CPC Paris:
Average Quality Score VS CPC Paris:
When looking at the graphs, you can also see that something happened around the time Google Instant was launched, and a week onwards. We therefore started examining our keywords to see what might have happened with them in that timeframe.
This is where we found out that some of our “broad” keywords related to France/Paris had received enormous amounts of impressions these days without us changing anything in the related ad-groups.
The keywords were: "Hotel Paris" and "Paris Hotels" (Broad). "Hotel Paris" received a large amount of impressions over several days, while "Paris Hotels" only received a large amount of impressions during one day.
Keywords VS Impressions:
This is where we are left with a question mark - and one in which there are no clear answers to:
- 1) Why was it only these two keywords that were really affected?
- 2) Why on those particular dates? [NB: it is these two keywords that caused the most difference in our accounts]
So, Google Insight was checked to see if it might have only been us that had the raise of impressions. But as you can see below, it does seem that in general the keyword "Hotel Paris" had an increase of hits in September, compared to the months before.
And its getting even more hits in October.
Google Insight for Kw Hotel Paris:
To be able to assure that it was indeed Google Instant that created the differences it was important for us to eliminate possible causes.
Therefore all numbers and figures that were found were consulted with our SEO team as well as our partner/affiliate’s team. Neither of these teams saw any large changes in their accounts the last month, except SEO who saw a slight decrease in traffic.
We also looked into seasonal changes, traveller behavior, sudden changes in browser usage, but none of these gave any specific answers.
To make sure that it was not us that had triggered the differences, we went thoroughly through our Adwords Change History to ensure that we had not overseen any mistakes. However, we could not find any direct leads.
Conclusion
We can see that we had a difference in our accounts after Google Instant was launched compared to before, and that this difference was mostly related to our France/Paris accounts/campaigns.
It is hard for us to pinpoint if this happened because of Google Instant or not, therefore we are curious to hear if others had any similar impact on SEM/SEO?
NB: This is a guest post by Maria Vasbotten, a search engine advertising intern at EasyToBook.