A vacancy for a forklift driver receives two hundred applications, and only after calling forty candidates does the recruiter learn that most of them have no valid licence. A nursing vacancy asks applicants for their full employment history, their hobbies and a motivation letter, and half of them stop halfway through the form. A colleague quietly adds a question about age to a form for a client, and nobody notices until a candidate complains. None of this comes from bad intent: each form was built by whoever happened to publish the vacancy, with whatever questions came to mind. This article explains how to set up screening questions on application forms in your Applicant Tracking System: why forms without agreed questions go wrong, what a screening question is and is not, which questions to ask and which never to ask, how to handle answers that do not meet a requirement, how to record the answers on the candidate, how the approach differs per segment, and how to check with your own data whether the questions actually help.
Screening questions matter in every segment, for different reasons. Recruitment agencies receive large numbers of applications on advertised roles and need to see quickly who meets the client’s hard requirements. Executive search firms rarely use open application forms, but still ask a few questions when a candidate registers interest in a mandate. Staffing desks screen on availability, location, certificates and right to work before anyone is scheduled. Secondment firms check notice period, mobility and specialist qualifications before a consultant is proposed. In-house recruitment teams want every applicant to be asked the same job-related questions, so that hiring decisions can be explained afterwards. In every case the question is the same: do you learn the few facts that decide whether an application can go further, without making the form longer or less fair than it needs to be?
Why forms without agreed questions go wrong
- Each recruiter builds a form per vacancy from memory, so the same role is screened on different questions every time
- Hard requirements from the client or hiring manager are not asked at all, and are only discovered in a phone call days later
- Forms ask for information that is not needed at this stage, which makes them long and makes good candidates give up
- Questions slip in that have nothing to do with the job, about age, health, family situation or origin
- Answers end up in an email or a separate file instead of on the candidate record, so nobody can find them later
The result is a first review that takes longer than it should, candidates who are called for roles they could never take, and a form you would not like to show to a regulator or an unhappy applicant.
What a screening question is and is not
A screening question is a short, job-related question on the application form that tells you whether a candidate meets a requirement agreed in the vacancy brief. It is not an interview, an assessment or a personality test, and it does not decide on its own who gets the job. It helps a recruiter sort applications at the first review; the decision to move a candidate forward or reject them is still made by a person, against the written entry criteria of the next stage. Answers are personal data like any other: they need a purpose, they fall under your retention period, and a candidate who asks what you hold about them receives them in a data access request.
Which questions to ask
Take the questions from the hard requirements in the brief, and keep only those that can be answered in a few seconds:
- Right to work — whether the candidate is allowed to work in the country of the job; the actual document check follows later, as part of the right-to-work check before the start
- Required certificates or licences — only those that are legally or contractually required for the role, such as a driving licence, a forklift certificate or a professional registration
- Availability — earliest start date, hours or shifts, and for staffing pools the availability you will confirm before scheduling
- Location and travel — whether the candidate can reach the work location or is willing to travel or relocate when the role requires it
- Language — the working language and the level the job really requires
- Notice period — when relevant for planning, recorded in the same field you use later for the notice period on the shortlist
Three to six questions per vacancy are usually enough. Keep a shared set of standard questions and let the recruiter pick from it, rather than writing new wording for every vacancy. Link skill questions to your shared skills and tag list, so an answer becomes a searchable tag on the candidate.
Questions you should not ask
Do not ask about age or date of birth, health or disability, pregnancy or family plans, religion, nationality or origin, marital status or anything else that is not needed for the job and could lead to discrimination. Ask whether a candidate has the right to work, not which passport they hold. Do not ask for documents such as copies of identity papers or certificates at the application stage; ask for them when they are actually needed. Let one person, usually recruitment operations or a team lead, approve the standard question set, and allow only that person to add new standard questions, following your user roles and access rights.
What to do with answers that do not meet a requirement
An answer that does not meet a hard requirement is a signal, not an automatic verdict. Let the system flag the application and show the recruiter which answer caused the flag, rather than rejecting the candidate silently. Candidates misread questions, a certificate may be in progress, and a client may accept an exception. The recruiter checks the flagged applications, decides, and records the outcome with one of your standard rejection reasons. The candidate receives a clear, polite message from your shared message templates, stating which requirement was not met. Never reject a candidate on an answer to an optional question.
