A Practical CRO Workflow

30 Actionable CRO Prompts You Can Actually Use

AI can generate 100 headline variations in seconds. That's not conversion rate optimization. The hard part is figuring out why people aren't converting in the first place.

These 30 prompts are arranged into a practical conversion rate optimization workflow that you can use repeatedly. Think of it as a simple CRO operating system.

30 Practical Prompts

6 CRO Phases

1 Repeatable Workflow

The Goal

The goal isn't to make every page look more "conversion optimized." The goal is to help the right visitors take the right next action with less confusion and unnecessary friction.

Find the problem → Understand the visitor → Improve the experience → Test the change → Learn from the result → Repeat

Don't use AI to make random changes to your website. Use AI to make better conversion decisions.

Phase 1: Find Conversion Problems

Before changing your website, figure out where the problem actually is. A low conversion rate is a symptom. Your first job is to find the likely cause.

Diagnose a Low-Converting Landing Page

Act as a conversion rate optimization strategist. Analyze this landing page: PAGE: [URL/TEXT] TARGET AUDIENCE: [AUDIENCE] PRIMARY GOAL: [GOAL] TRAFFIC SOURCE: [SOURCE] CONVERSION RATE: [RATE] Evaluate the page across: - Audience-message fit - Clarity - Value proposition - Headline - Subheadline - Offer - CTA - Trust - Social proof - Objection handling - Page structure - Content hierarchy - Visual distractions - Friction - Risk - Information gaps - Mobile considerations - Consistency with the traffic source Do not immediately recommend rewriting the page. First identify the most likely conversion problems, explain the evidence for each, and rank them by potential impact. Separate: Evidence-supported problems Likely hypotheses Things that require additional data Then provide the five highest-priority improvements.

Use When

A landing page receives traffic but converts poorly.

Give AI

The page, audience, traffic source and conversion data.

Get

A prioritized CRO diagnosis.

Investigate the highest-impact problems before changing copy, design or CTAs.

Find Where the Conversion Funnel Is Leaking

Analyze this conversion funnel: Traffic: [NUMBER] Landing page visits: [NUMBER] CTA clicks: [NUMBER] Form starts: [NUMBER] Form completions: [NUMBER] Checkout starts: [NUMBER] Purchases/leads: [NUMBER] Identify where the largest drop-offs occur. For each stage provide: - Drop-off rate - Possible causes - Evidence needed - Questions to investigate - Potential improvement - Priority Do not assume the biggest percentage drop is automatically the most important problem. Consider both: Magnitude of drop-off and Business value of the lost users Then identify the first funnel stage I should investigate.

Use When

You have analytics data but don't know where conversion optimization should begin.

Give AI

Funnel-stage numbers and relevant business context.

Get

A prioritized funnel diagnosis.

Fix the biggest meaningful bottleneck rather than randomly optimizing individual pages.

Audit a Page Like a Brutal CRO Editor

Audit this page: [PAGE] Assume I don't want compliments. Tell me what could prevent a qualified visitor from taking the intended action. Critically evaluate: - First impression - Clarity - Relevance - Value proposition - Specificity - Credibility - Proof - Objections - Offer - CTA - Information hierarchy - Cognitive load - Friction - Unnecessary content - Missing information - Confusing language - Weak transitions - Conversion distractions For every problem provide: Problem → Why it matters → Evidence → Recommended improvement Separate major problems from minor copy/design preferences. Do not recommend changes merely because something is different from common landing-page templates.

Use When

You want honest criticism rather than another generic "your page looks great" audit.

Give AI

The complete page and target conversion goal.

Get

A direct CRO critique.

Fix the biggest barriers first. Don't spend an hour polishing a button while the offer remains unclear.

