Form Conversion Rate Optimization (Form CRO) workflow skill. Use this skill when the user needs optimize any form that is NOT signup or account registration — including lead capture, contact, demo request, application, survey, quote, and checkout forms — while preserving the upstream workflow, copied support files, and provenance before merging or handing off.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
The command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
File Explorer
7 files
Showing SKILL.md
SKILL.md
Source instructions · Read-only preview
name
form-cro-v3
description
Form Conversion Rate Optimization (Form CRO) workflow skill. Use this skill when the user needs optimize any form that is NOT signup or account registration — including lead capture, contact, demo request, application, survey, quote, and checkout forms — while preserving the upstream workflow, copied support files, and provenance before merging or handing off.
Remove only fields that are unneeded for the stated goal.
When a field must stay, reduce friction through clearer labels, better defaults, examples, masks only when justified, and better keyboard/autofill behavior.
Prefer reversible front-end changes before risky schema or backend changes.
Distinguish quick wins from changes that need experiment design or stakeholder approval.
Redesign for completion and recovery
Improve layout, sequencing, and visual hierarchy.
Put easier fields first when it helps momentum, but do not hide critical disclosures.
Use inline help only where users genuinely hesitate.
Design validation to prevent avoidable errors and help users recover quickly when errors occur.
Make submit states, loading states, and success/failure outcomes explicit.
Protect trust, consent, and data quality
Make purpose statements specific: what the user gets, what happens next, and how quickly.
Keep consent language separate from core action when legally appropriate.
Do not pre-check consent boxes unless policy explicitly permits it.
Avoid deceptive urgency, hidden enrollment, or vague follow-up promises.
Define measurement and release plan
Record baseline metrics and the exact variants or code changes being proposed.
Define what should improve and what must not regress: completion rate, valid submission rate, average time, field error rate, downstream qualification, accessibility defects.
If experimentation is unavailable, still specify how post-release monitoring will detect regressions.
Document rationale and handoff
Summarize the main friction points, changes made, expected impact, and risks.
Preserve provenance and any upstream support files.
Include unresolved constraints, especially backend validation or compliance blockers.
Optimization Areas
Field-level optimization
Focus on necessity, clarity, and ease of entry.
Remove only fields that do not support the business goal or downstream process.
Rewrite labels to be explicit and unambiguous.
Prefer persistent labels over placeholder-only instructions.
Split combined questions only when it reduces confusion.
Use appropriate input types and input modes for email, phone, numeric, postal code, and date-related entry.
Configure autocomplete tokens where supported.
Mark optional fields clearly instead of forcing users to infer.
Layout and flow
Group related inputs logically.
Keep one-column layouts unless there is a strong reason otherwise.
Avoid visually dense blocks that make scanning hard on mobile.
For multi-step forms, ensure each step has a clear purpose and avoid hiding major effort until late in the flow.
Show progress only when it is truthful and stable.
Validation and error recovery
Prevent obvious errors early when possible.
Do not rely only on color to signal errors.
Keep error text specific, adjacent to the field, and actionable.
Preserve user-entered values after validation failure.
Do not trap the user in loops caused by front-end and backend rule mismatch.
Trust and expectation setting
Explain why sensitive data is needed.
State follow-up timing where relevant.
Make pricing, eligibility, or contact expectations visible before submission.
Remove copy that sounds manipulative or evasive.
Submission experience
Make the primary action label explicit, such as Request quote instead of Submit when the action is specific.
Disable duplicate submissions safely.
Show loading states when requests take noticeable time.
Return a confirmation state that explains next steps.
