What makes a good hypothesis?

Hi there 👋,

Did you ever watch that old film, Chicken Run?

I swear there’s a point to this, so bear with me.

Aside from being a great film, one scene stands out in particular right now. It’s when all the chickens are clucking around in a panic.

Well, they’re literally running around like ‘headless chickens’.

I mention this, because I see this kind of behaviour happen so often with experiments.

We get so excited about experimenting, that we go in blindly.

We don’t make a plan, we don’t have an outline for success, and we certainly don’t have a hypothesis.

But conducting an experiment without a hypothesis is the equivalent of chickens running around screaming. Nothing good can come of it.

What is a hypothesis?

Simply put, a hypothesis is an assumption based on the evidence provided.

This is one of the most important parts that you need to define before you do anything else.

Always make sure to set a clear hypothesis for your experiments.

Here is the basic structure I advise using:

• Because we saw A

• Then we expect B

• Due to change C

• We are right if metric D changes by E% (D = Success metric column, E = We are right column)

Let's try putting that into an example.

Because we saw User Generated Content videos 1 and 2 perform well on Meta ads.

Then we expect a combination video to perform even better.

Due to taking the best content from both of them.

We are right if, two weeks later, the Cost per Acquisition is at least 20% lower than the original ads with at least 50 purchases.

We noticed something, we acted upon this observation, and we outlined a marker of success. It's really as simple as that.

Recommendation

In every edition of Growth Waves, I also share a related book, individual or newsletter to check out related to the week's topic.

If you’d like to learn more about formulating a hypothesis and upgrading your experimentation process, check out my course, Growth Experiment Tracking.

In this 5-stage course, we cover everything you need to know about the experiment process.

From formulating a hypothesis to analysing your results, we’ll leave no stone unturned.

As part of the course, you’ll receive my experiment tracking spreadsheet, and the chance to ask questions throughout, for a total of £145.

If you haven’t done so already, click here to fill out my Growth Waves newsletter improvement survey. Your feedback is vital to allowing me to tailor my content to suit your needs and interests - I want to understand more about your industry, problem areas, and preferred content types.

Remember, one lucky respondent will win free access to one of my online courses:

• A Crash Course to Message-Market Fit - usually £45

• Growth Experiment Tracking - usually £145

Most experiments fail before they’ve even started, simply from a lack of a hypothesis.

But by outlining what you expect to find, and how you’ll discover this, you get to avoid that Headless Chicken Syndrome.

Instead, you can be like the happy, calm chickens at the end of the film.

To find out how they get there, you’ll have to grab your popcorn and settle down for a movie night.

Daphne