Automating Custom Application Questions Without Losing Your Voice

Custom questions are the extra fields an employer adds to an online application form: "Why do you want to work here?", "What is your notice period?", "Describe a project you led". You can automate the factual ones almost completely. The motivational ones need a draft tailored to the role and your approval before it is sent. This guide explains where the line sits, how to build a workflow around it, and what to watch for when you apply in Germany, Austria or Switzerland.

TL;DR

• Sort custom questions into three groups: facts (notice period, work permit, location), evidence (examples of your work) and motivation (why this employer, why this role). Automate the first group fully, the second with review, and the third only as a draft you approve.
• Keep one verified answer bank. Every automated answer should be traceable to it, so nothing is invented under time pressure.
• Never let automation fabricate experience. Public guidance for UK civil service applicants accepts AI for structuring real examples and rejects it for assessments.
• In Germany, gaps in a form can cost you the application, so check required fields, salary expectations and notice period wording before anything goes out.
• Appliqu offers two modes: "on approval", where you check each application, and Autopilot. For roles you care about, use "on approval".

Table of Contents

• What counts as a custom question
• Why custom questions slow down applications
• What you should automate, and what you should not
• A five-step workflow for custom questions
• Writing answers that survive a human reader
• What automated screening means for your answers
• Will recruiters notice automated answers?
• Our perspective
• Where Appliqu fits
• Privacy and accuracy checks before you automate
• Three safe approaches in Germany
• FAQ
• Sources

What counts as a custom question

Standard fields (name, address, CV upload) look the same on every form. Custom questions are different: the employer or its applicant tracking system (ATS) adds them per vacancy. They appear as free-text boxes, dropdowns, yes/no toggles and numeric fields, and they often sit on the last page of the form, which is where many applicants lose patience.

The table groups the types you will meet most often on German-language and international job boards.

The split matters because the risk is different in each row. A wrong notice period is an annoyance. A wrong claim about experience can end an interview or, later, an employment relationship.

Why custom questions slow down applications

Someone applying to ten roles a week rarely struggles with the CV. The time goes on the form: logging in to yet another employer portal, retyping the same notice period, rewriting a motivation paragraph at 22:00. The questions are repetitive in content but different in wording, so simple copy and paste produces answers that do not fit the question.

There is also a quality trap. If you reuse one generic paragraph for "Why us?", a recruiter sees it immediately. If you let a generic text generator write it from the question alone, you get fluent but empty text. The answer has to draw on two things at once: what the employer actually wrote in the posting, and what you have actually done. That is the reasoning behind how we built Appliqu: it reads the full job description rather than matching keywords, and the documents stay editable until you approve them.

What you should automate, and what you should not

Safe to automate

• Profile facts. Notice period, start date, location, languages, driving licence, willingness to relocate.
• Consistent eligibility answers. Your work authorisation status does not change between applications, so it should not be retyped each time.
• Standard consents. Data-processing consent checkboxes, as long as you have read what they cover once.

Automate with a review step

• Evidence questions. Ask for a draft built from examples you have already written down. A reliable structure is STAR (Situation, Task, Action, Result), which the Civil Service Careers guidance lists as an acceptable use of AI, alongside summarising public information about an organisation.
• Motivation questions. The draft should quote or reflect specifics of the posting and the employer. You then change anything that does not sound like you.
• Salary expectations. Check the range against the role, the region and the pay information in the posting before you accept a number.

Do not automate

• Assessments. The same Civil Service guidance names completing personality or ability assessments, such as situational judgement tests, as unacceptable use. Employers elsewhere often take the same view, and you can see why: the test measures you, not your tooling.
• Anything involving confidential data from a current employer.
• Knockout questions you cannot truthfully answer yes to. Automation should not make a borderline claim for you.