Find Conversion Friction

Analyze this conversion journey: [PAGE/FUNNEL] Identify every point where a visitor may experience unnecessary friction. Consider: - Too many steps - Too many fields - Unclear instructions - Confusing navigation - Repetitive information - Unclear pricing - Unexpected requirements - Weak CTA language - Uncertainty about what happens next - Trust concerns - Technical barriers - Slow or confusing transitions - Excessive choices - Unnecessary decisions For each friction point provide: Friction → Visitor concern → Severity → Possible solution → Evidence needed Rank the issues from highest to lowest priority. Do not recommend removing friction that serves an important qualification or security purpose without considering the trade-off.

Use When

Visitors appear interested but abandon before completing the desired action.

Give AI

The full conversion process and any available analytics.

Get

A friction map.

Remove unnecessary friction, not useful friction. Some friction protects the business or improves lead quality.

Find the Most Important CRO Problem to Solve

Here is my current CRO situation: WEBSITE: [WEBSITE] TRAFFIC: [TRAFFIC] CONVERSION RATE: [RATE] TOP CONVERSION GOAL: [GOAL] MAIN PAGES: [PAGES] FUNNEL DATA: [DATA] KNOWN PROBLEMS: [PROBLEMS] Create a CRO problem backlog. For every problem estimate: - Potential impact - Evidence strength - Number of visitors affected - Business value - Implementation difficulty - Testing difficulty - Risk of making things worse Then prioritize using: Potential impact × evidence strength × business value ÷ effort Do not pretend the score predicts exact revenue. Use it only as a prioritization framework. Finish with the three problems I should investigate first.

Use When

You have a long list of CRO ideas and no clear priority.

Give AI

Analytics, conversion data, website information and known issues.

Get

A prioritized CRO backlog.

Investigate the top problem before jumping to another optimization idea.

Phase 2: Understand Visitors & Intent

CRO isn't about optimizing for an imaginary "average visitor." You need to understand who is arriving, why they're there, what they expect, and what might stop them.

Identify Visitor Intent From Traffic Sources

Analyze these traffic sources and landing pages: [DATA] For each source, determine: - Likely visitor intent - Awareness level - Expected information - Expected offer - Likely objections - Appropriate page message - Desired next action Compare the visitor's likely expectation with what the landing page currently provides. Identify any message mismatch between: Traffic source → Search/ad/social promise → Landing page → Offer → CTA Rank the mismatches by potential conversion impact.

Use When

A page receives traffic from multiple sources but conversion performance varies significantly.

Give AI

Traffic-source data, campaign messages, keywords, ads, social posts and landing pages.

Get

An intent and message-match analysis.

Don't send every visitor to the same experience if their expectations are substantially different.

Build a Visitor Objection Map

For this offer: PRODUCT/SERVICE: [OFFER] TARGET AUDIENCE: [AUDIENCE] Create a visitor objection map. Identify potential objections related to: - Price - Trust - Quality - Risk - Timing - Need - Complexity - Alternatives - Results - Implementation - Compatibility - Commitment - Switching - Uncertainty For each objection provide: Objection → What the visitor may actually be worried about → Information needed → Proof that could help → Best place to address it Rank the objections by likely importance. Do not suggest manipulative techniques for eliminating legitimate concerns.

Use When

Visitors reach your offer but hesitate to convert.

Give AI

Offer details, audience research, sales objections, customer questions and reviews.

Get

A practical objection map.

Address the strongest objections where they naturally arise in the visitor journey.

Extract Conversion Insights From Customer Research

Analyze this customer research: [REVIEWS / INTERVIEWS / SUPPORT TICKETS / SALES CALLS / SURVEYS] Extract recurring: - Problems - Desired outcomes - Buying triggers - Objections - Questions - Frustrations - Alternatives - Decision criteria - Words customers use to describe the problem - Reasons customers choose the product - Reasons customers hesitate Group the findings by theme. Then identify: Messaging opportunities Proof opportunities Offer improvements FAQ opportunities CTA opportunities Landing-page content opportunities Do not invent customer language or conclusions unsupported by the source material.

Use When

Your website copy was written mainly from the company's perspective.

