Conversational form
Beta tester application form that sorts testers into the right cohort
A beta cohort fails when it fills up with people who applied out of curiosity and never opened the build. This eleven screen application asks what platform someone will really test on, what they use today, and how many hours a week they can honestly give.
The form itself
The first screen of the form. A conversational form shows one question per screen, so the remaining screens are listed in the field table below. Nothing here can be filled in or submitted, so read it as a reference and build your own from the button above.
Every field, and what it collects
The complete field list, in the order respondents meet it.
- Which platforms could you test on?Multiple choiceRequired
- Pick every one you would genuinely open during the six weeks, not every device you own. Cohorts are balanced across platforms, so this answer decides which one you land in.
- Options: macOS desktop app · Windows desktop app · iPhone · iPad · Android phone · Web browser only
- What device and operating system version will you mostly use?Short textRequired
- The version matters more than the model. Half of the bugs worth finding in a beta only appear on one operating system version.
- What do you use for bookkeeping today?DropdownRequired
- Options: A spreadsheet · Paper and a shoebox · A dedicated accounting package · My accountant does all of it · An invoicing tool that is not full bookkeeping · Nothing yet, the business is new
- How often would you open an app like this?Single choiceRequired
- Options: Several times a day · Once a day · A few times a week · Weekly, when I do the books · Monthly, at BAS or tax time
- Roughly how many invoices do you send in a month?Number
- An estimate is fine. This is a range question rather than a precise one, so answer with the number that feels typical.
- Which parts would you actually test?Multiple choiceRequired
- Options: Creating and sending invoices · Chasing overdue payments · Bank feed and reconciliation · Expenses and receipt capture · Reports · Importing from my current tool · Working offline
- How much time can you give the beta each week?Single choiceRequired
- Under an hour is a perfectly good answer and it is what most accepted testers pick. Overcommitting is the main reason people drop out in week three.
- Options: Under an hour · 1 to 2 hours · 2 to 4 hours · More than 4 hours · It depends on the week
- How would you rather send feedback?Single choiceRequired
- Options: In app feedback button · A shared chat channel with the team · Email when something breaks · A short call every fortnight · A written survey at the end only
- What would have to be true for you to switch tools for good?Long textRequired
- Before you applyMultiple choiceRequired
- Options: I agree not to share screenshots, prices or features of the beta publicly until launch · I understand the beta is unfinished software and I will not run my only set of books in it · I am happy to be contacted about the beta by email
- How do we reach you?ProfileRequired
What each screen of this application collects
The first two screens are the cohort sorter. Platforms is a tick list phrased as what you would genuinely open rather than what you own, and the device screen asks for an operating system version because that is where most beta bugs actually live. The next three screens establish the person current behaviour: the tool they use today, how often they would open a replacement, and roughly how many invoices they send, which is the single number that separates a hobby user from a business that would stress the product. Which parts would you actually test turns a vague willingness into a coverage map, so a team missing reconciliation testers can see who to invite. Time per week is asked as a band starting at under an hour, which is deliberately the first option. The switching question is the only written answer. The three ticks and the contact block finish it.
Why under an hour is the first option
Beta recruitment forms usually put the largest commitment first, and applicants read that as the expected answer. The result is a cohort full of people who promised four hours, tested for one week and then went quiet, which leaves the team with no coverage and no way to tell the difference between a quiet tester and a satisfied one. Putting under an hour first, and saying in the helper text that it is what most accepted testers pick, gets an honest distribution and a cohort that can be planned around. The same principle runs through the platform screen and the parts you would test screen: every one of them is worded so that a small honest answer is obviously acceptable.
How to adapt this form
A hardware beta adds a shipping address to the profile block and a screen asking whether the tester can keep a device confidential in a shared household, since a physical unit leaks differently from a build. A game beta swaps the invoices number for hours played a week and replaces the feedback channel question with a platform handle so crash reports can be matched. A regulated product beta, in health or finance, moves the agreement screen to the front and adds a screen recording consent, because recording a session in those contexts needs its own explicit permission rather than a line in a general acknowledgement.
What makes this form work
Platforms is worded as what you would open
Six options, and the helper text says pick what you would genuinely open during the six weeks rather than every device you own. That one sentence turns an aspirational tick list into a usable coverage map, because cohorts are balanced on this answer.
Under an hour is the first time option
The time band screen starts at under an hour and says it is what most accepted testers pick. Applicants answer honestly instead of guessing at the expected answer, and the team gets a cohort whose availability matches what actually happens in week three.
Which parts would you test is a coverage map
Seven options from invoicing to working offline. A team short of reconciliation testers can filter the responses and invite exactly those people, which is the difference between forty testers and forty testers who cover the product.
Questions people ask
Why ask for the operating system version rather than just the device?
Because a beta build breaks on a version, not on a model. Two people with the same phone on different releases will see different bugs, and a team that knows its cohort spans four versions can reproduce a report instead of guessing. It is one short text field and it saves a round trip on nearly every crash report.
Does a tick on this form create an enforceable NDA?
It records that the applicant agreed not to share screenshots, prices or features before launch, which is what most consumer betas actually need. Where a company needs something enforceable it sends a real agreement out for signature after acceptance. Treating a checkbox as a substitute for a contract is a mistake worth avoiding in either direction.
How does an application like this feed a beta list automatically?
Through the webhook add-on. Each submission is posted once as JSON to a URL of your choice, with the response keyed by field label and by field id, so a beta list or a CRM can pick up the platform and the time band without anyone copying a spreadsheet. It fires once per submission and is not retried, so the endpoint needs to accept quickly.
Why not just ask people to email to join a beta?
Because an email gives you a name and nothing else, and the sorting work then happens by hand for every applicant. Eleven structured screens take four minutes and produce a table that can be filtered by platform, by time available and by which parts of the product a person said they would test.
Should a beta application ask for a phone number?
Not for a software beta. The profile block here collects a name and an email only, because everything from the invitation to the build link travels by email anyway. Asking for a phone number costs completions and creates data that has to be protected without ever being used.
What does the switching question actually reveal?
It is the only free text screen, and it produces the product research the rest of the form cannot. Applicants describe the one thing their current setup does badly, which is both a feature signal and a good predictor of who will keep testing, because somebody with a specific complaint has a reason to come back to the build.
Build your own in about a minute
The button below opens the generator with this use case already described. Change the wording to match your own, generate, then edit anything you like.
Make my beta tester application form that sorts testers into the right cohortOther form examples
Brand ambassador application form that reads the content before the reach
Ambassador programmes drown in applications from accounts that bought their audience, and a follower count cannot tell those apart from the creators worth signing. This thirteen screen form asks for a sample post, a sixty second unscripted video and what an audience actually comes for, then asks about reach in bands.
Mentorship program application form for mentors and mentees
Running two separate forms for mentors and mentees means two sets of answers that cannot be compared, which is the hardest part of matching. This thirteen screen form asks both sides the same questions, including what they want out of it, and caps the expectations question at three ticks so a match has something specific to be built on.
Accelerator application form with traction numbers and a deck
An accelerator reads a hundred applications to shortlist twenty, so the form has to produce numbers that sort and answers that are hard to fake. This sixteen screen form asks for stage, last month revenue, active users, capital raised and a founder table, then a 90 second video pitch recorded in the browser.
Want the steps in the builder? Read Create a conversational form with AI, then Automate your form with add-ons. For everything this generator can do, see the form maker.
Sources
Written and checked by the OneCraft team. Last checked .