cro-methodology

Imported from wondelai/skills (cro-methodology/SKILL.md) via skills.sh. Install upstream with npx skills add wondelai/skills --skill cro-methodology. Copyright stays with the author (MIT).

CRO Methodology

Scientific, customer-centric approach to conversion rate optimization based on the CRE Methodology(TM). Extraordinary improvements come from understanding WHY visitors don't convert, not from copying competitors or applying generic tips.

Core Principle

Don't guess -- discover. Every visitor who doesn't convert has a reason. Discover those reasons through research, then systematically eliminate them with evidence and proof. This evidence-based approach consistently outperforms "best practices", intuition, competitor copying, and expert opinion.

Scoring

Goal: 10/10. Score any landing page, funnel, or conversion flow against the seven Quick Diagnostic rows below: award ~1.4 points per row answered "yes" (7 rows = 9.8, capped at 10). Bands: 9-10 = single clear action, research-grounded O/CO table, value prop legible in 5 seconds, proof at every friction point, funnel mapped, path free of UX blockers; 5-6 = guessed objections, generic best-practices copy, proof buried in FAQs; <=3 = competing CTAs, no funnel map, claims with no proof. Report the current score and the specific diagnostic rows failing.

The CRO Frameworks

1. The CRO Process

Core concept: A systematic 9-step process moving from defining success metrics through research and experimentation to scaling wins across the business.

Why it works: Random optimization skips research. The process forces you to understand visitors before changing anything, so every change rests on evidence, not opinion.

Key insights:

• Define success metrics aligned with business KPIs before touching any page
• Map the entire funnel to find "blocked arteries" (high-traffic underperforming paths) and "missing links" (absent funnel stages)
• Research visitors in three dimensions: who they are, what blocks them (UX problems), what stops them (objections)
• Gather market intelligence from competitors, reviews, and other industries
• Prioritize ideas with ICE scoring; design bold experiments, not "meek tweaks"
• Run experiments with statistical rigor (95% confidence minimum, full business cycles), then scale wins across the business

Product applications:

Copy patterns:

• "What's preventing you from [action] today?" (exit survey to discover objections)
• "Here's what [X] customers found..." (counter-objection with social proof)

See funnel-analysis.md when mapping the funnel -- step-by-step mapping, blocked-artery/missing-link diagnosis, industry funnel benchmarks, and impact-based prioritization.

2. Customer Research & Objections

Core concept: Visitors fail to convert for specific, discoverable reasons. Exit surveys, chat logs, support tickets, sales calls, and reviews reveal the "voice of the customer" and their real objections.

Why it works: Teams' guesses about why visitors leave are almost always wrong. Research uncovers objections no one anticipated, and the customer's own language out-persuades any copywriter's invention.

Key insights:

• Primary sources (exit surveys, live chat, tickets, sales calls) give direct visitor language; secondary sources (reviews, social media, competitors) reveal industry-wide objections
• The "Big 5" universal objections: Trust, Price, Fit, Timing, Effort
• Quantitative research (analytics, heatmaps) shows WHERE problems are; qualitative (surveys, interviews) shows WHY
• Non-converter surveys should ask ONE question for maximum response; post-purchase surveys ("What almost stopped you from buying?") reveal the objections that matter most

Product applications:

Copy patterns:

• Use exact customer language in headlines and body copy -- it outperforms polished marketing copy
• "What's the one thing we could change to make you [action]?"
• "How would you describe [product] to a friend?" (reveals positioning in customer terms)

Ethical boundary: Anonymize data, get consent for recordings, and don't survey so aggressively that you degrade the experience.

See RESEARCH.md when planning research -- ready-to-use survey questions per channel, recommended tools, and how to turn raw responses into a ranked objection list.

3. Persuasion Assets

Core concept: Every company sits on overlooked proof -- undisplayed testimonials, unmentioned awards, hidden credentials, buried guarantees. Inventory these "persuasion assets", acquire missing ones, display them.

Why it works: Visitors decide on evidence, not claims. A modest claim with overwhelming proof beats a bold claim with none.

Key insights:

• Audit five categories: Credentials & Authority, Social Proof, Risk Reversal, Data & Specificity, Process & Methodology
• Create a wish list for missing assets and actively acquire them (request testimonials, apply for awards, compile statistics)
• "Proof sandwich" structure: Claim (bold promise), then Proof (evidence), then Reinforcement (secondary proof)
• Proof hierarchy, strongest first: specific results with context > named testimonials with photos > case studies > statistics > logos > generic testimonials
• Place proof at points of friction, not in FAQs; specific numbers beat round ones ("47,832 customers" beats "About 50,000")

Product applications:

Copy patterns:

• "Here's how we did it for [Company X]..." (case study proof)
• "[Specific number] businesses trust us" (not "thousands of customers")
• Lead with benefits, not features: "Never delete another photo" beats "256GB storage"

Ethical boundary: Never fabricate testimonials, inflate statistics, or display fake trust badges -- all proof must be genuine and verifiable.

