For employers reading the Screening tab on a candidate profile.
Every screening question on a candidate's profile comes with an AI evaluation block. For each question you'll see Why We Ask This (what the question is testing), What We Look For (the specific traits being evaluated), Evaluation Criteria (what a strong vs. weak response looks like), and a per-question match percentage. The candidate's score on just that question. Use it to spot who performed well on the criteria that matter most, not just who had the highest overall match.
Where to find the per-question evaluation
To see per-question evaluation, open a candidate's profile and switch to the Screening tab.
Open the candidate from your position's candidate list.
Click the Screening tab.
Expand any question to see its evaluation block and the candidate's response.
Every question on the position gets its own evaluation block, so the answers are consistent across candidates. You can compare question-by-question across people.
The four pieces of the evaluation block
Each expanded question has four elements, generated by Truffle's AI from your job description and position details.
Why We Ask This. A short explanation of the question's purpose. Example: "This question is designed to reveal how the candidate approaches ambiguous problems — the kind of situations your role will put them in."
What We Look For. The traits being evaluated. Example: "Structured thinking, comfort with uncertainty, willingness to act with incomplete information."
Evaluation Criteria. Anchoring examples of what strong and weak responses look like for this question. Shows what "good" means so your team can calibrate.
Per-question match percentage. The candidate's match score on this specific question, using the same 0–100% scale as the overall match. A candidate can have a high overall match but a weak score on one question that matters. This makes that visible.
How to use per-question scores
Per-question scoring helps you identify the right candidate for the specific work, not just the highest-scoring candidate in general.
Compare candidates on one criterion. Expand the same question across three candidates to see who's strongest on, say, communication or customer empathy.
Spot mixed candidates. Someone at 65% overall match with 90% on your most important question might be a better hire than someone at 80% overall with 50% on that question.
Inform follow-up interviews. A strong score on one question and a weak score on another gives you a ready-made agenda for a live conversation.
How the evaluation is generated
The AI builds the Why / What / Criteria for each question from your job description, role details, and. If you've filled it in. The AI personalization section on the position. More detail in those fields produces sharper evaluation content.
You can edit the question itself (in Step 3 of the wizard) and Truffle regenerates the evaluation block. You can't currently edit the generated Why / What / Criteria by hand. They're always derived from the question and your role context.
Related
What does the match percentage mean?. The same scoring model, applied to the overall interview.