Give AI

Real customer research.

Get

Customer-derived conversion insights.

Use recurring customer language to make the page more relevant and specific.

Identify the Visitor's Unanswered Questions

Analyze this page: [PAGE] and this target audience: [AUDIENCE] Identify questions a qualified visitor may still have before taking the desired action. Group them into: - Understanding - Fit - Price/value - Trust - Implementation - Risk - Comparison - Results - Next steps For each question provide: Question → Why the visitor needs the answer → Where it should be answered → Recommended information Prioritize questions that could materially influence the decision. Do not add FAQs simply to make the page longer.

Use When

Visitors may be interested but still lack information needed to make a decision.

Give AI

Page, audience, offer and customer questions.

Get

A list of missing decision-support information.

Answer the strongest questions directly on the page or at the appropriate stage of the funnel.

Build a Visitor Decision Journey

Map the decision journey for a visitor considering: OFFER: [OFFER] AUDIENCE: [AUDIENCE] Start with: First awareness → Problem recognition → Solution research → Evaluation → Trust → Decision → Action For each stage identify: - Visitor thought - Main question - Concern - Information needed - Proof needed - Likely action - Potential conversion barrier Then compare this journey with my current page/funnel: [PAGE/FUNNEL] Identify which stages are well supported and which are weak.

Use When

Your page feels like it jumps directly from introduction to "Buy Now."

Give AI

Audience, offer, page and customer journey information.

Get

A decision-stage map.

Make sure the page provides enough information for the visitor's actual decision stage.

Phase 3: Improve Pages & Offers

Now we can start improving the experience. But don't optimize individual words in isolation. Optimize the whole conversion argument.

Clarify the Value Proposition

Analyze this offer and page: OFFER: [OFFER] PAGE: [PAGE] TARGET AUDIENCE: [AUDIENCE] Determine whether the value proposition clearly communicates: 1. Who this is for 2. What problem it solves 3. What outcome it helps create 4. Why the visitor should care 5. Why this solution is different 6. What the visitor should do next Then create five alternative value propositions. Each should be: - Specific - Audience-focused - Outcome-oriented - Credible - Easy to understand Avoid vague phrases such as: - "Take your business to the next level" - "Revolutionary solution" - "Unlock your potential" For each version explain the positioning choice.

Use When

Visitors don't immediately understand what you're offering or why it matters.

Give AI

Current page, product/service information and audience research.

Get

Clearer value propositions.

Choose the proposition that most accurately reflects the actual value of the offer.

Rewrite a Landing Page Around One Conversion Goal

Analyze this landing page: [PAGE] The single desired action is: [ACTION] Create a revised page structure that makes this action easier to understand and take. Before creating the structure, identify: - Visitor intent - Core problem - Desired outcome - Primary objection - Required proof - Offer - Conversion barrier Then structure the page as: Headline → Value proposition → Problem/context → Solution → Benefits → Proof → How it works → Objections → Offer → CTA → Risk reduction → Final CTA Do not automatically include every section. Explain why each recommended section exists. Remove sections that do not support the conversion goal.

Use When

Your landing page has grown into a long collection of sections without a clear conversion path.

Give AI

Current page, audience, offer and conversion goal.

Get

A conversion-focused page architecture.

Use the structure as a strategic guide, not a rigid template.

Turn Features Into Conversion-Focused Benefits

Here are the features of: PRODUCT/SERVICE: [PRODUCT] [FEATURES] Translate each feature into: Feature → Functional meaning → Customer benefit → Business outcome → Who cares → Evidence needed Then rank the benefits based on importance to: TARGET AUDIENCE: [AUDIENCE] Identify features that are impressive but unlikely to influence the purchase decision. Do not exaggerate the relationship between features and outcomes.

Use When

Your page reads like a product specification sheet.

Give AI

Verified product features and audience information.

Get

A buyer-focused benefit hierarchy.