A five-step workflow for custom questions

1. Build your answer bank once. Write down 6 to 10 real examples from your work (a project, a conflict, a result you delivered, a failure and what you changed). Add your fixed facts: notice period, salary range, languages, work authorisation. Keep each example in STAR form, in two or three sentences.
2. Set rules for the factual fields. Decide in advance how you answer on-site versus hybrid, travel, relocation and shift work. Rules prevent inconsistent answers across forms, which matters if a recruiter sees two of your applications.
3. Score fit before you send volume. Answering custom questions well for a role you are a poor match for is wasted effort. Our article on job search automation and fit scoring in DACH explains how to filter first.
4. Draft, then review. Let the draft use your answer bank and the posting. Read every free-text answer aloud. Remove anything you could not defend in an interview.
5. Check the form before submitting. Confirm required fields, dropdown values (a salary band, for instance), and that the CV you attached matches the answers you gave.

For the wider process, including how the stages fit together, see our five-stage workflow for automating job applications without losing tailoring.

Writing answers that survive a human reader

Motivation questions

A German recruiter reading "Why do you want to work for us?" expects a specific, factual answer: what attracted you to the employer's product, sector or position, and what you bring. Avoid superlatives and enthusiasm without substance. Two or three sentences usually suffice unless the form gives a character limit; respect it exactly, because some forms cut text off silently.

A workable structure:

1. One concrete detail from the posting or the employer's public information.
2. One piece of evidence from your own work that relates to it.
3. One sentence on what you want to do in the role.

Evidence questions

Use the STAR pattern and keep the result checkable. If you cannot state a figure honestly, describe the outcome in words instead of estimating. Do not borrow an example from an article or a template: interviewers ask follow-up questions, and the example has to be yours.

Salary and notice period

In Germany, forms commonly ask for annual gross salary (Jahresbruttogehalt). State the figure as annual gross and in euros, and keep the format the form asks for. If the field only accepts a single number, enter the lower end of a range you would genuinely accept, not a placeholder such as 0 or 1, which can look careless in the ATS. For notice period, state what your contract says, for example "three months to the end of a quarter", and do not guess.

Languages and eligibility

Use the levels of the Common European Framework (A1 to C2) only if you can back them up with a certificate or demonstrated use. Applicants from outside the EU should state work authorisation precisely, including whether sponsorship is required. A vague answer in a dropdown can be read as "yes, no sponsorship needed".

What automated screening means for your answers

Employers also automate. Some use algorithms to rank or filter applications, and custom questions are a convenient input because they are structured. The UK Information Commissioner's Office explains in its guidance for jobseekers that when an application score falls below a threshold, it may be rejected without a human recruiter seeing it, and that candidates can be entitled to be told about solely automated decisions and to ask for human review. That is a UK source and UK law. Germany, Austria and Switzerland have their own data protection rules (the GDPR applies in the first two, Switzerland has its own act), so check the employer's privacy notice for what applies to you.

The practical consequences for your answers:

• Answer knockout questions exactly as asked. A filter cannot interpret nuance. If you have 4.5 years of experience and the question asks for five, a truthful "no" with an explanation in the cover letter beats a false "yes".
• Use the vocabulary of the posting where it is true. The UK government's Responsible AI in Recruitment guide warns about tools that over-weight narrow terminology and criteria like postcode or employment recency, which can disadvantage candidates with career breaks. You cannot control that, but you can make sure your truthful experience is stated in the terms the employer uses.
• Add context where the form allows it. A free-text box is often the only place to explain a gap or a career change.

On the employer side, CIPD guidance on inclusive recruitment recommends comparing an automated sift with a manual screening for one round to see whether it disadvantages particular groups. The CIPD also advises meaningful human oversight: AI may recommend, people decide. The same principle applies to your side of the process, and it is why we keep a review step in the design.

If you want to see how the regulatory picture in the EU affects applicant-side automation, read our analysis of whether the EU AI Act affects auto-apply tools.

Will recruiters notice automated answers?

They notice generic answers, whether a person or a machine wrote them. The signals are consistent: the same opening sentence across applications, praise that could apply to any company, claims with no example, and answers that ignore the exact question. Our article on whether recruiters can tell that AI was used on your CV goes into the details. The fix is not to hide automation. It is to make sure every answer contains something specific to the role and something verifiably yours.

