Survey
A survey builder for product questions: it writes questions that do not lead the answer, picks the scale that suits each one, branches past what does not apply to a given person, and hosts the result at a link you can send. Answers come back per person, and you can compare one group against another.
- Asks for
- What you are trying to find out, and what you would do differently depending on the answer
- Who you are asking, and roughly how many of them
- Gives back
- A question set with leading phrasing removed, and a reason for every scale
- Branching, so nobody is asked what does not apply to them
- Themes grouping the questions, and the privacy wording for what you collect
- A public link to send, responses per person, and a way to compare groups
- Plan
- Free
What it gives you
Not: "How much did the new export improve your workflow?" Instead: "When did you last export data from PikoPM?" Never / This week / This month / Longer ago than that If "Never" → skip to section 3. Nobody rates what they have not used. "What did you do with the file after you exported it?" (open text) Behaviour, not satisfaction. The satisfaction question comes after this one.
How it works
Most surveys are written in an afternoon and are broken before they are sent. The questions carry the answer in them — how much did our new design improve your workflow — the scales are five-point when they should be behavioural, and everybody is asked everything, including the questions that make no sense for them. The data then looks clean and means nothing.
So the writing is the product here. You say what you are trying to find out; what comes back is a question set with the leading phrasings removed, a note on why each scale is what it is, and branching so somebody who has never used the feature is not asked to rate it. You edit anything you disagree with — it is your survey, and the point is that you are editing a sound draft rather than fixing a broken one.
It then hosts it. The survey has its own public address, so you can put it in an email, a Slack channel or an in-app message, and the person answering needs no account. Responses are kept per person rather than as totals, which is what lets you ask afterwards how the people who churned answered differently from the people who stayed.
Who it is for
A PM who needs numbers from more people than they can interview, and does not want to spend the week discovering the questions were loaded.
Where it stops
A survey tells you what people say at scale; interviews tell you why. If you do not already know roughly what the answers could be, a survey will give you tidy numbers about the wrong question — run a few conversations first. It also cannot fix a bad sample: a hundred replies from your most engaged users are a hundred replies from your most engaged users.
Questions people ask
Do the people answering need a PikoPM account?
No. The survey gets its own public address and anybody with the link can answer it, which is the entire point — the people you most need to hear from are usually your own customers, not users of this product. You keep an account; they do not need one, and the page they see carries your questions and the privacy wording for what you are collecting, not a sign-up form.
What makes a survey question leading?
Carrying the answer inside the question. "How much did our new design improve your workflow?" has already decided that it improved. So has any question with a positive adjective in it, and so has a scale running from good to excellent. The fix is usually to ask about behaviour instead of opinion — when you last did it, what you did next — and to put the satisfaction question after the factual one rather than before it.
Can I compare how different groups answered?
Yes. Responses are stored per person rather than as running totals, so afterwards you can split them by anything the survey captured or anything you know about who you sent it to, and read one group against another. That is usually where the finding is: the overall average of a product survey is almost always unremarkable, and the difference between the people who stayed and the people who left is not.
Run it on your own work
Survey needs an account, because it writes against what you give it and remembers what you have told it before. It runs on the Free plan. Or see what the plans cost first.