Introduction

In a hypothetical vending-related web form, the browser inserts an email address before the user types. A sales demonstration treats this as proof that the form is accessible. The buyer asks a narrower question: did the form programmatically identify the purpose of the field, or did the browser infer it from a label, remembered data or another heuristic? The same visible result can have different evidence behind it.

This article covers purpose identification for inputs that collect information about the user, where the purpose is in the relevant taxonomy and the technology supports identifying it. A personal email field and an email field for another recipient may accept the same data format but have different roles. Purpose review starts with the subject of the information, not with the keyboard that appears.

The reference is W3C WCAG 2.2 Success Criterion 1.3.5, Identify Input Purpose, at Level AA. This is web-content guidance, not a certification of a physical machine or a local legal conclusion. Phone pages, account forms and operator profiles mentioned here are proposed conditional procurement surfaces, not features verified for the shortlisted machines.

Quick Answer

W3C states that actual autofill is not relevant to evaluating this criterion; programmatically exposed purpose is. Browser heuristics that happen to fill the right value do not satisfy that requirement. Conversely, an eligible field can expose its purpose even when the particular browser supplies no suggestion.

For HTML, H98 describes appropriate autocomplete attributes. A broad type such as email or telephone does not necessarily identify whose information is expected. Visible labels remain important, but they are not a substitute for the requested machine-readable purpose. Keep privacy, session cleanup and saving preferences in a separate review.

Comparison Table

These examples are acceptance fixtures, not observed forms on the listed products. Other requirements can apply to fields outside this narrow purpose-identification scope.

Who Should Buy This

This guide suits buyers whose project includes forms collecting common user information on a supported web surface. Examples might include a customer profile, a quotation contact form or an operator’s own details. Confirm that the quoted system actually contains the form and identify its provider before adding it to the machine acceptance schedule.

It is especially useful where a supplier demonstrates autofill and calls the result complete. Procurement should distinguish authored metadata from browser behaviour. A demonstration using a familiar personal browser can conceal missing or incorrectly assigned purpose values because its heuristics already know the page.

Shared installations need a separate operational policy. A shared cabinet browser and a customer-owned phone should not be assumed to have the same storage, profile or cleanup behaviour. Purpose identification helps software interpret fields; it does not establish that personal data cannot persist or that the system’s account isolation has been tested.

How We Evaluate Smart Vending Machines

The three products below are a procurement shortlist based on WEIMI public listings reviewed on 10 October 2026. We have not independently tested machine forms, autofill behaviour, user profiles or metadata. We compare documented retail formats and the evidence buyers should request if the order includes applicable forms.

Our proposed field ledger records the software surface, field label, data subject, expected purpose, taxonomy match, technology support and actual designation. Mark fields outside scope with a reason. For a scoped HTML field, retain read-only DOM evidence of the attribute alongside a screenshot showing its visible meaning. The record should refer to the delivered build.

W3C H98 asks whether each relevant field has a valid, well-formed autocomplete attribute/value pair and whether the label’s purpose corresponds to the token. A syntactically valid token can still identify the wrong purpose. Review both conditions rather than accepting a validator’s green result without checking the form’s meaning.

We use only fictional values in a proposed controlled test fixture. No customer data submission is needed to inspect purpose metadata. If a later end-to-end form submission is part of acceptance, agree its test destination and data separately. That test serves another purpose and should not be confused with this metadata review.

Key Buying Factors

Begin with the person the information describes

An email input can ask for the user’s email, a colleague’s email or a recipient’s email. A broad email type describes the format rather than that relationship. The source specifically limits this criterion to information about the user. Ask the supplier to explain the field’s meaning in the workflow before assigning a personal-information purpose.

Use the recognised taxonomy rather than custom words

HTML autocomplete accepts defined fixed values. H98 links those taxonomic terms to inputs so software can interpret them consistently. A company-specific token that sounds descriptive is not automatically a valid alternative. Keep custom internal field identifiers separate from the recognised designation and verify the actual attribute on the rendered control.

Check agreement between meaning and token

A field labelled “Given name” should not be accepted merely because it carries some valid name-related token. H98 requires the label’s purpose to correspond to the token. Translation and form reconfiguration can break that agreement if the metadata remains unchanged after the visible label changes. Include those revisions in the regression scope.

Separate a broad input type from a specific purpose

W3C uses telephone, email and password types as examples of broad categories. They may help choose a keyboard or data format, but do not necessarily expose the more specific expected meaning. An input type review and a purpose review can both be useful; neither should be reported as evidence for the other without examining what is actually declared.

