Conversion Rate Optimization skill 04 of 13

Form Optimisation SKILL.md

The Form Optimisation skill makes Claude, Codex or another AI agent improve sign-up, lead, quote, contact and application forms. It uses field-level analytics, justifies every field, orders and groups questions, chooses input types and autocomplete, accepts flexible formats, adds inline validation and specific error messages, reassures on privacy and judges changes on completion and qualified leads.

form-optimisation/SKILL.md, version 1.0.0, 1,519 words.

By Afeez Adebayo, Marketing, SEO & Growth Expert with 10+ years of experience.

Created 10 October 2026. Updated 10 October 2026. Free to use under CC BY 4.0.

Follow me onX LinkedIn

Download .md Skill folder (.zip)All 13 in one folder (.zip)

Forms are where intent turns into action: sign-ups, demo requests, quotes, applications, checkouts. They are also where many visitors give up. Too many fields, unclear labels, aggressive validation, surprise requirements and slow mobile keyboards each add friction. Small improvements to a form that thousands of people start every month can add more conversions than any new campaign.

This skill audits and improves forms using field analytics, usability principles and research: what to ask, in what order, with which inputs and messages, balancing completion rate with lead quality.

Operating principles

Ask only what you need now. Every field must earn its place; collect the rest later.
Make errors hard and recovery easy. Clear labels, helpful formats, inline validation and kind messages.
Design for mobile first. Correct keyboard types, autofill, large tap targets.
Balance volume and quality. For lead forms, judge changes on qualified leads, not just submissions.
Measure at field level. Know which fields cause hesitation, errors and abandonment.

Common friction points

Friction	Fix
Too many fields	Remove, defer to later steps or progressive profiling
Unclear labels or placeholder-only labels	Visible labels above fields; examples as hint text
Strict formats (phone, postcode, dates)	Accept common formats; normalise on the server
Errors shown only after submit	Inline validation after the field is completed
Vague error messages	Say what is wrong and how to fix it
Password rules revealed late	Show requirements up front; allow show password
Required phone number without reason	Make optional or explain why it's needed
No autofill support	Correct autocomplete attributes
Wrong mobile keyboard	Input types for email, tel, number
CAPTCHA puzzles	Invisible bot protection where possible

When to use, and when not to

Use it for sign-up, lead, demo, quote, contact, application and checkout forms with high abandonment or low quality, and before launching new forms.

Don't use it for full checkout redesigns (checkout-optimisation) or onboarding after sign-up (signup-flow-optimisation).

Inputs to ask for, in one message

Form URL, purpose and what happens after submission.
Form starts, completions and field-level analytics if available.
Which fields are used downstream (by sales, onboarding, billing) and why.
Lead quality data for lead forms (qualification rate by source and field answers).
Recordings or user tests of the form.
Legal or compliance requirements (consent, data protection).

Raw SKILL.md file

.md

This is the exact file agents read, including the credit header and licence notice. Save it as form-optimisation/SKILL.md.

---
name: form-optimisation
description: "Improve sign-up, lead, quote, contact and application forms: field-level analytics, field justification, ordering and multi-step design, labels, input types and autocomplete, flexible formats, inline validation and specific errors, privacy reassurance, mobile and accessibility fixes and changes judged on completion and qualified leads. Use whenever someone asks about form conversion, form abandonment, too many form fields, multi-step forms, form validation errors, lead form quality or improving a sign-up or quote form."
license: CC-BY-4.0
metadata:
  author: Afeez Adebayo
  author_url: https://afeezadebayo.com
  author_linkedin: https://www.linkedin.com/in/afeez-adebayo/
  author_x: https://x.com/za_icon
  source: https://afeezadebayo.com/skills-md/marketing/conversion-rate-optimization/form-optimisation/
  category: marketing
  subcategory: conversion-rate-optimization
  version: "1.0.0"
  created: "2026-10-10"
  updated: "2026-10-10"
---

# Form Optimisation

Forms are where intent turns into action: sign-ups, demo requests, quotes, applications, checkouts. They are also where many visitors give up. Too many fields, unclear labels, aggressive validation, surprise requirements and slow mobile keyboards each add friction. Small improvements to a form that thousands of people start every month can add more conversions than any new campaign.

