How To Check if Affiliate Programs are Worth It with Universal Analytics

Following on from our article on are affiliate programs worth it, we now dive into the methodology of researching this concept using Universal Analytics (old style Google Analytics).

Open up Google Analytics

The first step is to open up your Google Analytics portal which is found at https://analytics.google.com (you will need access from your web team if you don’t have it already). This is how it will look when you first log in:

Find the Model Comparison Tool

Using the navigation on the left hand side of the page, you want to select “Conversions” which will open out three further menu options. You’ll be choosing the bottom one called Multi- Channel Funnels which then has another 8 options.

Lastly, select “Model Comparison Tool” as this fits with what you’re trying to do which is compare one way of thinking (last click rules) with another (introducer click rules). Upon successful navigation you will find yourself looking at a view similar to this:

Use the correct Channel Grouping

The first change to make on this page is to choose the “Source” (assuming that you are using source to identify your affiliate network’s traffic. We encourage the use of source rather than “default channel grouping” as Google isn’t always 100% accurate at recognising affiliate traffic:

You should also do a quick verification that all conversions are being included (see our screenshot above) as sometimes Google can default to selecting sales where only two channels or more are involved (especially when there is lots of data). 

It’s worth adding that in the example below GA recognises Affiliation as it’s own channel but sometimes this may appear as “Other” or even “Referral”. It will all depend on how your UTM-parameters are set up but in most cases, Google does the hard work for us!

Compare to the First Interaction Model

Now that you have a view that’s easy to follow; which demonstrates the number of sales and revenue on a last click basis, you need something to compare against. Where the greyed out text for “Select model” is written, you pick the drop down of “First Interaction” which should give a view like this:

You now have a way of comparing what each channel looks like in relation to how well it introduces consumers to your site (First Interaction) and how well it converts consumers that have already been on your site (Last Interaction). 

In this particular example it looks like the affiliate channel introduces consumers 65% less frequently than it converts consumers which is good for the revenue numbers of the affiliate channel but not so good for building brand awareness and making a wholesome contribution to your marketing mix.

Dig deeper into specific affiliates

There are warnings to heed as using the specific example above creates a view on the affiliate channel as a whole but of course the channel can have hundreds and sometimes thousands of partners, all acquiring traffic in different ways. It’s in your interests to dig deeper and understand which of your partners is effective at “introducing” and which are primarily “converters”. 

So let’s first hone in on just your affiliate performance by clicking on the “Affiliation” label (or whatever the equivalent is for your Source). You then choose the secondary dimension (in most cases this will be “Campaign”, but it can also be “Medium” or “Content” depending on your initial GA setup). Either way, you should end up with a view like this:

In the example above you can now see that there are some individual affiliates who introduce more often than they convert and these are the types of affiliates that you can more confidently say are “worth it”.

Once you’ve read and analysed your own results, you may find it useful to refer back to our article on are affiliate programs worth it utilising the benchmarks to infer which websites you work with make a difference to your brand growing efforts.