So just by being here and reading this sentence I'm going to assume you know you need to be using Google Analytics to improve your Search campaigns. Good, the first step is admitting there's a problem, and in this situation your problem is that your search accounts need some serious refinement to get them optimised. The second step is calling Persicopix to run your accounts! No, I joke, the second step is to use the tools available to you to begin optimising your accounts. Below I'm going to run over some of our go-to Google Analytics reports when we want to get some further insight into the behaviour of our users, which we can then use to optimise our search accounts.
Site Search Terms
I know we're all still mourning the loss of Organic keyword data (R.I.P), but we can still get some insight into keywords our users tend to use. Any site worth its salt will have a prominent search feature to aid users in finding the content that they're interested in. What I'm interested in is the search terms themselves, what are people looking for when they land on my site? You can grab these search terms relatively easily in Google Analytics by heading into your View settings, enabling Site search Tracking and entering your search query parameter.
If you're not sure what you search query parameter is, then head into your site and make a search. The URL of your results page should look something like the below, you just need to grab the parameter before the search term you entered.
You can find this report in Google Analytics by heading to
Behaviour -> Site Search -> Search Terms. Once you're here you'll be able to see what search terms users have been using and also useful information like the % Search Refinement (did they re-search?) and % Search Exit (did they exit from the results page?) of each search term. You can then use these search terms to bulk out your keyword lists, as inspiration for new campaigns or even as some insight into the quality of your website pages. For instance, if users are frequently searching for areas of your site that you're trying to heavily promote, it's likely they're not particularly easy to find on your site – change that.
Firstly if you're not sure what these are, head over to
my blog post here to get clued up. The Demographics reports show us some interesting information about a proportion of the users that land on our site. We can use these reports to investigate conversion rate by Age, by Gender and also by Interest group. Once we've identified a segment which is doing particularly well, why not create an ad group to specifically target this demographic on the Display Network? Let's run through an example quickly:
You can see from the table above that the age group '45-54' has a relatively high conversion rate compared to all the other age groups. Numbers from this age group may not be the highest, but at least we know that conversion rate is pretty good. So know we've got an age group we want to look into, let's create a Custom Segment to investigate the data further (again if you're not sure how to do this I cover it in
this previous blog post).
Now we've got our segment applied we can use it to gain some understanding of what this particular age group is looking for on our site. I'd head to the
Landing Page report, Site Search Terms, check which Goals are most frequently being completed and also have a look at your Ecommerce report. Each report will help you to build up a picture of how the majority of this segment behave on your site and what they're interested in seeing. In particular look out for demographic groups with high conversion rates, you can create specific targeted ad groups for these people and bid more aggressively on their searches. Armed with your demographic information head into your AdWords account, you should already have an idea of which campaign they're best suited to, based upon what the segment has been looking for on your site. In our instance, the majority of this segment has been looking at our training pages, so I'd create an ad group in our Training campaign. You can then use the data you've collected from Google Analytics to set up the ad group, such as selecting the best few Landing Pages as your destination URLs or adding a few of the most popular Search Terms into your keywords list. Once that's done, simply enable demographic targeting for the adgroup and you're on your way (info on how to do this here).
Landing page analysis
Understanding what's working and what's not working in your AdWords campaigns is a big part of optimisation. Your Landing Pages are arguably one of the most important parts of your account; sending users to a well-made Landing Page can really improve the likelihood of a conversion and therefore the effectiveness of your campaigns. The first thing to do in Google Analytics is head over to
Behaviour -> Site Content -> Landing Pages; once you're here you'll want to apply a segment to see just PPC traffic. With your segment applied, you'll now just be seeing landing pages that receive PPC click traffic.
Understanding which of your landing pages is performing poorly can be worked out relatively quickly by just looking at Bounce Rate. Click on the Bounce Rate metric to sort your Landing Pages by highest bounce rate, you then want to use a weighted sort type (this can be found next to the Secondary dimension drop down). Your report will now show the landing page with the highest bounce rate proportional to the amount of sessions they've received. Landing pages with a high bounce rate should be investigated further to understand what is causing users to leave the page without completing an action. To do this, add a secondary dimension of keyword to see what the users where searching for when they landed on the page. If the keyword was relevant to the page, we can assume something on the page isn't right and needs improving. If the keyword isn't relevant, can we create a landing page that would be more relevant or alternatively avoid these pages in your AdWords campaigns, if you are unable to affect the design of webpages.
Similar method to the above but now we're looking at Conversion rate of the landing pages. You can do this for All Goals or look at each goal specifically; this will depend on what you're actually trying to optimise. Now you've sorted by highest conversion rate, you again want to use a weighted sort type. You'll now be shown the landing pages with the highest conversion rates proportional to session numbers - these are you top landing pages, so use them frequently.
Service Provider report
The Service Provider report is, in my opinion, underused. Though Google Analytics can't show us IP addresses, it can show you the name of the Internet service provider your users belong to. Typically large enough organisations have a named Internet service provider, which we can see in Google Analytics. You'll tend to find named service providers from the following; councils, hospitals, schools, universities and large organisations. Knowing this, we can use the report to analyse whether a particular organisation or set of organisations is driving high quality traffic to our site.
Head over to
Audience -> Technology -> Network, once you're there you'll see a list of all the Service providers that brought traffic to your site. The majority will be home Internet providers (Sky, Virgin, etc) but once you've excluded those you'll be left with actual organisations. You can now identify particular organisations that are driving high quality traffic to your site, such as a high conversion rate, a high time on page or even a low bounce rate. Now you've identified a set of organisations you want to target head over to your Remarketing Lists, but make sure you've upgraded your code to support Display first. Now simply create a new remarketing list using the ISP dimension and enter your regex of organisation names. An example below:
You can then use this remarketing list in AdWords to target users from organisations who have previously delivered high quality traffic. We've even got a Google case study on this topic, in which we targeted users from banks to drive traffic to our client Watchfinder (who sell high end watches) -
The reports I've discussed above will help you gain a better understanding of what your users are doing on your site, which in turn will help you shape your search accounts to best serve their behaviour. Also I've mentioned some interesting targeting methods which should get you engaging with your different users groups in new and more personalised ways.
Give these reports a go; I guarantee you'll find some interesting stuff tucked away in your Google Analytics accounts. Until next time, adios.