Creating a survey for free is easy. Finishing one for free is the part that goes wrong, and it goes wrong in a specific way: collection stops partway through, or the answers are there but cannot be seen, and by then the link has already been sent to several hundred people.
Free plans in this category do not all limit the same thing. Some cap responses, some cap questions, some cap the number of surveys, and one caps what you are allowed to look at rather than what you are allowed to collect. Knowing which limit applies is the difference between a survey that completes and a survey that has to be rerun.
All figures below are from each company's published pricing and support documentation as of September 2026.
The four things a free plan can run out of
Responses. The most common cap and the most damaging, because it arrives when the survey is working. Once it is reached, later answers either stop being accepted or stop being visible.
Questions. Several free tiers allow ten questions per survey. That is enough for a real survey and not enough for a lazy one, which is arguably a feature. It becomes a problem when a survey needs demographic questions plus the actual subject.
The number of surveys. Caps on active surveys mean archiving or deleting old ones to make room, which throws away the thing that would have been fastest to duplicate next time.
People. Almost every free plan allows exactly one. Free tiers are built for an individual, and the moment a colleague needs to see the answers, the free plan has ended even if no other limit has been touched.
Any one of these can be the binding constraint. Work out which one applies to your survey before choosing a tool, because a plan that is generous on responses and capped at ten questions is useless for a long survey with a small audience, and the reverse is equally true.
Where each free plan stops
| Tool | Free plan limit | Users |
|---|---|---|
| Google Forms | No published response allowance | Collaborators you invite |
| Microsoft Forms | 200 responses per form on a free personal account | Collaborators you invite |
| Tally | Unlimited forms and submissions, under fair usage guidelines | 1 |
| Typeform | Response allowance not stated on the pricing page, unlimited questions | 1 |
| Jotform Starter | 5 forms, 100 submissions a month | 1 |
| Zoho Survey | Unlimited surveys, 10 questions each, 100 responses | 1 |
| QuestionPro | 200 responses per survey, 30 question types | 1 |
| SurveySparrow | 75 responses per quarter, 3 active surveys, 10 questions each | 1 |
| SurveyMonkey Basic | A limited number of viewable responses per survey, shown at sign up | 1 |
Three rows need explaining, because they are where most people get caught.
Microsoft Forms is two products. The editor is identical whether the account is a personal Microsoft account or a Microsoft 365 work or school account. The ceilings are not. A work or school account allows up to 5,000,000 responses on one form. A free personal account allows 200, and a paid personal one 1,000. Anyone who builds a survey on whichever account the browser happened to be signed into can hit 200 without ever having seen the number. Check the signed in account before writing the first question.
SurveyMonkey caps visibility rather than collection. The Basic plan lets you view a limited number of responses per survey even if more are collected, and the pricing page does not publish that figure, which appears once you sign up. The answers exist. Reading them is the paid part.
SurveySparrow counts by quarter. Seventy five responses per quarter is not 300 available whenever you need them. A survey that collects 200 answers in one week exhausts the quarter regardless of how quiet the other eleven weeks are.
Google Forms and Tally are the two genuinely open options: a Google Account with no published response allowance, and Tally's unlimited forms and submissions under fair usage guidelines.
Design the survey so it needs fewer responses
The cheapest way to stay inside a free tier is to need less from it, and that is a design decision rather than a tooling one.
Ask fewer questions. Every question costs completion. A survey of eight questions that 400 people finish produces more usable data than a survey of twenty five that 90 people finish, and it fits inside a ten question free tier if two of the eight can be merged.
Pick countable question types. Free text produces sentences that have to be read one at a time. Choice and rating questions produce something that can be counted by a chart the tool draws for you. A survey that is mostly free text is not a survey, it is a reading assignment, and that decision is made at the moment the question type is chosen rather than during analysis.
Do not survey everybody. If the goal is a number rather than a reply to each person, a sample is enough. Two hundred responses from a randomly chosen slice answers the same question as 2,000 from the whole list, and it fits inside caps that the full list would breach.
Close the survey on purpose. A survey left open indefinitely keeps consuming the allowance for answers nobody will read, because the decision it existed to inform was taken weeks ago. Set an end date at the point of creation.
Building it, without paying for anything
Two free routes cover most cases, and both take minutes.
Google Forms
Go to forms.google.com and click Blank form, then name it. The question type selector sits to the right of the question title, and changing it first is worth the habit, because the type determines what the answers look like later. Questions can carry a description, an image or a YouTube video, can be dragged into a new order, and can be marked Required.
Sections are the structural tool, and each one starts on a new page, which is how a long survey stops feeling long. A respondent can be routed to a particular section based on an answer, so nobody is asked to rate a service they just said they had not used.
Themes control the colour, the header image and the fonts for headers, questions and body text. Responder access can be restricted to the organisation or opened to anyone, and email addresses can be collected either as Verified, taken from the respondent's signed in Google Account, or as Responder input, typed by the respondent. Use Preview before publishing and answer the form yourself on a phone.
Microsoft Forms
Go to forms.office.com, sign in, and select New Form. Titles allow up to 90 characters and descriptions up to 1,000. Add new offers Choice, Text, Rating and Date, with Ranking, Likert, File upload and Net Promoter Score behind More question types, alongside Section.
