Skip to content

Accessibility

Surveys with accessibility in mind.

We design survey pages for keyboard use, screen readers, and small screens. This page describes the support available and the areas that still need review.

Keyboard operability

Surveys can be completed with a keyboard alone, in every template and both display modes. Every choice, scale point and button is a real, focusable control with a visible focus ring; dropdowns and sliders respond to the arrow keys, and ranking questions have Move up and Move down buttons as well as dragging. A signature can be typed instead of drawn: Type instead sets the respondent's name on the signature pad. In the one-question-at-a-time mode, including the Thread template, Enter moves on, and the keys 1 to 9 pick an answer in multiple-choice, dropdown, rating and single-answer image-choice questions; in the classic page mode, the Next button moves on. Every page starts with a skip-to-content link.

Screen reader semantics

Each question is a labeled group, and each choice is a real control, not a styled div, that reports whether it is selected. Typed answers are named by their question. Errors are announced as they appear and are tied to their question, and the countdown on a timed question is in a live region. When a respondent moves to another page or question, focus moves with them, and on a screen without questions (the welcome, the review, the end) it moves to the heading. On the review, each Edit button names its question, and a ranking says where a row landed when it moves. Progress is a labeled progress bar that a screen reader can read; it is not announced at every step.

No time pressure by default

Nothing times out while someone is answering, and a reload picks up where they left off in the same tab. A question with a minimum time shows a countdown before Next becomes available. Respondents can go back to earlier questions unless the survey's author has turned that off. Two things move on by themselves, both after the survey has ended: when the author has set a redirect, usually back to a research panel, the closing screen continues after about two and a half seconds, with a link to go at once; and on the Free plan, the closing screen moves on to plumeform.com after seven seconds unless the respondent chooses Stay on this page or presses any key.

Motion & visual comfort

With reduced motion switched on in the device's settings, animations and transitions finish at once, scrolling is instant, and the Thread template shows each question without its typing pause. The drawing in the Split stage template's panel and the Fluted template's glass background are still unless the survey's author sets them to drift slowly, and with reduced motion switched on they are always still. Question and answer text is sized in rem, so it follows the browser's text-size setting; a few small labels are fixed in pixels. The default survey page is warm off-white rather than pure white, and a survey can instead use one of 14 themes, five of them dark, or its own colors. Whatever is chosen, text colors are adjusted automatically to at least 4.5:1 contrast against what they sit on, and questions over a photograph, or over a gradient no single text color can be read on, sit on a card.

Mobile and zoom

Survey pages are designed for phones first. Answer choices run the full width of the column (in the Thread template they are message-style replies instead), and answer choices, scale buttons, grid cells, best-worst choices and the buttons that move between pages are at least 40px tall. Typed answer fields use 16px text on touch screens, so iOS does not zoom in on them. Split stage and Gallery, the two templates with a side column, become a single column whenever the survey is narrower than 900 CSS pixels: on a phone, in a narrow embed or when zoomed in.

No trackers on survey pages

Survey pages load no analytics, advertising or tracking scripts, which means less noise for assistive technology and faster loads on low-end devices, and their fonts are served by Plumeform itself. Anything a survey loads from elsewhere is there because its author added it, such as a logo, a picture or an embedded video.

Surveys embedded on other sites

Our embed widget opens popup and side-panel surveys in a labeled dialog. Focus moves to its close button, the rest of the page is set aside while it is open, and focus returns to where it was when the dialog closes. Escape closes the dialog, from its close button or from inside the survey; with a dropdown list open inside the survey, the first Escape closes the list. A site can also set a survey to open by itself after a delay or an event. Inline embeds carry a title for screen readers, and with reduced motion on, the popup appears without sliding in.

Audit status and limitations

We target WCAG 2.1 AA on respondent-facing pages but have not commissioned an independent audit. These are the gaps we know about:

  • Image-choice options can be chosen by keyboard and are read out by their text labels, so give every picture a label that describes it.
  • Multiple-choice answers are toggle buttons that report whether they are picked, rather than a radio group: each answer is its own Tab stop, and the arrow keys do not move between them.
  • A sketch question asks for a drawing, which needs a mouse, a pen or a finger. The builder warns an author who makes one required, and suggests a text question beside it.
  • The survey's author chooses its colors. Text is always adjusted to stay readable, but the color of a button itself is not; the builder warns when the brand color is too close to the background.

Test your survey with the assistive tools your audience uses, and report accessibility barriers to so we can investigate.

Last reviewed September 30, 2026.