Lead with benefits that matter to the visitor rather than giving every feature equal prominence.

Improve Social Proof

Analyze the social proof currently used on this page: [PAGE] Available proof: [TESTIMONIALS / REVIEWS / CASE STUDIES / RESULTS] Evaluate whether the proof addresses the visitor's actual concerns. Identify missing proof related to: - Product quality - Results - Reliability - Expertise - Customer type - Use case - Objections - Risk Then recommend: What proof should appear Where it should appear Which objection it addresses What evidence is still missing Do not invent testimonials, customer results or statistics. If a claim is weakly supported, flag it.

Use When

Your page has testimonials but they don't seem to reduce buyer uncertainty.

Give AI

Actual customer proof and page content.

Get

A strategic social-proof plan.

Use specific, relevant proof instead of collecting generic praise.

Reduce Cognitive Load

Analyze this page: [PAGE] Identify elements that may make the visitor work unnecessarily hard to understand or act. Evaluate: - Too many choices - Long paragraphs - Repetitive sections - Unclear hierarchy - Competing CTAs - Unnecessary navigation - Complex language - Dense forms - Ambiguous labels - Repeated claims - Distracting information - Unclear next steps For every issue provide: Problem → Why it increases cognitive load → Recommended simplification Then identify the five changes that could make the page easier to understand without removing important information.

Use When

The page contains useful information but feels overwhelming.

Give AI

The complete page and conversion goal.

Get

A cognitive-load audit.

Simplify decisions and presentation, not substance.

Phase 4: Optimize Forms, CTAs & Funnels

Small conversion elements matter—but only after you've understood the bigger problem.

Optimize a Lead-Generation Form

Analyze this lead-generation form: FORM: [FORM] OFFER: [OFFER] AUDIENCE: [AUDIENCE] CURRENT CONVERSION RATE: [RATE] Evaluate each field based on: - Business necessity - User effort - Qualification value - Information already available elsewhere - Potential abandonment risk Classify each field as: Essential Useful Optional Remove Then recommend: - Better field order - Clearer labels - Better instructions - Error-message improvements - CTA wording - Trust/risk-reduction elements Do not recommend removing fields that are genuinely required for the business process without considering lead quality.

Use When

A form receives starts but relatively few completions.

Give AI

Form fields, conversion data and sales requirements.

Get

A form optimization plan.

Balance conversion rate with lead quality. A shorter form isn't automatically better if it produces unusable leads.

Improve CTA Strategy Across a Page

Analyze every CTA on this page: [PAGE] Create a CTA inventory. For each CTA identify: - Location - Wording - Intended audience - Desired action - Visitor stage - Primary/secondary role - Relevance to surrounding content Then identify: - Competing CTAs - Weak CTAs - Missing CTAs - Premature CTAs - Repetitive CTAs Recommend a CTA hierarchy: Primary CTA Secondary CTA Low-commitment action Then explain where each should appear.

Use When

A page has multiple CTAs but no clear hierarchy.

Give AI

Page content and conversion objective.

Get

A CTA architecture.

Make the primary action obvious without eliminating useful lower-commitment options.

Analyze Checkout or Signup Friction

Analyze this checkout/signup journey: [STEPS] Performance data: [DATA] Identify potential friction across: - Account creation - Form fields - Payment - Pricing clarity - Shipping/fees - Trust - Error handling - Mobile experience - Required decisions - Unexpected steps - Confirmation - Technical issues For each issue provide: Step → Possible friction → Evidence needed → Severity → Potential solution Separate issues supported by data from hypotheses requiring testing. Then recommend the three highest-priority areas to investigate.

Use When

Users reach checkout or signup but abandon before completing.

Give AI

Journey steps, screenshots/text and funnel data.

Get

A checkout/signup friction map.

Use actual analytics, recordings, support tickets and usability feedback to validate hypotheses.

