CONTROL WORDS / PROGRAMMATIC NAMES
Visible words should
remain part of the name.
Review what the customer sees alongside what the supplied interface exposes.
Introduction
A button displays “Continue”, while the interface exposes a different name to software that operates controls. A customer using speech input may try the words on the screen and find that they do not identify the intended action. A sighted user hearing a screen reader may also encounter language different from the visible text. The procurement question is whether the two representations remain connected.
W3C WAI’s Understanding page for WCAG 2.2 Success Criterion 2.5.3, Label in Name, explains that a component with a visible label containing text or an image of text has a name containing that text. The explanation recommends putting the label text at the start of the name. This is informative web guidance, not proof of speech hardware or accessibility on a vending machine.
This guide proposes evidence for the actual supplied interface. The three WEIMI public listings below establish retail arrangements only. No microphone, speech-control software, accessible-name computation or machine testing was verified. Applicability and acceptance require review of the confirmed customer journey and supported environment.
Quick Answer
Ask the supplier to pair each confirmed visible control label with the accessible name exposed by the delivered interface. Record the actual words, the control and the relevant state. Where the criterion applies, the visually presented label text belongs in the name; a different hidden phrase is not established as equivalent merely because a designer considers it more descriptive.
The source focuses on text labels and images of text. It says the criterion does not apply where a visible text label does not exist. That boundary does not make an unnamed icon acceptable under every other requirement. Keep the label-in-name relationship separate from the broader question of whether a control has an appropriate accessible name.
Do not confuse this with buying a voice assistant. The source concerns information exposed for supported interaction technologies. Confirm whether any speech or text-to-speech arrangement is actually in the project, and review its operation separately. No built-in voice feature is claimed for the products in this article.
Comparison Table
| Public format | Naming question | Supplier evidence |
|---|---|---|
| Single-Door AI Vision Smart Fridge for Packaged Drinks | Which confirmed access/checkout controls have visible text labels? | Label/name pairs in the supplied connected journey |
| WM22 Snacks and Drinks Vending Machine | Do the ordered touchscreen labels remain in computed names? | Actual interface evidence for named controls |
| Two Cabinets, More Choice: Snack & Drink Vending Station | Who maintains names across the quoted menu/payment arrangement? | Confirmed control inventory and provider responsibility |
This is a shortlist based on public product listings, not an independently tested accessibility comparison. No speech support or programmatic naming was verified. Ask the provider to establish the real controls and software before treating the questions as included product features.
Who Should Buy This
Use this brief when accepting a customer interface, changing button copy or integrating a connected purchase page. It is especially useful when commercial teams approve visible wording while a different provider maintains programmatic names. The buyer needs evidence that a text update reaches both representations appropriately.
Include the buyer, interface developer, copy owner and provider of any connected steps. Appropriate evaluators should determine scope and naming behaviour. If the project includes supported speech input or text-to-speech, identify the actual environment used for evaluation. A sales screenshot cannot establish a software-exposed name or a delivered hardware capability.
This review differs from consistent help placement, focus visibility and touch-target geometry. A large visible button can expose an unrelated name. A clearly spoken name can still differ from its on-screen label. Keep the text relationship as its own acceptance item rather than treating a general accessibility statement as complete evidence.
How We Evaluate Smart Vending Machines
We use the W3C Understanding explanation to define the naming questions and public listings to distinguish equipment formats. No name computation, assistive software or physical-machine tests were performed. “Best” refers to a conditional retail shortlist, followed by supplier review of the ordered implementation.
Identify the visible label conservatively
The source recommends considering the nearby text conventionally associated with the control, rather than every relevant sentence on the page. A heading, instruction and group label may provide context without being the control’s own label. Request an annotated control inventory that shows which string the provider evaluates and why.
Read the actual exposed name
Ask the provider for the computed accessible name in the supplied environment, not merely the value of one attribute in a code excerpt. The explanation describes a precedence system in which only preferred naming sources contribute. The full linked computation specifications were not assessed here, so the supplier’s implementation evidence must support the result.
Compare the delivered text and state
Review label/name pairs after relevant copy, language or interface-state changes. A template can show matching words while the deployed version retains an old override. Preserve the software release, connected provider and visible state. This is a proposed procurement review, not a report of observed machine defects.
