Most people know about A/B testing… but what about A/B/C testing?
With any website and the larger sales funnel, the wise digital marketer is always testing for better conversion results. Does this have to be a major, time-consuming task? No. Do you have to push yourself? Yes.
Today we’ll look at A/B/C testing for a headline and consider both why we want to create such a test and how it compares to other forms of testing.
Let’s look at one we have running right now. We have written three landing page headlines and our testing hypothesis is that only one will compel more visitors to convert to leads. Luckily this is easy to do — no external testing provider required — using our preferred WordPress theme – Divi. Divi comes with built-in A/B (and C and D, etc.) testing. Here’s how it looks in the admin area:
We’re testing three headlines — an A/B/C test — using the built-in split tests with the Divi theme.
You can see above that #3 looks to be already winning, though it is early and we will need hundreds more unique visitors to have a valid testing result. In case you are curious, the test is running on this landing page. On this test we are looking at messaging factors for the word “Digital” (do more visitors respond to it when used?) and also for a stronger call to action like you see in version C.
Why do we use more than two variations in a split test?
Aside from the obvious reason of wanting to test three messaging variations to find out which will be more persuasive, there are a couple more reasons to use A/B/C testing:
- To exercise your copywriting chops. I find that writers become more persuasive this way because it forces us to focus more keenly on compelling phrasing.
- To look for messaging nuances beyond two choices. For example, with the analytics we get in Divi we can sometimes find out that one version caused fewer bounces, while another compelled more lead form submissions. You may then consider a “wedding” of wording in the final message.
- You have enough traffic to do a good A/B/C test, but not a full “multivariate” test.
Make sense? In case you’re wondering what a multivariate test is, think if we wanted to test three-page components — say headlines, plus images, plus wording on a form button. And, if we want to test three versions of each… but find the winning combination. The objective of multivariate testing is to determine which combination of variations performs the best out of all of the possible combinations. Advanced testing software lets you do this which in this case would mean that nine versions of a page are tested. This requires a lot more traffic to yield a valid result, usually, but it can be a powerful tactic to find the most persuasive combination of content elements on a page.
But what if your chief copywriter insists that they already know the best headline for a landing page? Short answer: they don’t. Neither does your boss. Nor does your expensive consultant. Why? Because the only ones who DO know are your site visitors, and everything else is just your first best guess. And an A/B/C test gives you the ability to test three guesses. Always be testing.