Best Practices
Do:
measure the current form before changing it
optimize for both completion rate and submission quality
use proper labels, field grouping, and standards-based input semantics
make mobile keyboard choice intentional with correct type and inputmode
preserve values and focus users on the first actionable error after validation failure
state required versus optional fields clearly
verify the happy path and the error path on mobile and desktop
test with autofill where name, email, phone, address, or payment-adjacent fields are involved
keep consent and purpose text understandable and specific
document which changes are safe UI refinements versus backend-dependent changes
Do not:
remove fields without understanding why downstream teams use them
replace labels with placeholders as the only instructions
use custom controls that reduce accessibility or keyboard usability without strong justification
block submission with generic messages like Invalid input when field-specific guidance is possible
introduce formatting masks that fight paste, autofill, or international formats unless truly required
treat all abandonment as a field-count problem
hide legal, pricing, eligibility, or consent information until after the user has committed
assume a conversion lift is valid if lead quality or successful fulfillment drops
Troubleshooting
Symptoms: Autofill does not populate expected fields, or users report repeated manual entry.
Solution: Check for correct autocomplete tokens, stable field names, proper label association, and realistic field splitting. Avoid decorative wrappers or custom widgets that break browser heuristics. Review the semantic checklist in references/domain-notes.md.
Symptoms: Mobile users abandon on numeric, phone, or date-like fields.
Solution: Verify type and inputmode match the expected input. Remove unnecessary masks, reduce forced punctuation, and confirm the mobile keyboard supports the intended format. Test copy-and-paste behavior, not just manual typing.
Symptoms: Users hit validation errors repeatedly and cannot tell how to recover.
Solution: Make errors field-specific, adjacent, and actionable. Preserve entered values, move focus to the first unresolved issue after submit, and ensure frontend rules do not conflict with backend validation requirements.
Symptoms: Multi-step forms have sharp drop-off after the first or second step.
Solution: Re-check whether the step split reflects user mental models. Move low-value fields later, surface effort earlier, and make progress indicators truthful. If a later step contains surprise requirements, expose them sooner.
Symptoms: Submission count improves, but lead quality or completion validity worsens.
Solution: Rebalance the optimization target. Restore or redesign qualifying questions instead of removing them outright. Compare valid submissions, downstream qualification, and support burden alongside raw completion rate.
Solution: Add explicit labels, connect help and errors with aria-describedby where needed, group related controls with fieldset and legend, and ensure error states are perceivable without relying only on color or placeholder text.
Examples
Example 1: Lead capture form review
Input
- Form type: demo request
- Fields: name, work email, company, phone, team size, free-text message
- Problem: strong mobile drop-off on phone field and repeated validation failures
- Constraint: sales requires company and work email; phone is optional
Expected optimization direction
- Keep company and work email
- Mark phone clearly optional
- Change phone input to mobile-friendly semantics
- Remove strict formatting mask if backend accepts normalized numbers
- Add inline error text explaining accepted formats only when validation fails
- Update submit label from "Submit" to "Request demo"
- Measure phone-field error rate and mobile completion rate after release
Example 2: Quote request form with trust friction
<!-- Before --><inputplaceholder="Email"><inputplaceholder="Phone"><textareaplaceholder="Tell us what you need"></textarea><button>Submit</button>
<!-- After --><labelfor="email">Work email</label><inputid="email"name="email"type="email"autocomplete="work email"required><labelfor="phone">Phone (optional)</label><inputid="phone"name="phone"type="tel"inputmode="tel"autocomplete="tel"><labelfor="details">Project details</label><textareaid="details"name="details"aria-describedby="details-help"></textarea><pid="details-help">Include timeline, budget range, or scope if known.</p><p>We use this information to prepare an accurate quote. We reply within 1 business day.</p><buttontype="submit">Request quote</button>
Expected outcome:
clearer field purpose
improved autofill and mobile entry
reduced ambiguity around optional phone collection
stronger trust through explicit expectation setting
Example 3: Worked optimization plan
Open examples/worked-example.md for a full before/after review with concrete findings, proposed edits, and measurement notes.
Additional Resources
references/domain-notes.md — open during implementation or review for semantic field setup, autocomplete guidance, validation and accessibility checks, and high-friction pattern notes.
examples/worked-example.md — open when you need a realistic end-to-end example showing how to turn a form audit into concrete changes and measurable outcomes.
Related Skills
No confirmed local related skills were provided in the source context.