This skill audits and improves forms using field analytics, usability principles and research: what to ask, in what order, with which inputs and messages, balancing completion rate with lead quality.

## Operating principles
1. **Ask only what you need now.** Every field must earn its place; collect the rest later.
2. **Make errors hard and recovery easy.** Clear labels, helpful formats, inline validation and kind messages.
3. **Design for mobile first.** Correct keyboard types, autofill, large tap targets.
4. **Balance volume and quality.** For lead forms, judge changes on qualified leads, not just submissions.
5. **Measure at field level.** Know which fields cause hesitation, errors and abandonment.

## Common friction points
| Friction | Fix |
|---|---|
| Too many fields | Remove, defer to later steps or progressive profiling |
| Unclear labels or placeholder-only labels | Visible labels above fields; examples as hint text |
| Strict formats (phone, postcode, dates) | Accept common formats; normalise on the server |
| Errors shown only after submit | Inline validation after the field is completed |
| Vague error messages | Say what is wrong and how to fix it |
| Password rules revealed late | Show requirements up front; allow show password |
| Required phone number without reason | Make optional or explain why it's needed |
| No autofill support | Correct autocomplete attributes |
| Wrong mobile keyboard | Input types for email, tel, number |
| CAPTCHA puzzles | Invisible bot protection where possible |

## When to use, and when not to
Use it for sign-up, lead, demo, quote, contact, application and checkout forms with high abandonment or low quality, and before launching new forms.

Don't use it for full checkout redesigns (`checkout-optimisation`) or onboarding after sign-up (`signup-flow-optimisation`).

## Inputs to ask for, in one message
1. Form URL, purpose and what happens after submission.
2. Form starts, completions and field-level analytics if available.
3. Which fields are used downstream (by sales, onboarding, billing) and why.
4. Lead quality data for lead forms (qualification rate by source and field answers).
5. Recordings or user tests of the form.
6. Legal or compliance requirements (consent, data protection).

## Workflow

### Step 1. Measure
Starts versus completions; time to complete; field-level drop-off, corrections and errors.

### Step 2. Justify every field
For each field: who uses it, for what, and could it be collected later? Remove or defer what isn't needed now.

### Step 3. Order and group
Start with easy questions, group related fields, use multi-step forms for longer flows with a progress indicator.

### Step 4. Improve inputs
Labels, hints, input types, autocomplete attributes, sensible defaults, address lookup where helpful.

### Step 5. Improve validation and errors
Inline validation, accepting common formats, specific error messages that keep entered data.

### Step 6. Reduce anxiety
Privacy reassurance near email and phone fields, what happens next, expected response times.

### Step 7. Check mobile and accessibility
Tap targets, keyboard types, focus order, labels for screen readers, error announcements.

### Step 8. Test and measure
Completion rate and, for lead forms, qualified lead rate; run tests where traffic allows.

## Field justification table
```
| Field | Used by | Purpose | Needed now? | Alternative |
| Company size | Sales | Routing and qualification | Yes | Enrich from domain |
| Phone | Sales | Call booking | No | Optional, or ask after booking |
| Job title | Marketing | Personalisation | No | Ask during onboarding |
```

## Multi-step form guidelines
- First step asks the easiest, least sensitive questions.
- Show progress (step 1 of 3).
- Save progress so users can return.
- Put qualifying questions where they help routing, not as a wall at the start.

## Error message examples
| Poor | Better |
|---|---|
| Invalid input | Enter an email address like name@company.com |
| Phone number invalid | Enter your phone number with or without spaces, for example 07700 900123 |
| Password does not meet requirements | Use at least 8 characters, including a number |
| Error | We could not reach the server. Your answers are saved; try again |
Good messages say what happened, how to fix it and keep what the user already typed.

## Autocomplete attributes worth setting
| Field | Attribute |
|---|---|
| Full name | name |
| Email | email |
| Phone | tel |
| Address lines | address-line1, address-line2 |
| Postcode | postal-code |
| Card number (checkout) | cc-number |
Correct attributes let browsers and password managers fill forms in seconds, which matters most on mobile.

