Retention, Churn Mitigation & Lifecycle Email AutomationPlaybook3 min readUpdated September 2026

Finding Out Whether Your Value Proposition Actually Lands

Most value proposition debates happen entirely inside the company, in a meeting where people argue about wording based on their own instincts, with no actual customer input involved until the messaging is already live. That's backwards: resonance is something customers determine, not something a team can reason its way to from a conference room.

Here's a runbook for testing whether your messaging genuinely lands, using signals you already have access to.

Vendors Covered in this Article

Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.

Start With Language Customers Already Use, Not Language You Invented

The strongest messaging usually borrows language customers already use to describe their own problem, pulled from support tickets, sales call notes, and customer interviews, rather than language a marketing team invented independently. Before testing new value proposition wording, pull a sample of how customers describe the problem you solve in their own words.

A gap between your internal language and customers' actual language is often the real reason messaging isn't landing, and it's a gap you can close before running any formal test at all.

Where should you test new messaging first?

A new value proposition doesn't need to launch on the homepage to get its first real signal. Test variants in a lower-stakes, faster-moving channel first, subject lines on a lifecycle email, ad headline variants, a landing page section, where you can read a response within days rather than committing to a major site change before you know whether the message works.

This sequencing, cheap channel first, expensive channel once validated, catches a weak message before it's baked into your highest-visibility surfaces.

Does behavior or stated preference show what resonates?

Asking customers directly which message they prefer in a survey produces a different answer than watching which message they actually respond to with real behavior, clicks, replies, conversions. Stated preference and revealed preference frequently diverge, and revealed preference is the more reliable signal when you can measure it.

Use surveys and interviews to understand why a message resonates or doesn't, but trust behavioral data, what people actually click and convert on, more than a stated ranking when the two disagree.

Watch for a Message That Wins the Click but Loses the Sale

A message can perform well at the top of the funnel, a strong open or click rate, while attracting the wrong kind of prospect who doesn't convert further down, or who converts and then churns quickly because the message oversold something the product doesn't fully deliver. Track a winning message all the way through to a downstream outcome, not just the immediate engagement metric, before declaring it the winner.

A message that wins clicks but loses the sale, or wins the sale but loses the customer within a quarter, isn't actually resonating, it's just attracting attention without matching reality.

Segment Resonance Testing by Audience, Not Just Overall

A message that resonates strongly with one segment of your customer base can perform poorly with another, and testing only at the aggregate level hides that split. Break resonance testing results down by the segments that matter for your business, industry, company size, how someone found you, rather than reading only the blended result.

A message optimized for the blended average can end up being mediocre for every individual segment rather than excellent for any of them, which is a common and avoidable outcome of skipping this breakdown.

Keep a Record of What's Already Been Tested

Message testing suffers from the same institutional memory problem as creative testing generally: the same wording gets retested because nobody remembers it was already tried and didn't work. Keep a shared log of tested value proposition variants, the segment, the result, and the date, so testing effort compounds into real knowledge about your specific audience rather than repeating in circles.

Over a year or two, this log tends to reveal a small number of recurring patterns specific to your audience, maybe a particular framing around time saved consistently beats one around cost saved, or a message emphasizing control outperforms one emphasizing simplicity for a certain segment. Those patterns are more durable and more useful than any single test result, and they only become visible once enough tests have been logged in one place to compare.

Martha, an AI CMO, can help maintain this log and pull the language customers use from support and sales notes, but interpreting why a specific message landed or didn't still benefits from someone close to the customer relationships.

A runbook for testing whether your messaging lands:

  1. Collect the language customers already use from support tickets, sales call notes, and customer interviews before writing new variants.
  2. Test the variants in a lower-stakes, faster-moving channel such as a lifecycle subject line, an ad headline, or a landing page section.
  3. Read actual clicks, replies, and conversions instead of relying only on which message people say they prefer in a survey.
  4. Follow winning messages down the funnel to confirm they attract customers who convert and stay, not just click.
  5. Break results down by segment and log every tested variant with its segment and outcome so wording is not retested.
Executive Capability Standard

What Good Looks Like

Good messaging resonance testing means starting from language customers already use, testing in a low-stakes channel before a major surface, trusting behavior over stated preference, and tracking results by segment and all the way to a downstream outcome.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Pull a sample of how customers describe your problem in their own words from support tickets and sales notes before drafting any new messaging.
2. Do Manually:Test a new message variant in a single lifecycle email subject line by hand and track it through to a downstream outcome, not just the open rate.
3. Delegate:Have a lifecycle or content marketer own a shared log of tested message variants, including results by segment, so testing effort compounds over time.
4. Automate:Set up your email or ad platform's native testing tools to run message variant tests automatically on your highest-volume, lowest-stakes surfaces.
5. Buy:Bring in a messaging or positioning consultant if internal debate about wording keeps stalling without a clear way to resolve it using real customer signal.

How to Get Started

Disclosure: We may earn a commission if you buy through some links on this page. It doesn't change what we recommend.

ConvertKit

A platform like ConvertKit is a reasonable place to run early, low-stakes message variant tests, like subject lines, before committing a winning message to a higher-visibility surface.

Visit ConvertKit→

Frequently Asked Questions

Where should new messaging be tested first?

In a lower-stakes, faster-moving channel, a lifecycle email subject line or an ad headline, rather than a major site surface like the homepage. This lets you read a signal within days and catch a weak message before it's built into your highest-visibility pages.

Should we trust survey answers or actual behavior more?

Behavior, when you can measure it. Stated preference in a survey and revealed preference through actual clicks and conversions frequently diverge, and behavioral data is the more reliable signal for which message genuinely resonates, even though surveys are still useful for understanding why.

How do we know if a winning message is actually good for the business?

Track it past the immediate engagement metric to a downstream outcome, whether it converts to a sale and whether that customer sticks around. A message that wins clicks but attracts customers who churn quickly isn't resonating, it's overselling something the product doesn't fully deliver.

About the numbers

This guide doesn't quote a sourced benchmark. Figures in it are estimates or general guidance, so check them against your own numbers.

Related Guides