Find Message Mismatch Between Ad and Landing Page

Compare: AD: [AD] LANDING PAGE: [PAGE] TARGET AUDIENCE: [AUDIENCE] Analyze the consistency between: - Audience - Problem - Promise - Language - Offer - Expectations - CTA - Proof Identify every important mismatch. For each mismatch explain: What the visitor expected → What the page says → Why the difference matters → Recommended correction Do not recommend making the landing page match an ad if the ad itself is misleading.

Use When

Paid traffic generates clicks but the landing page converts poorly.

Give AI

Ad copy/creative description and landing-page content.

Get

A message-match analysis.

Fix the expectation gap rather than simply making the page more aggressive.

Design a Low-Friction Conversion Path

Design the simplest practical conversion path for: BUSINESS: [BUSINESS] AUDIENCE: [AUDIENCE] OFFER: [OFFER] CURRENT FUNNEL: [FUNNEL] The desired outcome is: [OUTCOME] Map the current journey and identify unnecessary steps. Then design: Minimum viable conversion path For each step specify: - Visitor intent - Required information - Action - Friction - Trust requirement - Next step Also identify situations where a longer path may actually be better because the purchase requires education, qualification or trust.

Use When

Your funnel has accumulated too many steps.

Give AI

Current funnel, offer, audience and business requirements.

Get

A simpler conversion journey.

Don't shorten the funnel blindly. Shorter is useful only when it removes unnecessary friction.

Phase 5: Design Better Experiments

CRO isn't about making changes. It's about learning which changes improve the outcome.

Turn a CRO Idea Into a Testable Hypothesis

I want to test this CRO idea: [IDEA] Turn it into a properly structured hypothesis. Use: Because [OBSERVATION], we believe that [CHANGE] will cause [EXPECTED EFFECT] for [AUDIENCE], because [REASON]. Then define: - Primary metric - Secondary metrics - Control - Variant - Target audience - Possible confounding factors - Expected outcome - What result would support the hypothesis - What result would weaken it Do not claim the change will definitely increase conversions. The purpose of the test is to learn.

Use When

You have an optimization idea but haven't turned it into a proper experiment.

Give AI

Your observation and proposed change.

Get

A testable CRO hypothesis.

Write down the hypothesis before seeing the result. Otherwise it's easy to rewrite the story afterward.

Prioritize CRO Tests

Here are my potential CRO experiments: [TEST IDEAS] Score each based on: - Potential impact - Evidence strength - Number of users affected - Business value - Ease of implementation - Testing feasibility - Risk - Learning value Create three categories: Test Now Test Later Don't Test Yet Explain the reasoning. Do not prioritize an experiment solely because it is easy to implement. Also identify experiments that could produce useful learning even if the variant does not win.

Use When

You have more CRO ideas than testing capacity.

Give AI

Your experiment backlog and available data.

Get

A prioritized CRO testing roadmap.

Balance potential impact with evidence and learning value.

Decide What to Test First

I have these possible changes: [CHANGES] Determine which should be tested individually and which could reasonably be combined. Consider: - Interaction between changes - Ability to isolate causality - Traffic requirements - Business impact - Implementation effort - Risk Then create a recommended testing sequence. Explain why the sequence is preferable to testing everything simultaneously.

Use When

You have multiple page changes planned.

Give AI

The proposed changes and available traffic/testing capacity.

Get

A logical experiment sequence.

Avoid creating experiments so broad that you won't know what caused the outcome.

Design a Qualitative CRO Research Plan

I want to understand why visitors aren't converting on: PAGE: [PAGE] Design a qualitative research plan using appropriate methods such as: - User interviews - On-site surveys - Customer interviews - Support conversations - Sales-call analysis - Session recordings - Usability testing - Open-ended feedback For each method provide: - Research question - Who to study - What to ask/look for - Expected insight - Limitations - How the insight could influence CRO decisions Prioritize methods based on the questions I need answered. Do not recommend collecting research simply because "more data is better."

Use When

Analytics tells you what is happening but not why.