## Field-level metrics
| Metric | What it shows |
|---|---|
| Field drop-off | The last field people touched before leaving |
| Time in field | Hesitation or confusion |
| Corrections and refills | Unclear formats or labels |
| Error rate | Validation that is too strict or unclear |
| Blank optional fields | Fields people see no value in |

## Lead quality guardrail
When removing qualifying fields from lead forms, track the share of submissions that sales accepts or that become opportunities. If quality falls, replace the field with enrichment from the email domain or ask the question on the thank-you page or in the first call instead.

## Consent and privacy
Show consent options clearly and unticked where the law requires, link to the privacy notice near personal data fields and explain briefly how contact details will be used. Clear explanations reduce anxiety and abandonment.

## Edge cases and judgement calls
| Situation | What to do |
|---|---|
| Sales wants many qualifying fields | Use enrichment and fewer, better questions; test quality impact |
| Regulated forms (finance, health) | Keep required fields, improve explanations and layout |
| High spam | Invisible bot checks, honeypots and server validation rather than puzzles |
| International users | Flexible name, address and phone formats |
| Very long applications | Multi-step with save and resume |

## Handoffs to related skills
- Overall audit: `cro-audit`
- Research on hesitation: `conversion-research`
- Checkout forms: `checkout-optimisation`
- Sign-up flows: `signup-flow-optimisation`
- Mobile issues: `mobile-conversion`
- Lead quality and scoring: the CRM and Lifecycle `lead-scoring-routing` skill

## Worked example
A home insurance broker's quote form had 14 fields on one page, a strict phone format and errors shown only on submit. Field analytics showed most abandonment at phone and date of birth. The redesign started with postcode and property type, moved personal details to step two with a progress bar, accepted any common phone format, added inline validation and a line explaining why date of birth was needed. Completion rose substantially, and qualified quote rates held steady because the key qualifying fields remained.

## Output format
```
# Form review: <form>, <date>
## Funnel and field-level analytics
## Field justification table
## Recommended structure (steps, order)
## Input, validation and error improvements
## Mobile and accessibility fixes
## Hypotheses and test plan
```

## Common mistakes
- Asking for data nobody uses.
- Placeholder text instead of labels.
- Rejecting valid input formats.
- Errors that clear the whole form.
- Judging lead forms only on submission volume.

## Questions to ask before you change the form
- Who uses each field, and when?
- Where do users hesitate or make errors?
- How is lead quality measured?

## Self-check before you answer
- Is every field justified?
- Are validation and errors helpful and inline?
- Is the form mobile-friendly and accessible?
- Are changes judged on the right quality metric?

## Further reading
- Baymard Institute form usability research: https://baymard.com/research
- W3C Web Accessibility Initiative forms tutorial: https://www.w3.org/WAI/tutorials/forms/
- MDN autocomplete attribute reference: https://developer.mozilla.org/en-US/docs/Web/HTML/Attributes/autocomplete

---
Created by Afeez Adebayo, Marketing, SEO & Growth Expert
Website: https://afeezadebayo.com
LinkedIn: https://www.linkedin.com/in/afeez-adebayo/
X: https://x.com/za_icon
Source: https://afeezadebayo.com/skills-md/marketing/conversion-rate-optimization/form-optimisation/

Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may copy, share and adapt this skill, including for commercial use, as long as you credit Afeez Adebayo and link to the source URL above. Keep this notice in any copy or adapted version.

Workflow

Step 1. Measure

Starts versus completions; time to complete; field-level drop-off, corrections and errors.

Step 2. Justify every field

For each field: who uses it, for what, and could it be collected later? Remove or defer what isn't needed now.

Step 3. Order and group

Start with easy questions, group related fields, use multi-step forms for longer flows with a progress indicator.

Step 4. Improve inputs

Labels, hints, input types, autocomplete attributes, sensible defaults, address lookup where helpful.

Step 5. Improve validation and errors

Inline validation, accepting common formats, specific error messages that keep entered data.

Step 6. Reduce anxiety

Privacy reassurance near email and phone fields, what happens next, expected response times.

Step 7. Check mobile and accessibility

Tap targets, keyboard types, focus order, labels for screen readers, error announcements.

Step 8. Test and measure

Completion rate and, for lead forms, qualified lead rate; run tests where traffic allows.

