In the past, link building was less about bringing people straight to your site and more about making sure Google rank your site as high as possible.
These days, many link building methods focus on tactics that bring people straight to your site as well as helping Google understand that you should rank well for related terms. So, following on from Julie Joyce’s post on 3 Ways to Measure Link Building ROI, I’d like expand on ways that you can use Google Analytics to measure link building.
This post will assume that the link building you're doing is the type that has the potential to generate traffic directly from the links, rather than only through improving rankings. I’m not going to go in to detail about any of these techniques; there are plenty of great resources on this blog and others to help you with that.
Google Analytics is all about traffic – where it came from, what it led to on site and how much value this generated. The main focus for analyzing link building will be referral traffic data. You build a link, people use it to get to your site, you measure the value of this. But there is a lot more to it than just looking at this basic data.
This post will cover how to use Advanced Segments to your advantage, Multi-Channel Funnels for improved attribution figures, campaign tagging for easier analysis, a free custom report and how to easily use the API for exporting and combining data.
Firstly, let’s head to the Referrals data in Google Analytics.
Here we can see how many visits external websites brought to the site. Alongside this is interaction data such as pages per visit and time on site to help you understand the type of users coming from the site. By clicking the goal and ecommerce buttons above the graph you can also see how well these sites have performed with regards to conversions.
I’ve created a Custom Report to help you analyze these metrics, just sign in to Google Analytics, click this link, and choose which profile to apply the report to and enjoy the data:
All of this information can help you reach a decision as to whether the effort you went to for each link was worth it. However, you are like to wonder how much of the referral traffic can be attributed to link building, links gained naturally and social media.
You will need to decide at this stage whether you count traffic from naturally gained links and from social media as link building or whether you separate these out. You could argue the case either way, so it may be best to break it down based on whether you have separate marketing budgets for social, link building off site and creating link worthy content on site.
If there are three different budgets you will want to break them down, if it’s all under the same roof then you can analyze everything together. To break the three sections down we will use Advanced Segments.
Advanced segments enable data to be shown based on a set of instructions, be it, ‘traffic from Twitter’, ‘visits with more than 1 conversion’ or ‘visits from users in London’.
We have the following three sections to break down:
- Social media
- Natural links
- Link building
Of these, natural links is the hardest to identify and know which referring sites belong in this category. However, logic dictates that if we set up a filter to include all known social media sites, and another to include all known link building sites, we can then exclude these from the third advanced segment to leave only the natural links.
Link Building Advanced Segment
To create a segment that will only show traffic from the sites that you have targeted through link building follow these steps:
- Click Advanced Segments
- + New Custom Segment
- Name it
- Select Include
- Choose Source in the green box
- Change Containing to Matching RegExp
- Fill box with all link building domains, putting a pipe between each and a backslash infront of any dot or hyphen, like so:
The pipe means ‘or’ and the backslash means leave this as a standard character rather than matching a regular expression character.
Once this is set up and applied to the relevant Google Analytics profile, you will have filtered all data shown throughout all reports to only show data that included visits from your link building sites.
Social Media Advanced Segment
To create a social media only segment follow the steps above, but instead of filling it with link building sites we need to be using social media sites here. Like so:
Or, as I’ve already built it, you can apply this segment to your chosen profile just by clicking this link.
You may want to edit the sites in this profile, just remember to cancel out any characters with a backslash and you may need to start very short URLs with a carat (^) to prevent them matching other sites that end with the same text (see t.co for details!).
Natural Link Advanced Segment
So now that we have those two set up we can copy out the text used in each and then use these as excludes in another new segment, as shown in this image:
These three segments can all be applied together to help you compare the different results or individually to help you focus in on the performance of each method. Try applying them to different reports throughout Google Analytics for different insights.
For example, the Ecommerce report will show you the money you are generating from each segment, the location and language reports might show that you’re attracting a more diverse demographic than you expected.
If this method appeals to you and you want to take it further, later on you'll discover how to create custom channel groupings in Multi Channel Funnel reports.
If you don’t have referral data that is easy to break down you may want to consider using custom tagging on links that you build; adding information to help you identify them in Google Analytics. This would be done through the use of the URL Builder Tool which allows you to choose the campaign name, source, medium and other information for your URLs before you use them in link building or other campaigns.
The URL builder is a way to put together the campaign tracking data in the format that Google can use it. You would take the URL:
and make it relevant to the site you are promoting it on:
You might be familiar with this method from tracking other campaigns, but why shouldn’t it also be applied to link building? Obviously there is a limit to how much you can add tracking to and I’ve not investigated whether it has an impact on the quality of the link, but if it can help you evaluate the investment in resources and budget then it’s got a high chance of being useful!
Multi Channel Funnels
* Images in this section have been edited to protect actual data
In order to understand the bigger picture of your link building it can be beneficial to know how links have assisted conversions that they have not been attributed to it. Google Analytics uses last click attribution; this means that if a user has come to your site having searched for your product organically, then is encouraged to come back and make a purchase through a link you have built, the conversion is shown as referral. However if there is a third visit to the site where the user has searched for your brand name, the conversion will be organic and you won't know that it wouldn't have happened without your link building unless you use Multi Channel Funnels.
So does your link building assist conversions at all?
The best way to find out is to take your link building advanced segment from before and use the regular expression to create custom channel groupings, like so:
- Navigate to Top Conversion Report
- Above the data click Other
- Select Copy Basic Channel Grouping, rename
- Edit ‘Referral’ to be link building sites
- Add a new rule for your natural links (excluding the same data as above)
To reverse this and show the natural links you need to exclude the link building sources and ensure that remaining referral data is caught:
It may also be good to keep the social analysis consistent by editing the default social segment to include the sites you identified for your advanced segment.
So after a few short tweaks you will now have a report looking something like this:
Now, I like the path analysis for showing you the journey that users take, but if your site has many different journeys (as is common) the Assisted Conversions report will be better for you. The custom channel grouping that you just created can be applied to this report and will show you results like so:
This separated out the data for both assisted and last interaction conversions, both of which would never be understood if you relied only on the standard reports in Google Analytics.
API Export Fun
When working on a link building campaign, you will want to know both the traffic generated and how strong the promoted page of your site becomes. To do this you can export your traffic data from Google Analytics and combine it alongside link and social metrics.
If you haven't used the Google analytics API before, start by using SEO Tools for Excel, which makes it simple to export from Google Analytics as well as Majestic SEO and other tools.
Using this tool has created the following report to help analyze the full picture surrounding content marketing within a link building project:
This kind of analysis helps you bring everything together and understand not just the value of the traffic but the benefit that your link building has had in strengthening the link profile and social visibility of your pages.
As with many aspects of SEO, there is no single method that answers everything. There are different routes that you can take to analyze the value of link building and each project will require its own approach.
The ideas above are here to help you dig a bit deeper next time someone asks you what value the link building has brought, on top of the methods that Julie identified.
Title image source: jkfid/Flickr
Optimising Digital Marketing Campaigns with Search, Social and Analytics
At SES London (9-11 Feb) you'll get an overview of the latest tools, tips, and tactics in Paid, Owned, Earned, Integrated Media and Business Intelligence to streamline your marketing campaigns in 2015. Register by 31 October to take advantage of Early Bird Rates.