Give AI

Your conversion problem, audience and available research methods.

Get

A focused qualitative research plan.

Use qualitative research to generate and investigate hypotheses, not to pretend a handful of interviews statistically represents your entire audience.

Analyze a Failed CRO Experiment

Analyze this CRO experiment: HYPOTHESIS: [HYPOTHESIS] CONTROL: [CONTROL] VARIANT: [VARIANT] RESULTS: [RESULTS] TEST DURATION: [DURATION] AUDIENCE: [AUDIENCE] Determine: - What the experiment actually tells us - What it does not tell us - Whether the result is strong enough to act on - Possible reasons for the outcome - Segment differences worth investigating - Implementation issues - Potential interaction effects - New hypotheses generated by the result Do not invent statistical significance if the required information isn't available. Then recommend: Keep control Adopt variant Run another test Investigate further Explain why.

Use When

A CRO test doesn't produce the expected result.

Give AI

Experiment design and results.

Get

A structured interpretation of the outcome.

A failed experiment isn't necessarily wasted effort. It can eliminate a hypothesis and reveal what to investigate next.

Phase 6: Analyze Results & Prioritize

The final stage is where CRO becomes a continuous operating system rather than a collection of experiments.

Turn Analytics Into CRO Actions

Analyze this conversion data: [DATA] Identify: - High-converting pages - Low-converting pages - Major funnel drop-offs - Device differences - Traffic-source differences - Audience differences - Conversion trends - Unusual changes - Pages worth investigating For every important observation provide: Observation → Possible explanation → Evidence needed → CRO action Separate: What we know What we suspect What we need to investigate Then identify the five highest-priority CRO actions. Do not recommend changes solely because a page has a lower conversion rate than another page. Consider page intent, audience and traffic quality.

Use When

You have analytics data but aren't turning it into optimization decisions.

Give AI

Analytics and conversion data.

Get

A data-driven CRO action list.

Always compare similar pages and similar traffic before drawing conclusions.

Find Conversion Patterns Across High-Performing Pages

Analyze these high-performing pages: [PAGES + DATA] Compare them with these lower-performing pages: [PAGES + DATA] Identify recurring differences in: - Audience - Traffic source - Intent - Offer - Value proposition - Page structure - Proof - CTA - Content depth - Friction - Form complexity - Pricing - Positioning Identify: Patterns worth investigating Patterns worth testing Differences that may be caused by audience or traffic rather than page quality Do not assume correlation proves causation.

Use When

You want to learn from the strongest parts of your existing website.

Give AI

Multiple pages with comparable performance data.

Get

Potential CRO patterns.

Treat patterns as hypotheses until they are validated.

Build a 90-Day CRO Roadmap

Create a 90-day CRO roadmap for: WEBSITE: [WEBSITE] BUSINESS: [BUSINESS] MAIN CONVERSION: [CONVERSION] TRAFFIC: [TRAFFIC] CURRENT CONVERSION RATE: [RATE] KNOWN PROBLEMS: [PROBLEMS] AVAILABLE DATA: [DATA] Prioritize work across: - Analytics - Qualitative research - Messaging - Landing pages - Forms - CTAs - Offers - Trust - Funnel friction - Experimentation - Measurement Divide the roadmap into: Month 1 — Diagnose Month 2 — Test Month 3 — Learn & Scale For each priority provide: - Objective - Action - Hypothesis - Metric - Expected learning - Effort - Dependency Do not fill the roadmap with arbitrary A/B tests. Research and diagnosis should come before testing where evidence is insufficient.

Use When

You want a structured CRO program instead of random optimization work.

Give AI

Website data, conversion data, known issues and resources.

Get

A 90-day CRO roadmap.

Treat the roadmap as a learning system, not a fixed list of website changes.

