Real run

Form Field Friction Review

Sample 1 · 2026-10-29 · Grok Bot executor subagent (model family not disclosed)

Bot page

Form Field Friction Review: 'Book a demo' on tallyroute.example

1. Field review

Form goal: book a sales demo. Marks come from the field_use you gave, compared with that goal.

Q Field Type Required now Stated use Mark Reason
Q1 Work email email yes Sales, to route and prep the demo Keep Used for the demo itself
Q2 First name text yes Sales, to route and prep the demo Keep Used for the demo itself
Q3 Last name text yes CRM Make optional Used in the CRM, not named as needed to book the demo
Q4 Company text yes Sales, to route and prep the demo Keep Used for the demo itself
Q5 Phone tel yes Sales, only if email fails Make optional Backup contact only, per your note
Q6 Fleet size select yes Sales, to route and prep the demo Keep Used for the demo itself
Q7 Job title text yes Marketing, for segments Move later Used after the first contact, not for booking
Q8 How did you hear about us? text yes Monthly attribution report Make optional Reporting use, not needed to book the demo
Q9 Annual revenue select yes No one could name a use Remove candidate No use stated after asking; a candidate, not a decision

2. Label and autocomplete checks

Q Field Label check Required marker Autocomplete now Note
Q1 Work email Label present Not stated in the input email Purpose already set
Q2 First name Label present Not stated in the input none given-name is a value the saved W3C page lists
Q3 Last name Label present Not stated in the input none family-name is a value the saved W3C page lists
Q4 Company Label present Not stated in the input none Value not confirmed in the sources
Q5 Phone Label present Not stated in the input none tel is a value the saved W3C page lists
Q6 Fleet size Label present Not stated in the input n/a Not data about the user's identity
Q7 Job title Placeholder only: 'Job title' Not stated in the input none Needs a visible label; value not confirmed in the sources
Q8 How did you hear about us? Label present Not stated in the input n/a Not data about the user's identity
Q9 Annual revenue Label present Not stated in the input n/a Not data about the user's identity
  • Labels: W3C's labeling tutorial gives the wording "Labels need to describe the purpose" (https://www.w3.org/WAI/tutorials/forms/labels/). Q7 has only a placeholder, so it has no lasting label.
  • Required markers: W3C's instructions tutorial gives the wording "indicate any required and optional input" (https://www.w3.org/WAI/tutorials/forms/instructions/). Your paste does not say how required fields are marked, so I ask below rather than assume.
  • Autocomplete: the W3C Understanding page for Identify Input Purpose covers "The purpose of each input field" collecting information about the user (https://www.w3.org/WAI/WCAG22/Understanding/identify-input-purpose.html). My own note: Q2, Q3 and Q5 lack a purpose value; Q4 and Q7 values are not confirmed in the saved sources.

3. Counts

Fields per mark, counted by script from the Field review table:

Mark Fields Count
Keep Q1, Q2, Q4, Q6 4
Make optional Q3, Q5, Q8 3
Move later Q7 1
Remove candidate Q9 1
Total 9

Completion data you gave, computed by script (rule: submits divided by starts):

Period Starts Submits Submit rate
September 1,200 312 26.0%

This rests on one period and gives no per field drop off, so it is a baseline, not proof of which field causes friction.

4. Proposed form (held)

Every field stays in its current place. Nothing changes on the live form until you approve each HELD item.

  1. Work email (Q1, Keep)
  2. First name (Q2, Keep)
  3. Last name (Q3, Make optional, HELD)
  4. Company (Q4, Keep)
  5. Phone (Q5, Make optional, HELD)
  6. Fleet size (Q6, Keep)
  7. Job title (Q7, Move later, HELD; add a visible label)
  8. How did you hear about us? (Q8, Make optional, HELD)
  9. Annual revenue (Q9, Remove candidate, HELD)

5. Test plan

  • Hypothesis to test: a form with Q3, Q5 and Q8 optional and Q7 and Q9 off the first step may raise the submit rate against the September baseline in the Counts table. This is a hypothesis, not an expected result.
  • Measure: starts and submits per variant, same period length, plus the share of demos booked that sales can still prep from the kept fields.
  • Before the test: confirm where Job title would be collected later and who owns the attribution report if Q8 answers drop.

6. Questions for you

  • How are required fields marked on the live form now (asterisk, the word required, or nothing)?
  • Where would Job title be collected after the first contact, if you approve moving it?
  • Does anyone outside sales and marketing read Annual revenue, for example finance? If yes, tell me the use and I will re-mark Q9.
  • Can you share per field drop off? It would show which fields to test first.

Needs your decision

  • Approve changing Last name (Q3, Make optional, HELD)
  • Approve changing Phone (Q5, Make optional, HELD)
  • Approve moving Job title (Q7, Move later, HELD)
  • Approve changing How did you hear about us? (Q8, Make optional, HELD)
  • Approve removing Annual revenue (Q9, Remove candidate, HELD)
  • Approve any edit to the live form before it is made.