Plumeform help center.

Create and publish a survey from your dashboard:

1. 1Create it. Name a survey and choose Create. You can also start from a template, or on Pro paste questions, upload a Qualtrics .qsf file, or use the AI drafter.
2. 2Write the questions. Add questions from the type picker, drag to reorder, split across pages. Everything autosaves; undo with Ctrl+Z.
3. 3Preview. Preview opens the full-size survey in a new tab without storing answers. On larger desktop screens, a live mini preview also follows your edits. Share a reviewer link if someone without an account needs to sign off.
4. 4Publish. Publishing snapshots the survey. Respondents always see the published version, so later edits can never disturb answers already collected.
5. 5Share. You get a link like plumeform.com/s/your-slug. No respondent account is ever required. On Pro, rename it under Settings → Survey link; the old address keeps forwarding, panel IDs included.

Edits to a published survey stay in the draft until you hit Publish changes again. Respondents continue to see the previous published version until then.

Thirty types. Twenty-one are free to publish; file upload (and the voice answer it can record) and the research instruments (price sensitivity, pairwise comparison, Q-sort, card sorting, click maps, highlighting, conjoint and reaction-time tasks) need Pro, and every type can be built and previewed on any plan. Any type can be converted to another from the question editor, and the picker has ready-made versions of common ones: Yes / No, Website, Time, Date range, Country, US state, Faces, Sketch, First-click test, Voice answer and Implicit association (IAT). A page's reading material can also be shown once for a set time before its questions: a five-second test.

Every question can be required, or soft-required, which asks once whether the respondent wants to continue without an answer and lets them proceed.

Rules live in the builder's Logic tab and read as sentences: when some condition holds, then do something.

For repeated groups, select the page and choose Repeat for selected choices. Use {{item}} to include the selected label. Repetition supports up to 20 source choices and 180 expanded questions. Keep file uploads, branching rules, quotas and ending conditions outside the repeated group. Each choice gets its own stable columns in results and exports.

Conditions can test any question (answered, equals, contains, greater/less than) and combine with all or any. Rules can also branch on the assigned experiment arm or the running quiz score, checked at every page turn: “when the score is at least 3, jump to the main task”. That powers comprehension gates and performance-based paths for decision studies.

PipingPRO: add {{q:question-id}} into later question text and it's replaced with that respondent's earlier answer. The question editor has a picker that inserts the token for you.

A survey has two layers. The template is the shape of the page: where the question sits, what carries the progress, how dense the type is. The theme sets the colors, font and corners. Choose the template first, then adjust the theme under Design.

Classic is the free default layout. The other 5 layouts are Pro:

Your brand

Upload a logo (PNG, JPEG or WebP up to 512 KB) or point at one on your own site, and put your name beside it. Either half works on its own: a workspace with no image still brands the page. Choose whether the logo is landscape (a wordmark, sized by its height) or square (a mark or a photo, cropped to a box), how its corners are rounded, and how big it sits.

Upload background or Gallery page images up to 1 MB. Smaller files load faster on mobile. Questions sit on a card over a background photograph to keep the text readable. Each upload has its own URL; uploading another image preserves files used by other pages or surveys. Wait for Saved before leaving the builder.

Colors and text contrast

Choose from nine themes, each a palette of four colors: eight light (Plume, Honey, Cornflower, Iris, Lagoon, Admiral, Azure and Slate) and one dark (Charcoal). Themes set colors, font and shape, and can color the buttons, question card, headings and text from the palette too, while keeping your logo, brand kit and wording. You can customize the accent, background, gradient, font, corners and question cards.

The Design tab has three parts: Layout for the shape of the page (template, display mode, header, progress, reading width), Style for the paint (a look to start from, colors, page background, type, question cards) and Brand for who is asking (logo and name, the words on every page, links, the badge). Under Style, your brand color comes first; Fine-tune color roles sets headings, body and supporting text, buttons, the card, inputs and progress one by one. Text and progress colors adjust if your choice would be hard to read. Choosing another look replaces your colors and offers an undo.

Layout offers Inline, Centered and Floating headers, or none, and keeps the template’s progress position or moves it to the top, bottom or right side. Side progress moves to the bottom on phones and narrow embeds, and you can hide it. Brand adds an optional tagline and a website link on your logo and name; the link opens in a new tab so respondents keep their answers.

