My growth design checklist 🧪

How I decide what to test next

Download article PDF​

When I’m planning growth experiments, I try to keep things simple.

No big frameworks or long RICE score calculations. Just a loose rule-of-thumb process that’s helped me ship things that are more likely to move the needle.

I don’t believe in silver bullets or magic unicorn dust that fixes everything, but I do believe in cutting to the chase in order for people to make decisions quickly.

Speed of execution is one of the things that separates great teams, and I think it has become increasingly important in the past few years of product development.

So - at the risk of being too deterministic (i.e. leaning on truths in a world with no absolutes) - here’s the checklist I run through when prioritising. I’ll use referrals as the example throughout, as I’ve been diving into those recently.

My growth design checklist ✅

In short:

• ✅ Step 1: Start with the bottleneck
• ✅ Step 2: Test UX first: can people get to it?
• ✅ Step 3: Check if the behaviour is active or passive
• ✅ Step 4: Test UI next: can people see it?
• ✅ Step 5: Test the copy
• ✅ Step 6: Test the wildcards, the fun stuff, the frills

Now let’s dig into each:

Step 1. Start with the bottleneck

Most of you know this by now, but even I fall into the trap of sometimes coming up with an idea before I’ve even checked if there’s an issue.

Don’t start with the idea. Start with the problem.

Where are users dropping off? Not the ugly homepage. Not the MVP feature you’ve been itching to redesign. The actual bottleneck.

Get the data into a spreadsheet and look at the drop-off. Where are most users falling off? That’s where your attention should go.

Action to do now: do a full funnel analysis of your product or feature, ask:

• Top of funnel: how many people came in the door in the past 30 days?
• Middle of funnel: how many people saw or clicked on the thing?
• Bottom of funnel: how many people successfully used the thing?

With the example of referrals, this would be:

• Total new users in the past 30 days
• Total users who got to a screen with referral on it in the past 30 days
• Total users who clicked refer in the past 30 days
• Total invitees who accepted the referral in the past 30 days
• Total new users from referrals in the past 30 days

Then look at the conversion % between steps. Which one is painfully low? There’s your bottleneck.

​

Step 2. Test UX first: Can people get to it?

Once you know where the drop-off is, the next big win is making sure people can reach the thing you want them to do.

Funny story: I once ran 6 tests on a screen before realising only 1% of users got to that screen in the first place. Oops. As soon as I moved the screen earlier in the experience, we boosted conversion rate from 0.5% to 15% 😭

For example, with our example of referrals, if you saw that only 2% of people clicked a referral button, your first question might be: how many users have the chance to refer? Is it deep in the funnel, or surfaced early when people are most engaged?

Your first step should be UX tests — testing things like:

• Add more referral entry points
• Place referral CTA in top or side nav
• Include referrals in settings
• Introduce a referral step in onboarding
• Move naturally shareable moments earlier

Action now: brainstorm ways you can give more people a chance to interact with your feature

​

Step 3. Check if the behaviour is active or passive

This is similar to Step 2 but worth calling out separately.

Are you hoping someone stumbles across it? Or are you actively placing it in their path?

This is how I see the difference between active and passive:

• Passive: you leave something for people to find (i.e a referral reward in the navigation)
• Active: you put it in front of them (i.e. a referral reward pop-up on second entry into the product)

There’s a fine line here with annoying people vs providing value. You’ll win if you’ve looked into the psychology and deeper emotions driving people to take an action.

Action now: brainstorm ways you can literally put your feature in front of people’s faces.

​

Step 4. Test UI next: Can people SEE it?

The number of times that UI is just 😭 too 😭 small 😭

Another basic checkpoint: can people actually see the thing?

Big BIG thing here is mobile versus desktop. I had a client designing on desktop when 70% of new users are on mobile, and they couldn’t actually see the quiz buttons without scrolling.

Make it big enough. Make it obvious.

And if your audience skews older or less tech-literate? Go 30% larger than you think you need to. Get actual people of that audience to take a look, you’d be surprised, trust me.

Action now: scrunch your eyes on your screen, how much can you see? You should be able to see the headline and the key areas.

​

Step 5. Test copy.

Sometimes the fastest unlock is just better words.

More specifically: fewer words.

Cut cut cut cut cut ✂️

You can have the ugliest experience but the most compelling copy and get users to your conversion moment.

Use sites like Hemingway Editor to check you’re writing for a 5-year-old. Or, the good old five-second test.

Action now: cut words on your screen by 30%. And make them bigger (or sort the hierarchy out so people can read it more easily).

​

Step 6. Test the wildcards, the fun stuff, the frills

This is where the playfulness, delight, and creativity come in.

Gamification, social proof, incentives, variable rewards — whatever feels right for the moment and the user.

But don’t start here. These things amplify what’s already working — they don’t fix broken flows.

Action now: brainstorm 3 wild ideas. Or screenshot your UI and plop it into ChatGPT for a sparring partner.

​

To conclude, the world is your oyster 🦪

Ta da! That’s the checklist.

Loose, fast, and focused.

It’s based on what I’ve seen move the needle time and time again with clients.

If you think ‘damn I missed that,’ no stress.

If you think ‘Rosie, I know this all already,’ then great! There’s no secret sauce, just keep testing.

It’s so easy to focus on details when you’re working day-to-day in the weeds.

This checklist is designed as a cold bucket of icy water to break out of the pixels and take a bird’s-eye view of your product.

Use it, change it or share it with your team if you’re stuck on what to test next.

OR tell me what’s missing to get featured in V2 😎

Thank you SO much for reading (all the way to the bottom, wow look at you go).

See you next week,

Rosie 🚲 🚲 🚲 I'm probably in Brittany at the moment 🥞

​​