Automated rejection
Also called: auto-reject, auto-decline
When auto-rejection is the right call
The clean cases:
- Failed knockout question : candidate answered “no” to a hard requirement (right-to-work, mandatory certification). The rejection is mechanical; no judgment needed.
- Duplicate application : same candidate, same role, recent prior decline. Auto-rejection avoids re-running the process.
- Inactive candidate : 14+ days of no response after a clear next-step request. The rejection is a clean close, not a judgment.
In all three, the auto-rejection is the most respectful thing — silence would be worse.
When it’s the wrong call
Two patterns to avoid:
- AI-score-only rejection : rejecting based solely on a candidate-matching score, with no human review of the application . Even when the score is bad, the rejection email goes out from a human, not from the model.
- Pattern-matched dismissals : auto-rejecting candidates because their CV is “non-standard” — career change, gap year, unusual employer history. Often these are the candidates a manual reviewer would flag for a second look.
The principle: automate the mechanical part, keep the judgment.
What the email should say
A good auto-rejection email is:
- Specific to the role , not “your application.”
- Honest about the stage : “we reviewed your application and decided not to move forward at this time.”
- Localized : in the candidate’s language.
- Free of false hope : no “we’ll keep you in our talent pool ” if you won’t.
Where TalentGleam fits
TalentGleam’s auto-rejection logic fires on explicit triggers only (Knockout-tier screening questions, duplicate application, inactivity), with the email localized and the trigger logged. A recruiter can restore a candidate rejected in error: nothing is deleted. AI-score rejection requires manual confirmation. See the features page .