The Frosted question-card material lets the background show through, with a refracted edge in supporting browsers. Its tint adjusts for readable text over gradients and photographs. Save these choices as your workspace theme to reuse them. Custom branding, including your logo, name and brand color, requires Pro to publish; Free workspaces can build and preview it, though uploading images and saving a workspace theme need Pro.

Movement

Use Image framing to set desktop and mobile focal points for backgrounds, Gallery images, question media and image choices. Reading layout controls content width, text size and question spacing. Brand settings can include privacy and support links. These links open separately so respondents keep their place.

In Settings → Collecting, Completion screens lets you add images and two actions, or choose a different ending based on an answer. The first matching ending wins; otherwise the score outcome applies, followed by your default thank-you. Panel redirects still take precedence.

Templates use different transitions: Editorial has a gentle entrance, Split moves sideways, and Thread presents questions as messages. Animations respect the respondent’s reduced-motion setting.

Thread goes furthest, because there the movement is the layout. Answering sends your reply, the next question is typed out and arrives as a message, and its answers follow a beat behind it. The last few exchanges stay above it, the page follows the conversation down, and tapping an answer you have already given takes you back to change it. Nothing reloads between questions, or between pages.

We check every template and theme combination for question text, answers, small print, progress and footer contrast against the WCAG AA contrast standard, and check each template at phone and desktop widths for anything that overflows sideways. Pick a dark page and the questions turn light while the answer boxes keep dark text; pick a pale accent and the text on it darkens until it is legible. The builder warns you when an accent is too close to the background to see.

Configure research controls per question or in the builder settings. Advanced controls require Pro; response timing and test mode are available on every plan.

The results page automatically flags unusually fast responses, repeated answer patterns and failed attention checks. Speeder detection starts after five timed responses. Exact duplicate open-text answers are flagged on every survey.

Bot protectionPRO adds two more flags. Automated means the browser reported being driven by software, a script did the clicking, or answers arrived without any typing or pointer activity. Pasted text means most of an open-text answer was pasted rather than typed, which is worth a look but proves nothing: careful people draft elsewhere. While protection is on, the survey page counts key presses, pointer activity, pastes and tab switches. It never records what was typed, and nothing is rejected solely for these flags. Submissions without a valid page-load token, submissions within two seconds of loading, and replayed tokens are blocked.

You can run surveys with Prolific, CloudResearch Connect and MTurk. The panel setup guides include demos you can use to test the flow:

Prolific setup →Connect setup →Other sample providers →

The panel features include:

• Participant IDs (PROLIFIC_PID, STUDY_ID, SESSION_ID, participantId, workerId, or a custom parameter) captured from the study URL onto every response: up to 10 parameters per response, with names up to 40 characters and values up to 200.
• Completion through an on-screen code or a redirect that can include the response ID, score and captured URL parameters.
• Screen-outs routed to their own redirect and code, so disqualified people return marked correctly.
• One response per participant, enforced server-side on the panel ID.
• MTurk preview traffic handled without storing anything.
• Clean and flagged participant IDs copyable in bulk for approve/reject.
• Pro: a bonus list ready for Prolific's bulk bonus box (per completion, per clean completion, or per quiz point), a reject-review list of participants who failed attention checks, and a completion code per experimental condition ({completion_code} in the redirect).

Open a survey from Surveys. The Summary view shows collection performance; Questions shows charts and answer filters; Responses shows individual submissions; Analysis contains cross-tabs, segments and weighting; Quality contains flags and cleaning rules. Your filters carry between views. On a phone, swipe the view bar to reach every view.

On Pro the analysis goes furtherPRO. Cross-tabs get a chi-square test with Cramér's V and ▲▼ markers on the cells that drive the relationship. Compare a segment splits responses by one answer and tests every other question for a difference: chi-square for choices, Welch's t-test for ratings and numbers. Weighting rebalances your sample to population targets on one question; percentages and averages on the page, and a Weight column in CSV and XLSX exports, use it.

Clean datasetPRO: decide once which quality flags make a response unusable (failed attention checks, speeders, straight-liners, automated, duplicate text and, if you choose, pasted text). Matching responses are then left out of the results page, cross-tabs, segments, weighting, CSV and XLSX exports, the weekly export email and the shared results page, so every number you report comes from the same sample. Nothing is deleted: Show everything is one click away, and changing the rule changes the sample everywhere at once. The Results API and Google Sheets sync carry completed, non-test responses without these cleaning rules. On Free, Hide flagged does the same for the page you are on and the exports you download from it.