Our perspective

We built Appliqu around one belief: volume without fit is wasted effort, and fit without review is a risk. That is why Appliqu reads the entire job description and not only keywords, tailors your CV and cover letter to each role, and keeps the content editable and human-approved before submission.

Custom questions are where this belief is tested hardest. They are short, they feel minor, and they are the part of an application where a mistake is most visible. We therefore recommend working in "on approval" mode for any role you actually want, and keeping Autopilot for roles where the stakes are lower and your rules are clear. Autopilot is a convenience for repetitive work, not a replacement for judgement about which employers deserve your best answer.

Our advice on answers is simple. Keep your facts in one place. Write your examples yourself, once, carefully. Let automation do the repetitive assembly, and read what leaves under your name. If you are unsure about a claim, remove it.

Alex

Where Appliqu fits

Appliqu searches over 100 job boards around the clock, matches roles to your profile, tailors your CV and cover letter and tracks each application live. You can start on the permanent free Explorer plan, where Appliqu writes 5 applications for you. Sending starts with Core.

You can compare the plans on the Appliqu website. If you freelance, our guide to automating freelance proposals while keeping ATS fit and human review applies the same approach to proposal questions.

Privacy and accuracy checks before you automate

Automated applications move your personal data between more parties, so a few checks are worth making once:

• Read every draft. Our article on job automation privacy explains why reviewing each draft before it leaves protects both your data and your reputation.
• Remove sensitive details you do not want in an employer's database, such as health information, unless the role requires it.
• Keep a record of what you submitted. You will need it for interview preparation, and a live application tracker saves you reconstructing it.
• Stay truthful. If an answer is wrong, correct it before the interview rather than after.

The UK ICO's guidance for organisations buying AI recruitment tools also gives a useful picture of what employers are being asked to disclose about their own automated scoring, which you can use as a checklist when reading a privacy notice.

Three safe approaches in Germany

German applicants tend to expect complete, correct documents and precise answers. Automation fits that expectation when it reduces typing errors and inconsistency, and clashes with it when it produces filler. Our overview of three safe job search automation approaches for German candidates covers the options in more depth. In short:

1. Automate discovery, write by hand. Use automation to find and score roles; answer custom questions yourself.
2. Automate drafts, approve manually. The most balanced setup, and the one we recommend for most applicants.
3. Automate sending for well-defined roles. Only where your rules are clear, the questions are factual and you have reviewed the setup.

FAQ

Can I automate answers to custom application questions?

Yes, for factual questions such as notice period, location and work authorisation, which come from your profile and do not change. For motivation and evidence questions, use automation for a draft and approve it yourself. Knockout questions and assessments should be answered by you personally.

Should I use the same answer for "Why do you want to work here?" every time?

No. Recruiters recognise reused text quickly. Keep a fixed structure (a detail from the posting, evidence from your work, what you want to do), but change the content for each employer.

Is it acceptable to use AI to write application answers?

It depends on the employer and the task. The Civil Service Careers guidance in the UK accepts AI for structuring your examples and summarising public information, but not for completing assessments. Read the application instructions, and if the employer asks you not to use AI, follow that.

What should I enter for salary expectations?

Enter an annual gross figure in euros that you would genuinely accept, based on the role, the region and any pay information in the posting. Avoid placeholder numbers. If the form allows a range, state it and keep it consistent with what you say in the interview.

Does Appliqu send applications without my approval?

That depends on the mode you choose. In "on approval" mode you check each application before it goes out. Autopilot sends without that step, and sending itself starts with the Core plan.

Sources

• ICO: what jobseekers need to know about automated recruitment decisions
• GOV.UK: Responsible AI in Recruitment
• CIPD: Inclusive recruitment, guide for people professionals
• CIPD: Navigating AI standards
• Civil Service Careers: Artificial intelligence and recruitment
• ICO: Thinking of using AI to assist recruitment?

Recommended

• 5 Stage Workflow to Automate Job Applications Without Losing Tailoring
• Automate Freelance Proposals, Keep ATS Fit and Human Review
• Scale Bulk Job Applications in Germany With Supervised Automation
• Job Search Automation: Score Fit Before Volume in DACH