Branching is added from More settings for question, then Add branching, and it has one hard rule: it only moves forward. Pointing question 4 back at question 2 does not loop the respondent around, it skips questions 5 through 7 and drops them at the end of the form. Nothing warns anybody that it happened. End of the form is a valid destination and the right one for a disqualifying answer.
The free settings that protect the answers
Three settings do more for data quality than any paid feature.
Decide whether the survey is anonymous, and say so. Collecting an email address makes a response traceable and therefore answerable, and it also makes respondents more careful about what they write. Leaving it off produces franker answers and forfeits any follow up. Both are legitimate. Making neither choice is not, and a respondent who cannot tell which it is will assume the less comfortable one. Put it in the form description.
Limit to one response per person where the count matters. Microsoft Forms offers this directly, along with Record name. Without it, an enthusiastic respondent can move a small sample on their own.
Use a start and end date rather than a toggle. Microsoft Forms offers Start date and End date alongside the Accept responses switch. Dates close the survey whether or not anybody remembers. Toggles depend on somebody being at a keyboard on the right afternoon.
Getting the answers out, also for free
Exports are not the paywall in most of these tools, which is worth knowing before paying for one.
Google Forms offers Link to Sheets at the top right of the Responses tab, which creates a spreadsheet that keeps updating as answers arrive, and Download responses (.csv) from the More menu for a fixed snapshot. Microsoft Forms exports responses into an Excel workbook.
Understand which of those two you are holding. A linked sheet stays current. A CSV is a photograph of the moment the button was pressed, and teams who download repeatedly end up with several files that disagree, with no way to say afterwards which one a decision came from. If a spreadsheet is going to be the working surface, link it once and work in that one sheet.
Notifications are also free. Google Forms has Get email notifications for new responses on the Responses tab, plus a Form notifications add-on offering more options and customised follow up emails. Microsoft Forms has Get email notification of each response. Switch one on for anything time sensitive, because otherwise a survey is checked only when somebody remembers.
When free stops being free
The free tier ends before the limits do, and the thing that ends it is rarely on the pricing page.
It ends when a second person needs to handle the answers. Every free plan above allows one user, or in the bundled cases allows collaborators who can see and change everything with no way to divide the work between them. A Google Forms collaborator can edit any part of a form including the responses, and a Microsoft Forms collaborate link covers the structure, layout and responses. Neither puts an owner on an individual response.
At that point the tracking moves into a hand maintained column in a spreadsheet, and the cost stops being money. It becomes someone reconciling a sheet every morning, two people replying to the same respondent, and a respondent who never hears back at all. That work is invisible on any invoice and it grows with volume, which is why tools that keep an owner and a status on each response exist at all, and why the situations that need them look so similar from one organisation to the next.
For a survey that ends in a chart, none of this applies and free is genuinely free. For a survey that ends in a conversation, free has a price and it is paid in hours.
What to change first
Check which limit your free plan actually imposes, and check it against the busiest week you expect rather than the average one. If the binding constraint turns out to be people rather than responses, no free tier will fix it, and the choice is between a stricter process around the spreadsheet and a tool where the answer and the reply sit on the same screen.
Q1. Which tool lets you create an online survey for free with no response limit?
Google Forms comes with a Google Account and publishes no per form response allowance. Tally offers unlimited forms and unlimited submissions on its free plan, subject to fair usage guidelines. Among the research focused tools, free tiers are much tighter: 200 responses per survey on QuestionPro, 100 responses on Zoho Survey, and 75 per quarter on SurveySparrow.
Q2. Why did a Microsoft Forms survey stop accepting answers at 200?
Because it was built under a free personal Microsoft account, which caps responses at 200 per form. A paid personal account allows 1,000, and a Microsoft 365 work or school account allows up to 5,000,000. The editor is identical in all three cases, so the only way to know is to check which account is signed in.
Q3. Can you export responses on a free plan?
In the bundled tools, yes. Google Forms offers a live link to a Google Sheet and a static CSV download, and Microsoft Forms exports into an Excel workbook. Export restrictions are more common in the survey specialists, so check that specific point on a free tier before collecting anything you will need in a spreadsheet.
Q4. How many questions should a free survey have?
Fewer than feels natural, and the constraint helps. Several free tiers allow ten questions per survey, which is enough for a focused survey. Eight questions answered by 400 people produces better data than twenty five answered by 90, and each additional question costs completion whether or not the plan charges for it.
Q5. What is the catch with a free plan that allows one user?
That the answers arrive faster than one person can handle them, and nothing in the tool divides them up. Free plans generally allow a single login, and the bundled tools instead let collaborators see and change everything without assigning anything. The work then moves into a spreadsheet column maintained by hand, which costs time rather than money and grows with volume.
Q6. Should a free survey be anonymous?
It depends on whether anybody needs to reply. Anonymous surveys get franker answers to uncomfortable questions and close off any follow up. Identified surveys can be answered and will be written more cautiously. Decide before sending, and state which one it is in the form description, because respondents who cannot tell will assume the less comfortable option and answer accordingly.