The Form on the Landing Page: Short, Honest, Forgiving

Ask for the least information that lets you honour the promise the page made. If the offer is a quote, you need a way to reply and enough about the job to price it - not the visitor's full biography. If the offer is a download or a discount, you need an email address and nothing else; every extra field is a tax paid in completions. Sequence the form easy-first, one column, on one screen where possible; explain any field a stranger would question; and write error messages that help the visitor finish rather than scolding them for failing your form's private rules.

Take each field and finish the sentence: 'we cannot fulfil the promise without this because...' Name and email usually pass. Phone number passes only if a human genuinely calls back as part of the offer - and stating that ('we will call once, within one working day, to confirm the slot') converts a suspicious field into a service. Company name, job title, budget bands and 'how did you hear about us' almost never pass on a first form; they belong in the follow-up conversation, where trust exists to pay for them. Cutting fields is the most reliably positive form change available, because the mechanism is not persuasion - it is arithmetic: fewer tolls, more travellers.

Order fields so momentum builds: the interesting question first (what do you need?), contact details after the visitor has invested a little, anything effortful last. Multi-step forms can help when the request is genuinely complex - a quote with options - because each step is small and progress is visible; they hurt when they merely slice a short form into theatre. On mobile, one column, generous tap targets, the right keyboard for each field (email, number, phone), and autofill enabled are not polish; they are the difference between a form that gets finished on a bus and one that gets abandoned at the postcode.

Error handling is where form design stops being layout and becomes manners. Validate as the visitor goes, not in a red ambush at submit. Put the message at the field, in words that say what to do: 'Enter the email address the quote should go to - it needs an @ and a domain' beats 'Invalid input'. Never clear the form's other fields on an error; making a visitor retype everything because one field failed is the single most reliable way to lose them at the last step. Baymard's form usability research returns to the same theme: users rarely abandon forms for length alone - they abandon at friction, confusion and punishment.

The form is where the visitor hands over something of theirs, so the page's proof should gather there: one line on what happens next ('Quote by email within one working day, no calls unless you ask'), one line on data ('Your details go to the inbox, not a list - the privacy policy is here'), and any relevant guarantee. A privacy microcopy line under the button outperforms a badge wall higher up the page, because it answers the doubt at the moment the doubt occurs. Keep it true; a false promise at the form is the most expensive sentence on the site.

What happens after the tap decides whether the form felt worth it: an immediate confirmation that names what will happen and when, an email that arrives while the intent is warm, and - if the promise was a resource - the resource itself, delivered, not teased behind another step. Measure form completion as its own rate (starts to submits) so form work is judged separately from page traffic quality; a form fix cannot rescue an audience problem, and an audience problem should never be allowed to convict a good form.