Testing Affiliate Reinforcement Messaging
One of my favorite optimizations that I like to roll out with a new client is affiliate-messaging reinforcement. All too often, new client engagements are with folks that have never tested or are new to the functionality that is available from testing. Testing affiliate-messaging reinforcement is a relatively simple test to get up and running and has always provided quick wins. This type of test not only serves as a great opportunity to become familiar with the targeting capabilities of testing but also brings home the reason they are doing testing — to learn what works for their visitors and to leverage those learning’s for financial gains and better visitor experiences.
Affiliate traffic comes in many shapes and sizes. There are even companies out there like Commission Junction (CJ) that manage affiliate traffic for websites. Some of the more popular affiliate traffic generators out there include Upromise, Ebates and MyPoints. These types of affiliate sites have a group of regular visitors that are typically incentivized to make purchases or visits to other sites. Upromise, for example, incentivizes their traffic base to make purchases on the web through its referral network and by doing so you would get a discount from the merchant as well as some credit added to a an education account.
In order to get the incentive or the affiliate credit the visitor will have to make the purchase or convert only when directly arriving onto the website from the affiliate. Going directly to the referred site or going there after the session that brought you there from the affiliate will not work. This is so that affiliate gets the credit for bringing the visitor there. This is very important to keep in mind when setting up the test.
The premise is this optimization or test is simple. It is the continuance through messaging of the customer’s goal that leads to the increased conversion with high statistical confidence. Not all websites advertise or participate with affiliates but the principles of this test can be applied to any of your sources of traffic.
The test can be set up in no time at all. There are just two things needed to get up and running. The first is the creative you wish to present to the visitors that arrive from affiliate. If you get a decent amount of traffic then I would incorporate two variations of creative. Have one of the pieces of creative emphasize the affiliate source more prominently and the other piece emphasize the incentive that the visitor is entitled to get more prominently. The results of that scenario could be very interesting as you get to see what drives a greater impact to this segment of traffic.
The second necessary piece would be a way to differentiate affiliate traffic from other sources. If you don’t have many affiliates it might be easy to have a custom-landing page for each affiliate but this could lead to problems. You want to only have the affiliate messaging reinforcement to be presented to the visitor upon directly arriving from the affiliate. Having a custom landing page may allow visitors to think they are getting affiliate credit when in fact they wont if they arrive at the landing page another way. The targeting functionality of testing will allow you to avoid this problem by limiting the content to be shown only during the arriving session.
I recommend using URL parameters to differentiate these affiliate visitors. To do so simply provide to the affiliate a URL like this to your affiliate:
The value of “affiliate_x” being assigned to the URL parameter “sid” identifies that the visitor arrived from the affiliate.
Once you have these two pieces together you are good to go!
Musician’s Friend, a major online retailer of musical instruments implemented just such a test. Here is their ad that they had presented on the affiliate Upromise. In the ad they offer visitors free shipping on any order over $99.
When visitors clicked through on this ad they would qualify for the campaign or test that was quantifying the value of the reinforcement. If they visitor was placed in Recipe B — they would see this ad in the top center of the page:
You can see that they ad not only reinforces the free shipping but also that they came from Upromise and were entitled to receiving the credit in the education account within Upromise.
Musician’s Friend saw a 35% conversion lift over the control with a 97% confidence level with this test.
If you are a Test and Target customer and would like to see how an affiliate campaign is set up in the tool and how to use session level targeting, please just let me know. I will then set you up with a login to my demo account, which has this test as well as a few others that might prove to be helpful.
If you have a login already the campaign is called “Affiliate Reinforcement” and the profile targeter to limit the visitor to the campaign is called “user.singlesession”.