Find CRO Opportunities From Existing Customer Data

Analyze this customer and conversion information: [DATA] Look for opportunities hidden inside: - Customer segments - Purchase behavior - Lead quality - Product preferences - Repeat purchases - Conversion differences - Support questions - Refund reasons - Sales objections - Customer feedback - Device behavior - Traffic sources For each opportunity provide: Insight → Possible conversion problem/opportunity → Hypothesis → Recommended research → Potential test Prioritize opportunities where there is both meaningful business value and reasonable evidence. Do not assume that every difference in the data requires optimization.

Use When

You have customer data but mostly use it for reporting rather than CRO.

Give AI

Customer data, conversion data, sales information and feedback.

Get

New CRO hypotheses grounded in existing customer information.

Validate important hypotheses before making major website changes.

Find the Next Best CRO Action

Act as my CRO strategist. Here is my current situation: WEBSITE: [WEBSITE] BUSINESS: [BUSINESS] PRIMARY CONVERSION: [CONVERSION] TRAFFIC: [TRAFFIC] CONVERSION DATA: [DATA] TOP LANDING PAGES: [PAGES] FUNNEL: [FUNNEL] CUSTOMER RESEARCH: [RESEARCH] CURRENT TESTS: [TESTS] KNOWN PROBLEMS: [PROBLEMS] BUSINESS GOALS: [GOALS] I have limited time and don't want a giant list of CRO tasks. Identify the 5 highest-priority CRO actions. Evaluate each based on: - Potential business impact - Evidence strength - Number of visitors affected - Conversion relevance - Learning value - Implementation effort - Testing feasibility - Risk For each action provide: 1. What I should do 2. Why it matters 3. What evidence supports it 4. What we still need to learn 5. Recommended hypothesis 6. Suggested test or research method 7. Primary metric 8. First concrete step Do not recommend changing a headline, button color or form field merely because it is easy. Then answer: If I can work on only ONE CRO priority this week, what should it be? Give me one answer and defend it.

Use When

You have analytics, research and a long list of CRO ideas but no clear next step.

Give AI

Your website, analytics, research, funnel, tests and business goals.

Get

A prioritized CRO action plan.

Work on the highest-priority problem before jumping to another optimization idea.

The CRO Workflow in One Simple System

You don't need to use all 30 prompts every week. Use them according to the situation.

Phase 1 — Find Conversion Problems

#1, #2, #3, #4, #5 → Diagnose low-converting pages, find funnel leaks, audit pages, identify friction, and prioritize problems.

Phase 2 — Understand Visitors

#6, #7, #8, #9, #10 → Understand traffic intent, objections, customer language, unanswered questions and the visitor decision journey.

Phase 3 — Improve Pages & Offers

#11, #12, #13, #14, #15 → Clarify the value proposition, improve page structure, translate features into benefits, strengthen proof and reduce cognitive load.

Phase 4 — Optimize Conversion Paths

#16, #17, #18, #19, #20 → Improve forms, CTAs, checkout/signup journeys, message match and overall conversion paths.

Phase 5 — Design Better Experiments

#21, #22, #23, #24, #25 → Build hypotheses, prioritize tests, sequence experiments, conduct qualitative research and learn from failed experiments.

Phase 6 — Analyze & Prioritize

#26, #27, #28, #29, #30 → Turn data into actions, identify patterns, build a CRO roadmap, generate new hypotheses and decide what to do next.

But There's an Even Simpler Way to Remember It

When working on CRO, ask five questions:

1. Where is the problem?

Use: #1, #2, #3, #4, #5

2. Why is it happening?

Use: #6, #7, #8, #9, #10

3. What should I improve?

Use: #11, #12, #13, #14, #15, #16, #17, #18, #19, #20

4. What should I test?

Use: #21, #22, #23, #24, #25

5. What should I do next?

Use: #26, #27, #28, #29, #30

The Real Point of These Prompts

These aren't designed to give you another collection of random CRO tricks. They're designed to become a repeatable conversion optimization system. Don't use AI to make random changes to your website. Use AI to make better conversion decisions.

Download All Prompts (PDF)