Review a page before you start testing

Use this playbook when you have a live or nearly finished page and want to decide what is actually worth changing. The goal of a CRO review is not to generate a long list of subjective design opinions. It is to identify customer decisions that may be difficult, unclear, unsupported, or distracting, then turn those observations into testable hypotheses.

Start with the visitor, not the page

Before reviewing the design, define:

TRAFFIC SOURCE
[Meta / TikTok / Google / creator / email / other]

AUDIENCE
[audience and awareness]

CAMPAIGN PROMISE
[why they clicked]

PRIMARY ACTION
[what should they do]

PRIMARY BUSINESS METRIC
[CVR / RPS / AOV / other relevant metric]

A page can be “good” in isolation and still be wrong for its traffic.

Review the page through six questions

1. Does the first screen make sense after the click?

Check:

campaign message match
product continuity
offer continuity
clear primary action
enough context to know what the page is about

2. Is the value proposition specific?

Look for language that could belong to any competitor:

“premium quality”
“designed for you”
“game-changing”
“feel your best”

Replace generic language only when you have a more specific, approved product truth.

3. Is proof close to the claim it supports?

Ask whether the visitor sees enough credible evidence at the moment skepticism is likely to appear. Proof can include approved:

product facts
ratings and reviews
creator or founder material
comparison evidence
guarantees or policies
certifications or awards

Do not score a page higher simply because it has many badges. Relevance matters more than quantity.

4. Is the buying decision easy to understand?

Check:

what the visitor is buying
variant or option labels
one-time versus subscription choice
actual offer
CTA action
next step after the click

A visual product card can look clear while still sending the wrong variant or purchase choice to cart, so final QA remains necessary.

5. Is the page asking for too much attention?

Look for:

several competing CTAs
repeated benefits
long copy without new information
decorative UI that competes with product choice
sections that exist because “landing pages have this section” rather than because the visitor needs it

6. Does mobile preserve the same hierarchy?

Mobile is not a smaller desktop screenshot. Check whether the important message, proof, and action still appear in a useful order without excessive scrolling or cramped interactions.

Ask Jurni for a diagnosis, not an automatic redesign

Use:

Review this page as a CRO analyst before making any changes.

CONTEXT
Traffic: [source]
Audience: [audience]
Campaign promise: [promise]
Primary action: [action]
Primary metric: [metric]

Evaluate:
1. message match
2. value-proposition clarity
3. proof and trust
4. objection handling
5. purchase / CTA clarity
6. visual hierarchy
7. mobile hierarchy
8. unnecessary friction or distraction

For each finding provide:
- OBSERVATION: what is currently happening
- CUSTOMER DECISION: what decision it may make harder
- EVIDENCE: what on the page or campaign suggests this
- HYPOTHESIS: the single change worth testing
- METRIC: the metric most directly connected to that hypothesis

Prioritize the top 5.
Do not edit the page yet.
Do not invent analytics data or claim a change will increase conversion.

Separate evidence from opinion

Not all observations have the same confidence. Useful evidence can come from:

experiment results
page analytics
high drop-off before a key step
recurring support or customer objections
campaign-to-page mismatch you can directly observe
repeated usability issues
clear inconsistencies in product/offer information

Lower-confidence ideas can still be worth testing, but label them honestly.

Prioritize by decision impact

A practical prioritization question is:

If this hypothesis is right, how important is the customer decision it changes?

Higher-impact decisions often include:

whether the page feels relevant after the click
whether the visitor understands the product
whether they trust the main claim
whether the offer is clear
whether they understand the purchase choices
whether the CTA takes them where expected

Changing a border radius can matter if it solves a real comprehension issue. Usually it does not deserve the same priority as the questions above.

Turn one finding into a test brief

Example:

OBSERVATION
The Meta ad leads with “no afternoon crash,” but the hero headline is a broad brand slogan.

CUSTOMER DECISION
The visitor must decide whether the page is still relevant to the ad they clicked.

HYPOTHESIS
Repeating the approved “no afternoon crash” angle in the hero will make the campaign more immediately relevant.

CHANGE
Hero headline and supporting sentence only.

KEEP
Product, offer, page structure, CTA action, proof, and everything below the hero.

PRIMARY METRIC
CVR.

SECONDARY DIAGNOSTIC
CTR / ATC rate, depending on the page flow.

Now you have a useful test—not “make the hero better.”

What good looks like

A good CRO review ends with fewer, stronger hypotheses. You should know:

what problem you observed
what customer decision may be affected
what single change you want to test
what will remain constant
which metric matches the hypothesis

Common mistakes

Asking AI to optimize the entire page. You may get a nicer page but learn very little. Treating best practices as evidence. “Put the CTA above the fold” is a hypothesis unless your own context makes the issue obvious. Choosing tests because they are easy. Test important decisions first. Using analytics without context. A number does not tell you why it happened. Changing several major things in one variant. If it wins, you will not know which change mattered.

Page Building Guides

Prompting Guides

Templates & Reusable Sections

Experimentation Guides

Review a page before you start testing

Review a page before you start testing

Start with the visitor, not the page

Review the page through six questions

1. Does the first screen make sense after the click?

2. Is the value proposition specific?

3. Is proof close to the claim it supports?

4. Is the buying decision easy to understand?

5. Is the page asking for too much attention?

6. Does mobile preserve the same hierarchy?

Ask Jurni for a diagnosis, not an automatic redesign

Separate evidence from opinion

Prioritize by decision impact

Turn one finding into a test brief

What good looks like

Common mistakes

Next playbooks

​Review a page before you start testing

​Start with the visitor, not the page

​Review the page through six questions

​1. Does the first screen make sense after the click?

​2. Is the value proposition specific?

​3. Is proof close to the claim it supports?

​4. Is the buying decision easy to understand?

​5. Is the page asking for too much attention?

​6. Does mobile preserve the same hierarchy?

​Ask Jurni for a diagnosis, not an automatic redesign

​Separate evidence from opinion

​Prioritize by decision impact

​Turn one finding into a test brief

​What good looks like

​Common mistakes

​Next playbooks

Review a page before you start testing

Start with the visitor, not the page

Review the page through six questions

1. Does the first screen make sense after the click?

2. Is the value proposition specific?

3. Is proof close to the claim it supports?

4. Is the buying decision easy to understand?

5. Is the page asking for too much attention?

6. Does mobile preserve the same hierarchy?

Ask Jurni for a diagnosis, not an automatic redesign

Separate evidence from opinion

Prioritize by decision impact

Turn one finding into a test brief

What good looks like

Common mistakes

Next playbooks