See PERSUASION.md when auditing or acquiring proof -- the full five-category asset checklist and psychological triggers. See COPYWRITING.md when writing the proof copy itself -- headline formulas, benefit-led phrasing, and proof-element wording.

4. The O/CO Framework

Core concept: The Objection/Counter-Objection table is the core CRE technique: map every visitor objection to a specific, evidence-backed counter-objection.

Why it works: The table forces every counter to be placed where its objection arises in the reading flow, so a concern is answered the instant the visitor feels it -- not pages later, by which point they have already left.

Key insights:

• Research objections from surveys, chat logs, tickets, and sales calls -- don't guess
• Implicit objections (ones visitors won't admit) require "CO Only": counter without stating the objection
• Place counter-objections at the point of friction (credit-card objection near the payment form), not buried in FAQ
• Address primary objections above the fold; repeat the same counter in multiple formats (text, video, testimonial, data)
• Canned support responses are goldmines of tested counter-objections

Product applications:

Copy patterns:

• Bad (states implicit objection): "Worried you're too lazy to learn a language?"
• Good (CO Only): "Let the audio do the work for you."
• "What almost stopped you from buying?" (post-purchase survey to validate the O/CO table)

Ethical boundary: A counter-objection must resolve the concern with real evidence, not dismiss a legitimate worry as unfounded.

See OBJECTIONS.md when building the O/CO table -- per-category counter-objection technique catalogs, CO-Only patterns for implicit objections, and how to mine objections from support logs.

5. Hypothesis Design

Core concept: Every experiment needs a documented hypothesis linking a specific change to an expected outcome for a research-grounded reason, prioritized with ICE scoring (Impact, Confidence, Ease).

Why it works: A hypothesis forces you to articulate WHY a change should work, grounding it in customer research. ICE scoring stops teams wasting traffic on low-impact tweaks.

Key insights:

• Format: "If we [change X], then [metric Y] will improve because [reason based on research]"
• Define primary (decides winner), secondary (monitoring), and guardrail (must not decrease) metrics before testing
• ICE, 1-10 each: Impact (could this double conversion?), Confidence (how strong is the research?), Ease (how easy to implement?); prioritize by the average
• The 10x screen: if a change couldn't 10x results, deprioritize it. Worth testing: complete redesign, new value proposition, fundamentally different offer. Not worth testing: button color, font size, image swap

Worked example: "Customer language from surveys will lift signups because visitors see their own words" scores I:8, C:9, E:10 = 9.0 -- a top-priority test. A button-color swap scores ~I:2, C:2, E:10 = 4.7 and gets skipped despite being trivial to build.

Copy patterns:

• "Based on our research, visitors' #1 objection is [X]. This test addresses it by [Y]."
• Document before: hypothesis, primary metric, sample size, duration. Document after: raw numbers, confidence interval, learnings, next steps

See testing-methodology.md when prioritizing or scoring a backlog -- per-axis ICE scoring rubrics, a worked prioritization table, and the weighted-ICE variant.

6. A/B Testing Methodology

Core concept: Run controlled experiments comparing page versions with proper statistical rigor, so results reflect reality rather than random noise.

Why it works: Without rigor you can't distinguish real improvements from random variation -- peeking, undersized samples, and ignored practical significance all manufacture false winners.

Key insights:

• Calculate required sample size BEFORE starting (baseline rate, minimum detectable effect, 80% power, 95% significance)
• Run at least one full business cycle (1-2 weeks), covering weekdays AND weekends
• Never peek at results and stop early -- it dramatically inflates false positives
• Practical significance matters: a statistically significant 0.1% lift isn't worth implementation complexity
• Use multivariate only with 100k+ monthly visitors on a proven winning page
• Promote winners to the new control; a failed test that teaches you something beats a win you don't understand

Product applications:

Copy patterns:

• "We increased [metric] by [X]% with [Y]% confidence over [Z] weeks"
• "Test showed no significant difference, teaching us that [insight about customers]"
• Document learnings: Test, Hypothesis, Result, Learning, Applicable to

Reporting rule: Decide sample size and duration up front, then report whatever the pre-set test returns -- never stop early on a peeked "winner," rerun a test until it yields the answer you want, or bury an inconclusive result. (This is the one honest-reporting constraint for the whole methodology.)

Common Mistakes

Quick Diagnostic

Audit any landing page or conversion flow:

Further Reading

For the complete CRE Methodology(TM), detailed case studies, and advanced techniques:

• "Making Websites Win: Apply the Customer-Centric Methodology That Has Doubled the Sales of Many Leading Websites" by Dr. Karl Blanks and Ben Jesson

About the Author

Dr. Karl Blanks and Ben Jesson are cofounders of Conversion Rate Experts, the agency whose CRE Methodology has doubled the sales of many leading websites -- clients include Google, Apple, Amazon, Facebook, and Dropbox -- and earned a Queen's Award for Enterprise (Innovation). Blanks holds a PhD and led usability teams at Hewlett-Packard; Jesson's background is direct-response marketing. Their book Making Websites Win distills the methodology into a repeatable, evidence-based process.