Do not promise autofill in every environment

The source makes a clear distinction: purpose must be programmatically determinable, while actual autofill depends on the user agent and settings. A browser with no saved information may offer nothing. A browser using heuristics may offer something even without suitable metadata. Record observed suggestions as behaviour, not as the criterion’s pass condition.

Handle mixed-purpose fields with the stated allowance

W3C discusses an input accepting two purposes, such as username or email. Where the technology does not allow multiple purpose values, the source permits either one value or no designation. Record the actual mixed purpose and technology limit. Do not generalise that allowance into permission to omit purpose metadata from every ordinary personal-information field.

Keep saving and shared-session policy independent

An organisation may want to prevent autofill in some environments, but W3C says input purposes still need to be programmatically determinable. Ask for an implementation explanation and separate behavioural evidence. An autocomplete token does not prove browser storage has been disabled, old personal values are cleared, or the next cabinet user cannot see them. Those operational controls need their own review.

Best Smart Vending Machines

“Best” means a possible format match that merits a configured supplier review. These three real products are not independently tested winners, and their public listings do not verify user-information forms or purpose metadata.

Feature Comparison

The comparison cannot rank these products by autocomplete quality because the necessary form evidence is not in the listings. Request a delivered-build review from the responsible software provider. A design proposal is useful, but it remains a proposal until the actual rendered controls are checked.

Cost & ROI Analysis

This is a hypothetical software review allowance, not a WEIMI price or measured saving. Assume a field inventory and metadata review takes eight hours at $90 per hour, or $720. Assume corrections and translation alignment cost $680, and final verification costs $240. The initial allowance is $1,640. Machine purchase, logistics, tax, stock and payment fees are excluded.

Simple recovery is allowance divided by assumed annual benefit, multiplied by 12. It ignores discounting and subsequent maintenance. These values do not establish faster completion, better conversion or reduced support demand. Replace the assumptions with recorded costs and a supplier quotation; without supported benefits, financial return remains unproven.

Separate purpose inventory, corrections, regression review and shared-device storage work in the quotation. Those are different deliverables. A proposal that changes several tokens may not address browser profile configuration or session cleanup. Define those responsibilities explicitly if the project requires them.

Best Choice by Scenario

For packaged drinks sold through an access-based fridge, shortlist the single-door AI format after confirming the actual route. If it includes eligible user-information fields, request their purpose evidence. If it does not, avoid choosing a machine on an invented autofill requirement.

For touchscreen-led snack and drink retail, shortlist WM22 after confirming its mechanism fits the product range. Any separately quoted customer or operator form should have a named provider and a field ledger. Screen size and form semantics are different procurement facts.

For two selling areas, shortlist the dual-cabinet station after layout and mapping review. If accompanying web content contains forms, distinguish personal information from machine identifiers and from another recipient’s data. Choose on operational fit and a credible evidence plan, rather than an unverified promise that every form automatically fills itself.

Applications

A proposed contact form asks for the user’s given name and email. Its visible labels explain the request, and its eligible controls expose the matching purposes. The reviewer records the values on the rendered controls. Whether the current browser suggests stored data is observed separately and does not determine the metadata result.

A proposed service form asks for a machine serial number and a colleague’s contact email. Those fields should not be treated automatically as personal information about the person filling in the form. Document their subjects and scope. Clear labels, validation and other applicable requirements still need review even where this criterion does not require a listed user-purpose designation.

A proposed operator profile is translated into another language. The purpose taxonomy is language-independent, but the visible meaning must still correspond to the designation. Review the translated label and actual metadata together. A translation approval that considers only wording can overlook a field whose expected purpose changed during reconfiguration.

FAQ

Final Recommendation

Approve a purpose ledger for the confirmed software surfaces. Each eligible field should have a clear subject, expected meaning and valid matching programmatic designation. Keep the screenshot and read-only control evidence with the build identifier. Explain exclusions rather than assigning personal-purpose tokens to every input.

A browser suggestion can be convenient, but it is not the criterion’s proof. Choose the vending format on its confirmed retail fit, then request the software evidence the public listing cannot supply. Maintain labels, metadata and any separate storage policy through configuration and language changes.

CTA

Send WEIMI the intended machine format and a list of any quoted forms on customer or operator web surfaces. Identify whose information each field collects and request a quotation separating purpose metadata review from shared-device storage and cleanup work.