Search 37 guides and answers, or browse by topic below.
Three ways to start — blank, from a template, or by describing it to AI Author — and what each one sets up for you.
What each question type is for, and the ones people most often reach for when something else would serve better.
Ready-made surveys you can field as-is, and reusable blocks you can drop into a survey you are already building.
The block added automatically to new surveys, how to remove or extend it, and why the wording is what it is.
What you can still change once data is coming in, what is deliberately locked, and how to make bigger changes safely.
Ask people only the questions that apply to them, based on how they answered earlier.
Drop what someone already told you into the wording of a later question, so the survey reads like it is paying attention.
Use an earlier question's selections as the options, rows or columns of a later one, so people only rate what is relevant to them.
Repeat a question once per option someone picked, or per grid row they rated a particular way, with the item's own wording piped into the question.
Show each person one concept route, keep that route on resume, and cap cells, audiences or answer groups.
Publish and share a link, or send it to a list by email and track who has responded without breaking anonymity.
Queue invites or a reminder to go out at a set time in a chosen time zone, and change your mind up until it fires.
How anonymous surveys work, and the minimum group size that stops filters from identifying individuals.
Automatic flagging for speeders, straight-liners and duplicates, plus attention checks you place yourself.
Buy responses from our panel when you need an audience you do not already have.
End collection without breaking the link: respondents see a closing page, your results stay put, and you can reopen whenever you like.
Charts chosen to match each question type, filters for slicing the data, and where individual responses live.
CSV, Google Sheets, Excel and SPSS, with a choice of layouts for the questions that do not fit one column.
What each AI feature does — drafting a survey, reviewing it, theming open text, summarising results, and probing respondents live.
Copy a survey to fork it, or field it again as the next wave of a series so results trend against the previous run.
Logo, colours, fonts and background — and brand kits, so you set it up once instead of on every survey.
The panel down the right of the editor — completion screen, display mode, progress bar, incentives, randomisation and notifications.
Invite someone to one survey as an editor or a viewer, and what each of those two actually lets them do.
What each plan includes, which limits are real, and what happens when you reach one.
Because in 2026, AI in a survey tool is table stakes. AI Author, AI Reviewer, AI Themes, and AI Exec Summary cost us a few dollars of inference per active user per month — small enough to bake into Pro. Only AI Interviewer is metered on Research, because per-respondent probe costs scale with study size.
Unlimited surveys, unlimited questions, unlimited responses from your own audience, 17 question types, full conditional logic, piping, looping, carry-forward, and quotas. The Free-tier caps: 3 active surveys at a time (upgrade for unlimited), 10 AI Author runs per month, 10 AI Reviewer runs per month with one-click fixes included. Basic fraud + attention checks are on every tier.
You'll see a friendly nudge with the option to upgrade to Pro for unlimited AI usage. Your existing surveys, responses, and previously-saved Reviewer findings stay accessible — only new AI runs are gated.
Probably not. Pro's soft cap is 100 AI Author runs per month — that's 5+ surveys a workday. If you're really crushing it, reach out and we'll work something out for high-volume needs.
Pro and Research include survey collaboration — invite a colleague to any survey at view or edit access. No per-seat math, no separate seat purchases. If your team is large enough that everyone needs their own account, talk to us about Enterprise.
10 completes is the practical minimum. A 10-complete, 2-minute study is ~$30. No subscription required — buy as many or as few completes as you want, whenever you want.
A US consumer panel with particular depth in women 25–40, Gen Z, and active shoppers of CPG categories. We support targeting by age, gender, region, income, household, parent status, occupation, shopping behavior, category usage, brand affinity, and custom screener. For hard-to-reach audiences (specific clinicians, homeowners with solar, etc.), our Enterprise plan offers custom recruitment.
Because it buys you better data. On most panels, respondents get 10–15% of what you pay — which attracts professional survey-takers, farms, and bots. Paying 50% makes the panel a real job for people who care about answering carefully. Our completion rates and fraud scores are the receipts. Every panel-access dollar is itemized at checkout so you can verify where it goes.
CSV and Google Sheets export are on the Free tier. XLSX export and the enhanced export dialog (Standard / Condensed / Expanded / Raw JSON) ship on Pro. SPSS export is on Research. Custom integrations available on Enterprise — talk to us. Your data is always yours, no vendor lock-in.
Open AI Author from the editor and switch to the “Parse existing” tab. Paste the survey as plain text (Word doc, Google Doc, PDF, an export from another tool, even just a numbered list) and we’ll detect the questions, types, options, and scales automatically. You’ll review and edit before saving — nothing publishes without your sign-off.
Share copies the public link (and publishes the survey if it wasn't already) — that's the fastest way to get it in front of respondents, and it works in Slack, an intranet page or a QR code. Send opens a recipient list you paste addresses into, sends each person their own invite, and tracks who has responded so you can remind the people who haven't. Neither one is how you bring in a colleague: the Share button inside the editor does that, and it's a different thing — see “Sharing a survey with a colleague”.
Both can read responses, filter the dashboard, and export. An editor can also change questions and settings, publish and unpublish, send the survey out by email, and run the AI analyst. Neither can invite other people, change roles, or delete the survey — those stay with the owner. Give view access to a stakeholder who wants the findings, and edit access to whoever is building or fielding the survey with you.
Their answers autosave as they go, so a network blip or accidentally-closed tab won't wipe their progress — if they return within 7 days, the survey picks up where they left off. In-progress drafts don't count toward your response total (that stays scoped to submitted responses), but you can see them on the Responses tab if you want to track drop-off. Drafts that sit untouched for more than 7 days are automatically marked "abandoned" so your dashboard stays clean. Nothing is deleted — you can always see abandoned rows to analyze where respondents drop off.
Starting a survey, choosing question types, and getting the structure right before you send it.
Showing the right question to the right person, and pulling earlier answers into later ones.
Getting your survey in front of people, and keeping the data that comes back trustworthy.
Reading the dashboard, slicing results, analysing with AI, and getting your data out.
Making a survey look like yours, and the per-survey toggles worth knowing about.
Plans and limits, sharing work with colleagues, and wiring SurveyGenie into your own systems.
Help is still growing, and we write guides in response to what people actually get stuck on. If it is not here, ask — and it usually shows up here afterwards.