4 MISTAKES MAKING YOUR A/B TESTS INEFFECTIVE
A/B tests are about more than moving CTAs around, changing labels and playing with colors?
It's a waste of time anyway
These days, 99% of people know how to use your website
So unless it's a complete mess
Chances are the problems that need fixing are actually in the messaging
So before I break this down any further
What I'm about to describe below is only a fraction of what's contained in the guide
So save yourself some time and just download it - it's free after all
But...
For those of you who need a little more convincing
I'm not saying there's never a need to test the UI
In fact we do it all the time
But if you've already got the basics in place
And are following standard web design practices
I've found that the majority of A/B tests that manipulate the UI
Are often showing a winner, not because of the design
But rather because of some bug that wasn't initially obvious on one of the iterations
And after you've resolve this
There's very few UI changes that will match the impact of simply revising the wording a nature of the offer
Which...
And this is where A/B tests can get interesting
Because once you've properly A/B tested your messaging
Then you can start building out much better UI interfaces
Tailored around exactly what you're saying to your customers
For example...
We had a client who we ran a series of tests for in order to help them increase average order value
And it turned out that simply directing the users attention to the fact they could buy something to accompany whatever item of clothing they had just added to the bag
Well this had a big impact
So once we got that messaging right
Then we started displaying visuals of the "related products" in the basket
Because the conclusion was that users just weren't thinking of buying another product
And this wouldn't have become clear without A/B testing the messaging first
Now
As I said above
This is literally a fraction of the detail that's contained in the guide so go it and let me know what you think
An innovative and actionable way to run A/B tests