Key Buying Factors
1. Overrides can replace visible text
The explanation says aria-label and aria-labelledby can take precedence over native naming, even where visible text is associated with a control. It recommends care when visible labels already exist. Ask the provider to demonstrate the resulting name rather than assuming that adding a naming attribute always improves the relationship.
2. Name and description have different jobs
The source explains that aria-describedby contributes to the accessible description rather than the name. Headings and instructions may provide contextual information through a description. A helpful description does not by itself establish that the visible label appears in the name. Keep the name evidence and supplementary context separate in the review record.
3. Starting with the label is a best practice
The criterion says the name contains the visible label text; the note recommends putting it at the beginning. A provider can add necessary context carefully, but should preserve the visible phrase rather than replace it with a more elaborate hidden instruction. Ask for the actual result and appropriate assessment instead of inventing a universal wording formula.
4. Placeholders do not solve general labelling
The explanation says placeholder text is not an appropriate substitute for a label. It also notes that, for this criterion’s visual interpretation, a placeholder can be a candidate when no other nearby string exists. This limited observation does not justify removing persistent labels from a form. Review the real implementation and broader requirements separately.
5. Symbolic characters differ from language
The source distinguishes symbolic characters used as icons from text that expresses human language. It gives text-editor and arrow examples, where the name describes the function rather than merely repeating a symbol. Do not turn that distinction into a claim about every icon on a vending screen; identify the actual control and visible label before assessment.
6. Copy and translation need coordinated ownership
When visible words change, request a corresponding review of the accessible name on the supplied version. The source notes that capitalization and most punctuation differences are not relevant to this criterion, while matching meaningful capitalization is good practice. Those points do not excuse different words or an outdated translation. Assign ownership of both representations in the order.
Best Smart Vending Machines
These three real public-listed products describe retail formats. Their programmatic names, voice controls and accessibility implementation are unverified. Use the equipment arrangement to choose a quotation scope, then request name evidence for the actual interface.
Single-Door AI Vision Smart Fridge for Packaged Drinks
The single-door AI vision fridge listing describes packaged drinks and compatible snacks, camera checkout, five shelf levels and five baskets. A top screen or lightbox and optional cooling are listed. It is packaged-product retail, not juice preparation. Confirm the supplied display, access and payment arrangement.
For confirmed access and checkout controls, ask which visible words label them and what names the responsible provider exposes. Camera recognition does not establish speech input or matching control names. Review the connected journey actually included in the quotation.
WM22 Snacks and Drinks Vending Machine
The WM22 listing describes touchscreen snacks-and-drinks retail with cooling and inventory functions. Spiral, conveyor, direct-push and hanging dispensing arrangements are optional. Confirm the ordered mechanism and interface rather than assuming every listed option is included.
For the ordered touchscreen software, request paired label/name evidence for confirmed menus, buttons and fields. Display size and inventory functions do not establish the accessible naming implementation. Preserve the tested software release and any connected-provider scope.
Two Cabinets, More Choice: Snack & Drink Vending Station
The dual-cabinet listing presents a main display and a secondary section with visible spiral lanes, with two saleable spaces under the menu/payment arrangement. Confirm actual routes and installation details. Independent cooling zones and a specific software architecture were not established.
For the supplied menu/payment arrangement, identify the owner of visible copy and programmatic names across relevant views. Two cabinets do not establish one shared software interface or separate speech systems. Evaluate confirmed controls without inventing a cart or integration architecture.
Feature Comparison
| Review item | Evidence requested | Boundary |
|---|---|---|
| Visible label | Nearby text or image-of-text associated with control | Page-wide instructions are not automatically labels |
| Computed name | Actual name exposed by supplied interface | One attribute excerpt may be incomplete |
| Override | Result after native or ARIA naming precedence | Extra naming code is not automatic proof |
| Description | Context identified separately from name | Description does not replace label-in-name evidence |
| Copy revision | Updated label/name pair on delivered release | Template wording is not deployed evidence |
Use the matrix to request evidence and record unresolved items. It is not a score for the public products. Ask the supplier to identify the interface version and responsible provider before accepting a relationship that may change during a release or translation update.
