Have you ever wondered what happened to the change in your pocket, only to discover that you had a hole?
Running split tests is like finding that hole, and then fixing it, so you don’t keep tossing away your money. The problem is that many people don’t know how to do a proper split test.
Today, I want to caution you to avoid making a mistake that lots of marketers make; they don’t do the split test in real time.
What I mean by that is, they might post one version of their ad or their sales page one week, and the next week post a different version. And they think that is a split test.
It’s not. Or at least, it’s not one that will give you any information that you can actually use.
Imagine what would happen if you took one direct mail letter and mailed it one week and you took another direct mail letter and mailed it another week. Let’s say the second of those weeks included Easter, Passover, and Good Friday.
What would happen? No doubt the second one that you mail out is going to suffer. That’s not necessarily at all because of the marketing piece; it’s because of what’s going on.
The only way to do a true A B split test is to do them in real time. You mail them both the same week – ideally on the same day – so that whatever noise is going on in the world, or in the marketplace, affects both tests equally.
Get a baseline measure
Let’s say you designed a funnel from scratch, never having driven traffic to it before. You use that as a baseline measure, so that any change after that point is always done via a split test. Maybe you try a video on one sales page, and a long-form sales letter on another.
Assuming that one of them performs worse than the other, you want to minimize the damage from that version as quickly as possible, and cancel it.
If you don’t run the split test in real time, your numbers are susceptible to “external noise”, like holidays, weather, etc.
This is especially true if you are running ads in multiple countries. You have no idea if there are different holidays or events going on; you’re making decisions about which ad performs best, so you need to be able to compare them in real time.
If you run one ad today, and another one tomorrow, you could find that one of them performs poorly. But you wouldn’t be able to tell if it was due to “external noise” or just because the ad wasn’t working.
So don’t make the mistake that a lot of marketers make. Compare your split tests in real time in order to make a decision with confidence about which one is truly working for you.
Getting this right is like putting money in your pocket.