Pre-screen candidates — automatic screening via knockout criteria

Knockout questions are the most important filters in the recruiting funnel. They filter out applicants who do not meet formal minimum requirements — such as work permit, residence, minimum qualification or availability at the desired date. Anyone who does not fit here will also not fit after a 60-minute interview. Mark these questions clearly as mandatory and communicate honestly what the consequence of a certain answer is.

Formulate knockout questions so that the answer is unambiguous. "Do you have a valid work permit for Germany?" is clear. "Do you speak English?" is not — fluent, professional, school knowledge? Use single-choice answers with defined levels instead of collecting free text and interpreting manually. Pay attention to the AGG: knockout questions must not be discriminatory. Gender, religion, marital status are taboo unless they are exceptionally mandatory for the position (such as a female actress for a female role). When in doubt, consult the legal department.

Beyond hard knockout criteria, a scoring model for the soft factors is worthwhile. Years of experience in a relevant domain, English level, willingness to travel, salary expectation relative to budget — all of these are dimensions that do not mean a hard exit but influence suitability. Award points per dimension and sum them to an overall score. This creates a comparable order of applicants that makes it easier for HR to enter the manual review.

The calculation engine calculates the score automatically on submission. Transparency is important: document the model internally and disclose how points are awarded. Anyone who understands the score can also question it — and that is good, because a model without human correction options reinforces existing biases. Let HR upgrade individual applicants manually at any time if a CV shows special quality that the form does not capture. Review the model every three months against the actual hire outcomes — anyone who had a low score and was ultimately hired successfully points to a gap in the model.

Pre-qualification becomes particularly efficient when the follow-up questions are based on the first answers. Anyone who specifies "backend" for a software engineering role sees different technical questions than someone who chooses "frontend". Anyone who chooses "field sales" for a sales role gets questions about travel readiness, "inside sales" does not. With conditional logic, the form stays short for each applicant and complete for HR.

Think the paths through in personas before you build the form. Which minimum information do you need per typical profile? Which detail questions are nice-to-have? Write the paths on paper or in a simple table and only then translate them into the builder logic. Avoid deep nesting — three to four branches are the practical upper limit, anything more becomes unmaintainable. Test several paths after building and check that no dead ends or infinite loops arise.

Pre-qualification is only efficient if qualified applications land seamlessly in the applicant tracking system. Trigger a webhook into the ATS after successful pre-qualification and hand over all fields in the appropriate format. Anyone who does not yet use an ATS gets far with a shared spreadsheet in which one row per application captures score, answers and status field.

Many ATS offer REST APIs or similar interfaces. Stick to the official documentation and avoid self-built mappings that are difficult to maintain later. If no native API exists, a generic webhook is the most pragmatic way — do not build a native integration that does not exist. In addition to the standard fields, hand over a unique application ID and a timestamp; this makes later audits traceable. Applicants who fail the knockout criteria should receive an immediate, friendly rejection — automated via email trigger, with a short reasoning. This courtesy costs little and strengthens employer branding.