Tracker.ly can distribute traffic from multiple redirect links between two or more destinations. This is often used to split-test two or more affiliate links to learn which one is more profitable or converts better. Depending on your goals, the winner could be the one that converts the best or the one that earns the most and that may or may not be the same product. For example, when split-testing two products, the less expensive product may convert better than the more expensive competitor, but the higher payouts on the more expensive one, might mean that overall, it earns more, despite the lower conversions.
Currently, you will need to login to the affiliate program's back-end and compare sales and revenue over the period you are running your split-test to determine the winner. Later, we'll have the ability to import sales data from affiliate programs that allow you to export it.
To send redirect links to multiple destinations, you simply select two or more destinations instead of one on the screen to select the destination. That will start the process of defining how the split-test will work and create a collection of destinations, which you can send other redirect links to in the future.
You can point redirects to a split-test in four ways:
1) Go to the management page of any redirect link, and click the button to point the redirect link to a new destination.
2) Go to the management page of any destination link, and click the button to point the destination's redirect links to a new destination.
3) Go to the "Create / Manage Split-Test Offers" button on the home page of the wizard in the "Boost Traffic and Profits" section, and then click the "New Split-Test" button.
4) Create a new redirect link and point it to several destinations.
In this example, I've created a new redirect link, and on the page where you select the destination, I've selected two destinations. When on this page, if the new destination doesn't exist yet, you can create it using the "New Destination" button, and you'll return to this page when done:
Once you have selected multiple destinations and click the "Create Split-Test" button, you next give the split-test a reference name and tell it what type of split-test to create. There are four different types, but for split-testing affiliate links, you only use the first two types.
1) Split-test that remembers destination:
This is recommended for split-testing affiliate links. We set a cookie in each visitor's browser to remember which destination we sent them to the first time. Then, every time they click the redirect link, they go to the same destination. So, each person sees the same destination every time they click your redirect links, even though it is part of a split-test. This gives each person a consistent experience and avoids the confusion of wondering why the link is taking them to a new place each time they click it.
2) Split-test that forgets destination:
Alternatively, you can send each person to a random destination in the split-test each time they click, by using this type of split-test. This is also recommended for testing banner ads, where the redirect link goes to one of several destination images. It does not remember what each visitor saw before, so each click or impression will get a random destination. It may be the same, or it may be different.
3) Rotate through list over and over:
Each time someone clicks on a link, they are given the next destination in sequence, rotating through the list of destinations continuously. For example, if you were to put a redirect link on your page to a banner image, every time you visit the page, you'll see the next image, rotating through the list over and over. If you had 3 images, and visited six times, you would see images 1,2,3,1,2,3. Someone else coming to the same page would always start at the beginning of the sequence. If you wanted everyone to get a random image (destination) on each visit and make sure that each image gets equal distribution, you would use a split-test that forgets the destination.
4) Go through list and stay on last:
Each time someone clicks on a link, they are given the next destination in sequence, until they reach the last one. Then, they will always see the last one. For example, if you had a sequence of four pages, like a primary offer, their second chance, their last chance offer, and the missed offer page, each time they click the link, they go to the next page in the sequence, and when they reach the missed offer page, they keep seeing that. Each visitor starts at the beginning of the sequence.
Here, I am continuing to create a split-test of my two amazon mouse affiliate links from above. I'm giving it a reference name, "What's the best mouse?", and setting the type to the one best for split-testing:
With Tracker.ly, you can have multiple redirect links going to a split-test. Because of that, we use some clever math to let us randomly assign a destination to every click, while still making sure that, overall, each destination gets even distribution of clicks. That means that when you test your link, it's possible that you'll get the same destination more than once. Rest assured, the math works out when you have enough clicks to get any meaningful results.
It also means that on the second step of creating a spilt-test, you can assign other redirect links to also be traffic sources for the split-test. In this case, I have 2 existing redirect links, one for each destination, in addition to the new redirect link I'm creating. I'll point both of them to the split-test as well. If I had lots of redirect links pointed to the destinations being split-tested, I could use option three below to select some of them. Or, I can choose from all of the redirect links I have setup in Tracker.ly, to move them to the split-test:
In this case, once you click "Save", you'll be taken to the management page of the new redirect link you're creating. Here, you can see that the redirect link is sending traffic to a split-test, with 3 redirect links pointed to 2 destinations:
Sending more redirect links to an existing split-test:
Once you have a split-test created, you might want to create a new redirect link as a new campaign to send traffic to the split-test, or you might want to point some other existing redirect link to the split-test. In the first case, you create a new redirect as normal and in the second case, you go to the management page of an existing redirect link and click the "Point To Different Or New Destination" button below:
When you get to the screen to select the destinations, instead of selecting a destination, click the "My Split-Tests" button to list all of your existing split-tests:
Pick the split-test you want to send the traffic to, and you're done. If you pick a finished or cancelled split-test, Tracker.ly will create a copy of the split-test, so that your history remains intact. You can use that as a starting point for setting up the new split-test. In this case, I'm selecting the active split-test I made earlier:
Notes on Split-Testing:
1) If you have both scarcity and a split-test running on the same destination:
First we do the split-test, sending people to one of several destinations you setup in Tracker.ly.
- If the destination has a scarcity offer, we send people to the expired destination if it's expired.
2) Split-testing and link juice (301/302):
It probably is better to set your links to 302, which tells search engines that they are temporary redirects.
3) When testing a split-test:
If you have it set to remember, it will keep sending you to the same place, even though it is working correctly.
Switching browsers treats you like a new visitor, so it won't remember you between browsers.
- We use some cool math to randomly distribute the links equally across all visitors, so don't expect destinations to appear in sequence when testing. Over larger numbers it evens out.