Record the answers on the candidate
Store every answer on the candidate record, linked to the vacancy it was given for, together with the question wording and version and the date. When the same person applies again later, the recruiter sees earlier answers and only asks again what may have changed, such as availability or notice period. When the same person applies twice under different email addresses, the answers help to recognise a duplicate that should be merged. Record the candidate source on the same application, so you can later see which sources bring applications that pass the screening. When you change a question, keep the old answers with the old wording, so a report never mixes two different questions.
Differences per segment
Agencies usually keep a standard question set per client or per role family, built from the hard requirements the client has agreed, and do not show the client name on the form when the vacancy is confidential. Executive search firms keep questions to a minimum and use them mainly for registration on a mandate; screening happens in the conversation. Staffing desks rely most on screening questions, because availability, location, shifts and certificates decide whether a candidate can be scheduled at all, often within the same day. Secondment firms add questions about mobility and specialist qualifications for the assignment. In-house teams ask every applicant for the same role the same questions, and make sure that internal applicants are screened on the same requirements as external ones.
Let the system help
Much of the work around screening repeats and suits automation: suggesting the standard questions that belong to a role family when a vacancy is created, showing the questions in the language of the job advert, flagging applications with an answer that does not meet a requirement, sorting the first review so flagged and complete applications are easy to find, prefilling answers from an earlier application, and setting a task when flagged applications wait longer than the review deadline. An artificial intelligence helper can suggest clearer wording for a question, point out a question that seems unrelated to the job, or summarise free-text answers for the recruiter. The recruiter reviews every flagged application and decides; artificial intelligence does not replace recruiters and does not reject candidates on its own.
Common mistakes to avoid
- Too many questions — a long form loses good candidates before they finish
- Questions not taken from the brief — a requirement that nobody agreed should not filter anyone out
- Silent automatic rejection — candidates are rejected on a misread question and nobody ever sees it
- Answers outside the record — answers in an inbox or a spreadsheet cannot be searched, reused or deleted on time
- Never reviewed — a question set written for last year’s requirements keeps screening on last year’s requirements
Questions teams ask
- Should a wrong answer reject the candidate automatically? — No; flag it, let a recruiter check it and record the decision with a reason
- How many screening questions should a form have? — Usually three to six, each linked to a hard requirement in the brief
- May we ask for salary expectations on the form? — You can, as an optional question, but many teams prefer to record the salary expectation in the first conversation
- Can we ask open questions? — One short open question can help, but it should never be the reason for a rejection without a person reading it
- Who decides which questions are standard? — One owner of the question set; recruiters suggest new questions, the owner checks them and adds them
Check it with your own data
Count from your own records, per team, per segment and per role family: how many applications came in with and without screening questions, how many were flagged and how many of those were still moved forward after review, how many candidates started the form but did not finish it, how many candidates who passed the screening were later rejected for a requirement the form should have covered, and how long the first review took on average for vacancies with and without screening questions. Look at whether the share of candidates reaching an interview changes after you introduce a question set, and keep those figures next to your time-to-fill. Review flagged applications that are still waiting in your weekly pipeline review. Report the number of applications behind each figure. Do not compare with industry averages; how your own forms perform on your own vacancies is the number that matters.
A practical start
- Collect the hard requirements from your most common vacancy briefs and turn each into one short question
- Agree a shared standard question set per role family and name one owner for it
- Remove every question that is not needed for the job at this stage
- Flag answers that do not meet a requirement instead of rejecting automatically
- Record every answer on the candidate record, linked to the vacancy and the question version
Fewer questions, better first review
A short, agreed set of screening questions tells your recruiters within seconds which applications meet the basic requirements, keeps the form fair and short for candidates, and leaves a record you can explain afterwards. Take the questions from the brief, keep them job-related, flag instead of rejecting automatically, record every answer on the candidate and review the set whenever your requirements change. Explore the product or book a demo — we show how screening questions, flags and recorded answers are set up in BoldATS for your segment.