Field justification table

| Field | Used by | Purpose | Needed now? | Alternative |
| Company size | Sales | Routing and qualification | Yes | Enrich from domain |
| Phone | Sales | Call booking | No | Optional, or ask after booking |
| Job title | Marketing | Personalisation | No | Ask during onboarding |

Multi-step form guidelines

First step asks the easiest, least sensitive questions.
Show progress (step 1 of 3).
Save progress so users can return.
Put qualifying questions where they help routing, not as a wall at the start.

Error message examples

Poor	Better
Invalid input	Enter an email address like name@company.com
Phone number invalid	Enter your phone number with or without spaces, for example 07700 900123
Password does not meet requirements	Use at least 8 characters, including a number
Error	We could not reach the server. Your answers are saved; try again

Good messages say what happened, how to fix it and keep what the user already typed.

Autocomplete attributes worth setting

Field	Attribute
Full name	name
Email	email
Phone	tel
Address lines	address-line1, address-line2
Postcode	postal-code
Card number (checkout)	cc-number

Correct attributes let browsers and password managers fill forms in seconds, which matters most on mobile.

Field-level metrics

Metric	What it shows
Field drop-off	The last field people touched before leaving
Time in field	Hesitation or confusion
Corrections and refills	Unclear formats or labels
Error rate	Validation that is too strict or unclear
Blank optional fields	Fields people see no value in

Lead quality guardrail

When removing qualifying fields from lead forms, track the share of submissions that sales accepts or that become opportunities. If quality falls, replace the field with enrichment from the email domain or ask the question on the thank-you page or in the first call instead.

Show consent options clearly and unticked where the law requires, link to the privacy notice near personal data fields and explain briefly how contact details will be used. Clear explanations reduce anxiety and abandonment.

Edge cases and judgement calls

Situation	What to do
Sales wants many qualifying fields	Use enrichment and fewer, better questions; test quality impact
Regulated forms (finance, health)	Keep required fields, improve explanations and layout
High spam	Invisible bot checks, honeypots and server validation rather than puzzles
International users	Flexible name, address and phone formats
Very long applications	Multi-step with save and resume

Overall audit: cro-audit
Research on hesitation: conversion-research
Checkout forms: checkout-optimisation
Sign-up flows: signup-flow-optimisation
Mobile issues: mobile-conversion
Lead quality and scoring: the CRM and Lifecycle lead-scoring-routing skill

Worked example

A home insurance broker's quote form had 14 fields on one page, a strict phone format and errors shown only on submit. Field analytics showed most abandonment at phone and date of birth. The redesign started with postcode and property type, moved personal details to step two with a progress bar, accepted any common phone format, added inline validation and a line explaining why date of birth was needed. Completion rose substantially, and qualified quote rates held steady because the key qualifying fields remained.

Output format

# Form review: <form>, <date>
## Funnel and field-level analytics
## Field justification table
## Recommended structure (steps, order)
## Input, validation and error improvements
## Mobile and accessibility fixes
## Hypotheses and test plan

Common mistakes

Asking for data nobody uses.
Placeholder text instead of labels.
Rejecting valid input formats.
Errors that clear the whole form.
Judging lead forms only on submission volume.

Questions to ask before you change the form

Who uses each field, and when?
Where do users hesitate or make errors?
How is lead quality measured?

Self-check before you answer

Is every field justified?
Are validation and errors helpful and inline?
Is the form mobile-friendly and accessible?
Are changes judged on the right quality metric?

When to use this skill, and when not to

Use it for forms with high abandonment or low lead quality, and before launching new forms.

Forms are where intent turns into action and where many visitors give up. Too many fields, unclear labels, strict validation and awkward mobile keyboards each add friction. Small improvements to a form thousands of people start each month can add more conversions than a new campaign.

Don’t use it for full checkout redesigns (use Checkout) or onboarding after sign-up (use Sign-Up Flow).

Example prompts

“Our quote form has 14 fields. Improve it.”
“Which form fields cause the most abandonment?”
“Should we use a multi-step form?”
“Rewrite our form error messages.”
“Remove fields without hurting lead quality.”

Example output

The skill returns a fixed structure. A shortened example for an imaginary site:

# Form review: example insurance quote form, 11 October 2026

Drop-off: phone and date of birth fields
Changes: start with postcode and property type; personal details on step 2 with progress bar; any phone format accepted; inline validation; reason given for date of birth
Guardrail: qualified quote rate
Result: completion up substantially; quality held

Lead forms are judged on qualified leads as well as submissions, so removing fields never quietly lowers lead quality.