Cost & ROI Analysis
Illustrative review budget only; every figure and outcome is assumed. Suppose a buyer allocates USD 300 to a control-label inventory, USD 450 to name-evidence review and USD 200 to pilot observation. Initial spending is USD 950. Assume annual follow-up costs USD 250. Equipment, implementation, speech software and formal assessment are excluded.
For sensitivity analysis, assume annual gross contribution associated with additional completed purchases is USD 1,150, USD 650 or USD 200. These are not measured gains or forecasts. After the assumed USD 250 annual review cost, net contribution would be USD 900, USD 400 or negative USD 50. Contribution is different from sales revenue.
| Assumed annual gross contribution | After USD 250 annual review | Simple payback on USD 950 |
|---|---|---|
| USD 1,150 | USD 900 | 1.06 years |
| USD 650 | USD 400 | 2.38 years |
| USD 200 | −USD 50 | No positive payback in this model |
First-year review spending is USD 1,200. Simple payback divides initial spending by positive assumed annual net contribution and omits timing, financing and other costs. A real pilot must distinguish naming effects from traffic, assortment, connected-service reliability and unrelated interface changes. The arithmetic establishes no causal commercial gain or equipment return.
Best Choice by Scenario
Packaged-product access: shortlist the single-door fridge when its retail format fits. Request label/name evidence for confirmed access and checkout controls from the provider responsible for those steps.
Touchscreen snacks and drinks: shortlist WM22 when the ordered mechanism suits the assortment. Ask for paired visible and exposed names on the supplied interface, including relevant copy revisions.
Two saleable sections: shortlist the dual-cabinet station when its layout serves the site. Define the real menu/payment journey and naming ownership rather than inferring software behaviour from cabinet count.
Applications
Quotation attachment
Attach a control inventory showing the visible label, exposed name, relevant state and provider owner. Mark unconfirmed controls as proposed. Specify the actual supported evaluation environment without assuming a built-in microphone or speech assistant. Keep supplementary descriptions distinct from the naming requirement.
Copy approval
When approving customer-facing wording, request a check of the delivered name as well as the screenshot. Preserve the language version and release. A marketing edit can be visually correct while a naming override remains unchanged, so assign coordination responsibility before publication of the interface.
Pilot review
Use appropriate evaluators and agreed tasks on the supplied environment. Where supported speech input is part of the project, record that arrangement explicitly. This article reports no spoken-command tests, participants, completion rates or machine accessibility findings.
FAQ
Does this require a built-in microphone?
The cited criterion concerns the relationship between visible text and a control’s name. It does not establish microphone hardware or a built-in voice assistant. Neither feature was verified on these machines.
Must the accessible name be exactly the label?
The criterion says the name contains the visually presented text. The explanation recommends putting the label text at the start of the name. Additional context needs careful assessment rather than arbitrary replacement of the visible words.
Can descriptive help text replace the name?
The explanation distinguishes the accessible name from the description and notes that aria-describedby is not part of name computation. A description alone does not establish the required label-in-name relationship.
Do punctuation and capitalization have to be identical?
The source says capitalization and most punctuation differences are not relevant to this speech-oriented criterion. It still recommends matching capitalization where meaning matters. This is not a licence to change the label’s words.
Does an icon-only control automatically pass every naming requirement?
No. The explanation says this criterion does not apply where no visible text label exists. Other naming and accessibility requirements remain separate; no icon-only machine control was assessed.
Were these machines tested with speech input?
No. Public product descriptions establish retail formats. No speech software, accessible-name computation or physical-machine accessibility test was performed. Request evidence on the supplied environment.
Final Recommendation
Choose the retail equipment format, then request paired evidence for visible control words and the names the supplied interface exposes. Identify the label conservatively, review actual naming precedence and preserve the delivered version. Coordinate copy and translation ownership without confusing name consistency with a promise of voice hardware or complete accessibility.
Source: W3C WAI, Understanding SC 2.5.3: Label in Name, read on 10 October 2026, page updated 5 April 2026. This is informative web guidance. Linked naming specifications, test rules and the full normative standard were not assessed. Product facts rely on linked public listings and same-date recorded evidence. No machine testing, compliance finding or commercial improvement is claimed.
CTA
Request a WEIMI quotation with configuration-specific control-name evidence. Describe the assortment, customer journey and intended supported input environment. Ask which visible labels are supplied, what names the interface exposes and who maintains both when wording changes.
