What everyone gets wrong with A/B testing

Marketing porn.

Here’s a definition:

”Stuff that makes you feel sophisticated when you talk about it at conferences.”

Chief culprit?

AB testing.

Oh, AB testing…

Don’t you just feel grown up when you’re AB testing?

The most AB testing I ever did was in my first 1-2 years of business.

(Tells you something, doesn’t it?)

I was sold by the stories of large companies that split tested their way to a bazillion dollars.

I would spend money on fancy software that could AB test things anywhere on your website, and feel extremely clever whenever I hit “start”.

I was acting like a teenager who just discovered their parent’s credit card

However…

99% of my time on AB testing was completely wasted.

Here’s the problem…

AB testing only works if you can get meaningful results fast enough.

There are two parts to that sentence:

  1. Meaningful results
  2. Fast enough

Example…

Let’s say you’re split testing the headline on your sales page.

When can you claim to have a meaningful result?

Answer:

At least 100 conversions on each variation.

That means 200+ sales.

Can you get that?

Great!

But how long did it take?

For a lot of businesses, it’ll take you 6+ months to get that result.

During which time you can’t touch ANYTHING ELSE on your sales page, because it’ll invalidate the test.

And if you’re smart enough to get a lift, how long till you recover the money you’ve spent on the work?

If you’re paying a CRO team $20k/mth to conceive of and run these tests, you’ll likely be old and grey by the time you actually get a hard cash return.

See the problem?

It’s one thing if you have Amazon-level traffic…

But if you’re Joe the YouTuber collecting 50 leads a day, it’s probably a waste of time AB testing anything on your sales page.

Numbers are too low.

Conversions take too long.

So does that mean you should give up on AB testing altogether?

No…

Because if you move further up the funnel, you CAN get more meaningful results, because you have a much volume of conversions to measure.

For example, landing pages:

  • squeeze pages
  • newsletter signup pages
  • lead magnets

Most of your available traffic is probably pointing to these pages, as opposed to the 1% who make it to your sales page.

More:

Every email you capture is super impactful, because everything else is downstream from lead capture.

Get more emails? Make more sales.

So if you’re going to AB test anywhere — do it on the pages where you collect emails.

Just make sure you’re running smart tests.

Namaste,

Olly

Share your thoughts

Your email address will not be published. Required fields are marked

{"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}