What the agent needs from you

The skill asks for the form URL and purpose, starts, completions and field analytics, which fields are used downstream and why, lead quality data, recordings or user tests and any legal or consent requirements.

How this skill hands off to related skills

Forms connect to the set: the CRO Audit, hesitation research in Research, payment forms in Checkout, sign-up in Sign-Up Flow, phones in Mobile and lead quality with the CRM and Lifecycle Lead Scoring skill.

How to run Form Optimisation step by step

Install the skill as form-optimisation/SKILL.md.
Measure starts, completions and field drop-off.
Justify every field.
Reorder, group and improve inputs.
Fix validation, errors and reassurance.
Test with a quality guardrail.

Situations this skill handles

Forms differ by purpose. The skill handles these situations:

Situation	What the skill does
Sales wants many fields	Enrichment and quality tests
Regulated forms	Keeps fields, improves explanations
High spam	Invisible bot checks
International users	Flexible names, addresses and phones
Long applications	Save and resume
Low completion on mobile	Keyboards and autofill

What makes this skill different

Most form advice stops at removing fields. This skill measures each field, justifies it with the team that uses it, improves inputs and errors, and protects lead quality with a guardrail metric.

It treats accessibility and mobile behaviour as conversion issues, not afterthoughts.

Practical tips from real projects

Ask only what you need now.
Make errors easy to fix.
Design for mobile first.
Balance volume and quality.

A worked example: a home insurance broker quote form had 14 fields on one page, a strict phone format and errors shown only on submit. Field analytics showed most abandonment at phone and date of birth. The redesign started with postcode and property type, moved personal details to step two with a progress bar, accepted any common phone format, added inline validation and a line explaining why date of birth was needed. Completion rose substantially, and qualified quote rates held steady because the key qualifying fields remained.

Common friction points have known fixes. Too many fields are removed, deferred or collected through progressive profiling. Placeholder-only labels become visible labels with hint text. Strict phone, postcode and date formats are relaxed and normalised on the server. Errors shown only after submit move inline. Vague messages explain what is wrong and how to fix it. Password rules appear up front with a show password option. Required phone numbers become optional or explained. Missing autocomplete attributes are added. Wrong mobile keyboards are fixed with correct input types. CAPTCHA puzzles give way to invisible bot protection.

The field justification table lists each field, who uses it, its purpose, whether it is needed now and an alternative. Company size might stay because sales routes on it, though enrichment from the email domain could replace it; phone might become optional or move after booking; job title might move to onboarding.

Multi-step forms start with the easiest, least sensitive questions, show progress, save answers so users can return and place qualifying questions where they help routing rather than as a wall at the start.

Error messages say what happened, how to fix it and keep what the user typed. Instead of invalid input, ask for an email address like name@company.com; instead of phone number invalid, accept numbers with or without spaces and show an example.

Autocomplete attributes for name, email, phone, address lines, postcode and card number let browsers and password managers fill forms in seconds, which matters most on mobile.

Field-level metrics show where friction lives: field drop-off shows the last field touched before leaving, time in field shows hesitation, corrections show unclear formats, error rates show validation that is too strict and blank optional fields show questions people see no value in.

Consent options are shown clearly and unticked where the law requires, with a link to the privacy notice near personal data fields and a short explanation of how details will be used.

Lead quality is protected with a guardrail. When a qualifying field is removed, the team tracks the share of submissions that sales accepts or that become opportunities. If quality falls, the field is replaced with enrichment from the email domain or asked on the thank-you page or in the first call. Accessibility checks cover focus order, labels for screen readers and error announcements, and they usually improve completion for everyone, not only users of assistive technology. Spam is handled with invisible bot checks, honeypot fields and server validation rather than puzzles that frustrate real people.

Every change is checked on real phones before launch.

Changelog

1.0.0, 10 October 2026: first release.

Use, adapt and share this skill under CC BY 4.0: credit Afeez Adebayo and link to this page.

Previous skillLanding Page and Homepage Optimisation Next skillPricing Page Optimisation

All Conversion Rate Optimization skills