It's no surprise that the paid search reports in Google Analytics encourage and help you to slice and dice your data using multiple dimensions. But for unpaid search you get the keyword and the search engine and… well, that's it. There are no built-in equivalents of 'campaigns' or 'ad groups' to help you structure your projects and establish business processes to get the best results.
But custom advanced segments can fill the gap. You can divide your keywords into groups to get an easy overview of how you're doing and find the key to some powerful analysis.
What's more, you can then configure Google Analytics Intelligence alerts to let you know when things change. Do that and you can get yet more advanced insight into the major contributors to the change as a bonus.
The combination of custom segments and Intelligence alerts configured for those custom segments is extremely powerful in the way it provides both a top-level view and detailed suggestions about the causes of change.
That thought may be enough to set your mind thinking of different possibilities, without need of further advice. But if you'd like some detailed instructions as starting point, please read on.
Use GA Custom Segments to Create 'Rolled-up' Keyword Groups
At the moment you're probably reporting on this kind of work using external systems. I'll assume that you have a document of some kind which contains your keyword research and that it lists a series of phrases, typically grouped in clusters associated with different master phrases.
If you configure each of these groups as an advanced segment you can then use the segments to report on 'rolled-up' groups of organic keywords and also to create custom 'Intelligence' alerts which will give you access to Intelligence's 'Major Contributors' analysis of what's causing the change in performance.
All being well, when your traffic for these phrases goes up, GA will send you an email to say so, and when you log in it will provide valuable clues about what caused the change. Was it some of your own work? Or did you get lucky?
Here's How to Set Up Custom Segments for Keyword Groups
Click on Advanced Segments in the left navigation of Google Analytics, then click on the 'Default Segments' area to open up the list of the built-in Segments. You can use these as the starting point for building a custom segment.
So in this case, click the 'Copy' link to the right of 'Non-paid Search Traffic':
This will open up the configuration screen for the segment so that you can make changes and then save it as a new segment.
The first thing to do is to give the segment a new name. GA does not insert any form of automatic name and so you'll get an annoying error message when you try to test your segment unless you do this. I always used to get bitten by this, so I now make it a habit to do it first!
I name such segments on the basis of the master keyword phrase for the phrases in the associated group.
The first step is click the 'Add "and" statement' button, because what you want is a segment where all the visits are from organic search "and" the keywords used for the search were part of your target group.
Next, open up the green 'Dimensions' menu on the left hand side of the screen and drag the 'Keyword' dimension over in to the report area and drop it on the 'dimension or metric' box.
At this point there are two ways to add your keyword list.
If your prefer a 'drag and drop' approach (I don't) type your first keyword into the box to the right of the keyword dimension. If you want to be very precise with your matching, stick with the 'Matches exactly' default option. Or you may choose to change the drop down menu to 'Contains' instead.
Then you need to add your keyword variations. To do this you need to 'Add an "or" statement'. Be careful about this. You need to add the 'Or' to the list of keywords, not to the 'Medium'. Be sure to use the 'Add an "or" statement' link beneath the keywords.
The grey border area around these input areas shows how they are grouped together, but it's not very prominent so many people miss this visual clue.
You can then repeat this adding of an 'or' statement for each of the keywords in your list.
Personally, I find the drag and drop approach can result in long and unwieldy screens. I prefer to use a regular expression match so that all the relevant keywords are included in one match.
To do this, you only need to add one instance of the keyword match statement but then click on the drop down menu to the right of the 'Keyword' dimension and change it from the default 'Matches exactly' to 'Matches regular expression'.
The regular expression you need consists of your list of keywords separated by the 'pipe' or 'vertical bar' symbol, which is usually shown on your keyboard as a vertical line, or vertical split line in one of the shifted positions.
There's no need to spell out all the keyword variations in full. The list above would also be matched by
keyword|keeword or even key|kee (if you're not too worried about matching more than you anticipated), but the long version should do no harm. This is not the place to discuss the use of regular expressions in Google Analytics when there are other great resources on the subject, notably Robbin Steif's free Ebook on "Regular Expressions for Google Analytics"
This may seem a more complicated technique, but it can be more practical and more powerful in the long run. It's up to you!
In the example below, I've combined the two approaches in a way which adds clarity for longer-term management of such a project.
The keywords in this case relate to an imaginary project based around two cult skateboard clothing brands from the 1980s. So in this example the keywords have been divided into two sub-groups, one for each brand.
- Each of the groups has been configured as a regular expression match
- The second group was added by 'Adding an "or" statement' to first one
It's a good idea to use the 'Test' button on the Custom Segment configuration screen to check whether your matches are working properly and producing sensible numbers. (Tip: that 'test' function can be a great way of doing 'on-the-fly' reporting on the numbers of visits which matched a complex set of conditions. It can be very handy for one-off funnel reporting.)
You're all set. You can now get a nice rolled-up view of your group of keywords just like you could if they were an Adwords Campaign or Ad Group. But the extra benefit of this approach is that you can now use the Segment to configure an Intelligence Custom Alert
Bonus: the Extra Power of Custom Alerts
To set up a Custom Alert for your new segment, click on Intelligence in the left navigation, then choose 'Create a new alert'.
Start by giving your alert a name, then choose the period for the alert. In this example, I've chosen monthly, but you may find that a week works better for you. For SEO work, daily is going to be a bit too short, I think.
Don't forget to select the option to have GA send you an email when the alert is triggered.
Next, open up the 'This applies to' green menu and choose your new segment from within the 'Custom Segments' list.
Now move down to the blue 'Alert me when' menu below and open up 'Site Usage' and choose 'Visits'.
Finally, drop down the condition menu and choose the '% increases by more than' option and specify a percentage increase. In the example I've chosen 10%, so I should receive an email if the visits from this segment goes up by more than 10% compared with the month before. You will want to use a figure which makes sense to you according to your site's performance. Then hit the 'Create alert' button, to save your changes, and you're done.
You can use your custom segments in day to day reporting straight away. But the benefits of setting up an Intelligence Custom Alert will take time to appear.
The first obvious benefit is the email to tell you that something has changed and things are moving in the right direction.
But there's more to it than that. The Intelligence Alerts system also reports on the factors which may have contributed to the change. When you click through from your email alert you should be able to open up an extra reporting area beneath the alert to see the 'Major Contributors' to the change.
At the time of writing this post I don't have an example of this on test site which I can show in public. So no screen grab here, I'm afraid. To see what it looks like, please read the Official Google Analytics Blog announcement: Major New Features Added to Intelligence.
You'll want to experiment with different settings for that alert, which may take time. But in the meantime, taking a structured and segmented view of the normal reports should help you get to grips with your day to day work.
I help small and medium sized retailers make their ecommerce sites better for their customers and better for their business by analyzing what people do on web sites and what they say about them.
I've been fascinated by how people use sites ever since I first started digging into log files back in 1995. I've used a lot of clickstream tools since then, but these days I use Google Analytics. I never cease to learn new ways of using it.
If you've read this far you'll know that I don't do "short". If you'd like to read even more like this, please check out my blog on ecommerce optimization.