Public results links show charts and counts on a live, read-only page. They do not include free-text answers or individual responses. On Pro, hiding the survey's Plumeform badge also removes it from the shared page.

Minimum group size (Settings → Fielding → Anonymity, free): no result describes fewer people than the number you choose, 3, 5, 7 or 10. Until enough have answered, and under any filter that narrows the answers below it, the Questions and Analysis tabs show nothing. Cross-tab rows and columns below it are merged into one group or left out, and the totals count only what is shown. Segment comparisons, research measures, AI summaries of written answers, study trends, the Report tab and public results links follow it too; a shared link never receives the hidden numbers. It applies as soon as it is saved. Individual responses and exports are not affected.

NPS and eNPS show promoters, passives and detractors beside the score. Mark an NPS question as Employee NPS to label it eNPS. When the answers span two weeks or more, the card also charts the score by week, or by month past ten weeks, with a dash for any period under the minimum group size.

Research measures and reports

Analysis now includes Kano classifications and Van Westendorp price curves when the question mappings are configured under Settings → Fielding. Unmodified SUS, UMUX-Lite, TIPI, SWLS, WHO-5 and CES batteries from the library receive their standard scores. Results show the complete and excluded counts. These measures use the current published version, active filters and unweighted answers.

In Report, choose charts, add findings and chart notes, set a date range and an answer filter, and optionally compare a baseline period. Save to update the existing public results link. Report filters are saved separately from dashboard filters. Print / save PDF opens your browser print dialog. Review notes before sharing: anything you write there is visible to link holders.

Group surveys into waves of a studyPRO and link participants across them using a panel ID or a custom URL parameter.

• Retention funnel: participants per wave, share of wave 1, and how many also completed the previous wave.
• Wave trends: scale averages and their changes, or answer-option shares, for the same question across waves.
• Linked CSV: one row per participant with each wave's answers in w1_, w2_ columns.
• Multi-wave import: paste or upload several waves at once and get a linked study.

Questions are matched across waves by their wording, so duplicating a wave and editing it still lines up in the trend view.

Add languages in the builder's Languages tabPRO. Each locale is a translation of every string in your survey; the surrounding chrome (buttons, validation, progress) is already localized for English, Spanish, French, German, Italian, Portuguese and Dutch.

Respondents get a language picker and are matched to their browser language automatically; adding ?lang=es to the link forces one. Answers stay a single dataset, with translated wording and the same answer values. AI can draft a locale, and you can review and edit every translation.

The API provides read-only access to surveys and responses in your workspace. Revoking a key stops its access immediately. The API reference includes requests, response fields, pagination, sync guidance and error handling.

Slack and HubSpot

Workspace owners on Pro can configure Connected delivery under Settings → Delivery. Slack uses an incoming webhook for the chosen channel and sends a survey title and dashboard link. HubSpot uses a private app token with contact-write permission; map an email question and optional name, company and phone fields. Contacts match by email. This does not subscribe anyone to marketing.

Optional answer thresholds control which completions are sent. Deliveries run about once a minute, with up to four attempts in total, and show recent status with a manual retry for failures. Test and screened-out responses are excluded. Zapier and Make work through a Catch Hook or the Results API; the native Zapier connector is awaiting developer-account publication and is not yet listed publicly.

Team seatsPRO: from Team the owner invites people by email. Editors build, publish and manage surveys and studies; viewers see results and download exports, and the builder opens read-only for them. Invites last 7 days and only work for the invited address. Billing, API keys and the team itself stay with the owner. If the workspace drops to Free, teammates lose access until it's back on Pro; nothing is deleted. Belong to more than one workspace? Switch under Team → Your workspaces.

Your dashboard shows responses used this month. If you cross the cap, collection continues until midnight UTC, up to 10% over the limit, so fieldwork is not cut off mid-day. Upgrade or manage billing (card, invoices, cancellation) from Plan. Cancellation keeps Pro active until the billing period ends, then switches to Free with all your data intact.

Where data lives, who can reach it, retention and GDPR: security & data practices. Participant accessibility: accessibility statement.

Delete your account: open Account & team → Delete account. Review the workspaces listed and type the requested workspace name exactly (your email if you own no workspace). This permanently removes your account, owned survey data and uploaded files, and cancels associated subscriptions immediately. Other owners' workspaces remain intact. Export anything you need first; deletion cannot be undone and does not issue a refund. If cleanup is interrupted, it retries automatically and you can retry from the account page.