Performance-Based Testing: Avoid the Most Common Affiliate Mistake

Psst… Do you test?

Everyone says they test. Do you?

Sure, it takes extra time, which you say you don’t have. You stress, develop creative, throw it out to the millions, and start again, repeating the same old mistakes. That approach takes lots more time.

It’s tiring for you and your customer.

Everyone talks about testing, but few do it. It’s easy to throw a bunch of ideas against the wall, hoping one ad sticks, and call that testing:

• It’s not a test and it’s not direct marketing, which is what affiliate programs are all about. It’s guessing, which you can’t learn from.
• It’s not rocket science, but it is a science. It’s learned more through experience than in a classroom. Most entry-level marketers know little or nothing about direct marketing, although it’s the central function of their job.
• It’s a discipline. It doesn’t take genius, but it requires tenacity. The foundation of direct marketing is the practice of testing.

Successful affiliate programs incorporate direct marketing principles that have been practiced for years. They require good headlines, copy, and creative. They focus on conversions. They live and breathe numbers, not opinions.

What you like doesn’t matter. What the customers like does. The numbers will tell you whether they like your offer or not, quickly. Especially if you test.

How to Conduct an Affiliate Email Testing Campaign

The basis of testing is using a control, A, against a test, B.

It’s important to understand basic factors of testing. We’ll focus on an affiliate email campaign to keep things specific (this practice can be applied to virtually any marketing medium, such as banners or pops):

1. First ask yourself: What are you testing? Each element you test is a variable. For example, you can test these variables:
   • Subject line
   • Link placement in the message (top, bottom, middle)
   • Ad copy effectiveness (what makes them click and act?)

2. Isolate the variables. You have one control email and a test email with variables. For example, you could use a control sender line and a sender line with a different approach. You could test how many people open the message and/or the click-through rate of the creative.
3. When you know what you want to test, the rest is easy. Send the A email (control) to your list. Send the B email (test) to a representative sample. Know the control results, and try to improve them by tweaking copy, headlines, and subject lines.
4. Give the email a week to return results, and adapt the campaign from there.
5. Practice direct marketing with your affiliates. If you’re lucky, 5 percent of them have evolved into resellers for you. Deliver all the messages with the creative already coded. Make it easy for them to send a few tests.
6. Try to test a minimum of 50,000 emails through affiliates (or 20 percent of the list, if possible. If that’s not possible, the lists may be too small).
7. Many bigger affiliates won’t test at all, so focus on smaller niche partners to help your efforts and theirs. Use affiliates to benchmark your cost per action (CPA) (see my previous article).
8. Click here to see a sample test Excel spreadsheet with the variables outlined.
9. Focus on just a few variables. Don’t try to test everything.
10. A valid test of a list depends on the size of the list. A good estimate is to test to 10 to 20 percent of the entire list. Use the control for your standard emails.

Completing the Process: Create an A Email and
a B Email

Implementing an email campaign involves the next steps you are about to take. You simply create two versions of the email, differing only by the specific variables you mentioned.

To start your testing process:

1. Create an A email version in HTML, AOL, and text.
2. Create a B email version in HTML, AOL, and text.
3. Track open rates, click-throughs, and conversions; use your affiliate software to track to transaction or the action you want them to take. Use a graphic from your affiliate program in the HTML email to determine open rates (total opened emails/HTML emails = open rate).
4. Upload your creative to your affiliate program.
5. Be sure to enter your campaign into the email test spreadsheet and track your results.

Direct marketing is based on numbers. Numbers come from testing small and scaling out your offer.

Because affiliate advertising is so cheap, many marketers get sloppy. We throw stuff against the wall. Who cares what sticks, as long as something does? The customers — the people you’re trying to market to — care. Don’t treat them like a wall you’re constantly throwing ads against. They’ll get tired. They’ll tell you by their lack of response, as do your affiliates when your program doesn’t sell.

Select a few partners, pay to test your own creative, and work with affiliate partners based on results, not hope.

When it’s a game of numbers, you can win. When it’s a guessing game, you’ll lose. Branding is important, but revenue pays the bills.