How to test offers (without breaking your brand)

We're wrapping up this series today.

Over the last four emails, I've walked you through …

Today I want to talk about testing.

Because here's what usually happens …

A founder reads everything I just sent, nods along, and then says …

"Okay, I'm going to test this."

Which sounds smart.

But then they test it the wrong way … and either (a) declare it a failure too early, or (b) call it a win based on metrics that don't actually matter.

So let me show you how to test offer changes without accidentally breaking your brand in the process.

Here's the most common mistake …

You change your offer. You watch conversion rate for a week. If CVR goes up, you call it a win. If it drops, you revert.

This is how you optimize yourself into a corner.

Because CVR is a leading indicator … but it doesn't tell you if you attracted the right behavior.

You can easily boost CVR by …

And all of those tactics can work short-term.

But they often attract low-commitment customers who churn fast, need more support, and never hit positive LTV.

So you "won" the test, but you quietly made your business worse.

When you test an offer change, you need to measure downstream health … not just checkout behavior.

Here's the scoreboard I use with clients …

If your CVR goes up 10% but your refund rate doubles and subscription survival drops, you didn't win. You just delayed the problem.

The goal is not to maximize conversion.

The goal is to maximize successful starts.

Here's the framework …

Bad hypothesis … "If we add urgency, CVR will increase."

Good hypothesis … "If we make the 90-day option the obvious default, customers will commit more deeply and churn less."

You're testing guidance, not force.

Offer changes need at least 30–60 days to show their real impact.

Because the downstream effects (retention, reorder rate, support load) don't show up in week one.

If you're testing on a 7-day window, you're just measuring noise.

Don't just look at "overall CVR."

Break it down …

Sometimes an offer change works great for one segment and poorly for another. You need to know which.

When you change the offer, keep an eye on …

These are the signals that tell you if you're building momentum or creating friction.

If you're not sure where to start, here's the simplest, highest-leverage test …

Make your "Start Here" option unmistakably obvious.

Don't add urgency. Don't discount. Don't change the price.

Just make the correct starting path structurally clear.

Visually bigger. Labeled "Start Here" or "Best Way to Begin." Positioned as the default.

Then measure …

I've run this test dozens of times with clients.

Sometimes CVR dips slightly in the first week (because you're not bribing people anymore).

But almost always, the downstream metrics improve … fewer cancels, better retention, smoother cash flow.

And that's the trade you want to make.

Testing is good.

But testing without a hypothesis about customer success is just gambling.

You're not trying to find the offer that converts the most people.

You're trying to find the offer that guides the most people to start correctly.

Because customers who start correctly …

And that's how you build a business that scales profitably … not just a business that converts.

Alright, that's it.

Five emails. One big idea …

Belief without a path creates paralysis.

Your customers aren't hesitating because they don't believe you.

They're hesitating because they don't know how to start safely.

And if your offer doesn't make the first step obvious, they'll under-commit, evaluate too early, and quietly disappear … no matter how good your product is.

The fix isn't more urgency, more discounts, or more options.

It's clarity.

Make the right way to begin unmistakable. Align your offer to the real timeline. Encode expectations up front.

Do that, and everything downstream gets easier.

If this series resonated with you … or if you're stuck on how to apply it to your business … just hit reply and tell me where you are.

I read every response. And if I can help, I will.

See you tomorrow,

Jeremiah

P.S. … One last thing … I know this series was dense. If you want to revisit any part of it, just reply and ask. I'm happy to go deeper on any piece … whether it's the one-bottle trap, the testing framework, or how to handle multiple product lines. I'm here.

100% Typo Guarantee … This message was hand-crafted by a human being … me. While I use AI heavily for my research and the work I do, I respect you too much to automate my email content creation.

There was no review queue, no editorial process, no post-facto revisions. I just wrote it and sent it … therefore, I can pretty much guarantee some sort of typo or grammatical error that would make all my past english teachers cringe.

Anonymous Data Disclaimer … Most of my clients prefer that I not share the inner workings of their businesses or the exact details of the marketing strategies we develop. In order to be able to share my own proprietary intellectual property without violating the sensitive nature of my relationship with them, I often anonymize what I share with you. This may include changing the specifics of their industry, what actually happened, or what we developed together. When I make these changes, I work to preserve the success principle I want to convey to you while obscuring sensitive data. This is necessary.