Customer service assessment test for hiring: questions, rubric, and workflow

What to screen for in a customer service hire

Post a customer service role and you'll have 150 applicants by Friday. Every one of them is "a people person." The resume can't tell you who stays calm with an angry customer, who writes a clear refund email, and who quietly bends policy to end a hard conversation. Those are the behaviors that decide whether your CSAT holds, and they're exactly what this kit screens for.

If you're hiring this role now, pair this test with the customer service representative hiring guide for the full process.

Nine skill areas map to the quality standards you already care about:

• Empathy and rapport: acknowledges emotion, validates the experience, respectful tone.
• Active listening and information gathering: asks targeted questions, confirms understanding.
• De-escalation: maintains calm, reduces friction, sets boundaries professionally.
• Ownership and follow-through: next steps, timelines, accountability. No handoffs into the void.
• Problem solving: structured reasoning, isolates variables, tests solutions.
• Policy judgment: applies policy consistently, knows when to escalate, avoids compliance breaches.
• Prioritization: triages by impact and urgency, protects SLAs.
• Written communication: clear, concise, correct, customer-friendly.
• Documentation: captures the right details for continuity.

The test structure

A balanced screen has two components and takes 25 to 35 minutes:

• Situational judgment questions: 12 to 18 multiple-choice scenarios showing how a candidate approaches trade-offs under pressure.
• Written work sample: one email reply and one chat reply, scored with the rubric below. Writing is what your customers actually receive, so score it like QA would.

Optionally add a lightweight policy overlay for your industry (returns windows, identity verification, privacy rules).

The scenario bank: 10 questions with answer guides

One note before you use these: there isn't a universal right answer. Teams weigh speed vs thoroughness and strict policy vs retention differently. The answer guide below reflects common preferences. Adjust the key to match your environment, then apply it to every candidate identically.

What strong options usually do: acknowledge the customer, clarify key facts, resolve within policy, set expectations, and document or escalate appropriately.

Question 1: angry customer, late delivery (ecommerce)

Context: customer says: "My order was supposed to be here yesterday. This is ridiculous. I'm never buying from you again." Tracking shows a carrier delay due to weather; the package is now expected tomorrow. What do you do first?

• A. Apologize and immediately offer a full refund.
• B. Explain that weather delays are not your fault and the carrier is responsible.
• C. Acknowledge frustration, confirm order details, share the updated ETA, and offer a reasonable option if policy allows (shipping fee refund, replacement, or monitoring until delivery).
• D. Ask them to wait 48 hours and contact you again if it doesn't arrive.

Most-aligned choice for many teams: C. It balances empathy, clarity, and options while staying within policy. A creates cost and inconsistency, B escalates emotion, D increases repeat contacts.

Question 2: identity verification (finance and telecom)

Context: a caller demands access to an account but fails a security question.

• A. Provide limited details since they sound genuine.
• B. Calmly explain you can't access the account without verification; offer the approved verification reset path.
• C. Transfer to a supervisor immediately.
• D. End the call to avoid risk.

Most-aligned choice for regulated environments: B. It protects compliance and still provides a path forward. Treat this as a gate item: a miss here is disqualifying regardless of total score.

Question 3: queue triage

Context: three tickets: a VIP customer can't log in (business-critical), a refund request within the policy window, and a non-urgent feature question.

• A. Handle tickets in the order received.
• B. Start with the easiest one to reduce backlog.
• C. Triage by impact and urgency: login issue first, then refund, then feature question.
• D. Ask your manager what to do.

Most-aligned choice: C.

Question 4: abusive language in chat

Context: a customer swears at you repeatedly.

• A. Respond with the same intensity so they understand it's unacceptable.
• B. Ignore the language and keep troubleshooting.
• C. Set a respectful boundary, refocus on resolution, and warn of chat termination per policy if it continues.
• D. End the chat immediately without explanation.

Most-aligned choice: C.

Question 5: ambiguous issue (SaaS)

Context: customer reports: "Your app is broken. It won't work." No screenshots.

• A. Send a generic help article link.
• B. Ask targeted questions (device, browser or app version, error message, steps to reproduce) and propose the first diagnostic step.
• C. Escalate to engineering immediately.
• D. Tell them to reinstall and close the ticket.

Most-aligned choice: B.

Question 6: refund outside the policy window

Context: a customer requests a refund 10 days past the window. They claim they never used the product. You can see usage logs.

• A. Approve the refund to preserve goodwill.
• B. Deny the refund bluntly and close the ticket.
• C. Explain the policy, reference usage data neutrally, offer an alternative (credit, troubleshooting, downgrade), and escalate only through a documented exception path.
• D. Ignore the usage logs and ask them to send proof.

Most-aligned choice: C.

Question 7: ownership on a handoff

Context: billing must investigate. The customer is already frustrated from being transferred twice.

• A. Transfer immediately and end the interaction.
• B. Tell the customer billing will contact them "soon."
• C. Summarize the issue, set a timeline, provide a case number, and tell them what you'll do next.
• D. Ask the customer to contact billing directly.

Most-aligned choice: C.

Question 8: written clarity

Context: customer asks a simple question: "How do I reset my password?"

• A. Send a long paragraph explaining all security features.
• B. Provide 3-5 clear steps, include the reset link, and offer help if they get an error.
• C. Tell them to search the help center.
• D. Ask why they want to reset it.

Most-aligned choice: B.

Question 9: ticket notes

Context: you're ending a call after troubleshooting a connectivity issue.

• A. Note: "Helped customer, issue fixed."
• B. Note: full reproduction steps, what you checked, what worked, customer confirmation, next steps if it recurs, and relevant tags.
• C. No notes needed if resolved.
• D. Copy and paste the customer's first message only.

Most-aligned choice: B.

Question 10: "bend the rules" request

Context: a customer asks you to mark an item "not delivered" so they can get a replacement faster, but tracking shows delivered.

• A. Do it to help the customer.
• B. Refuse and accuse them of lying.
• C. Explain you can't take that action, offer a legitimate path (carrier claim process, proof of delivery review), and document the interaction.
• D. Transfer to another agent.

Most-aligned choice: C.

The written work samples

Prompt A (email): refund with policy constraints

Scenario: a customer emails: "I want a refund. The product didn't work for me." They purchased 45 days ago; the refund policy is 30 days. The candidate can offer store credit or troubleshooting. Ask for a reply of 6-10 sentences.

Prompt B (chat): escalated complaint

Scenario: customer in chat: "I've contacted you three times. No one helps. Fix this now." The ticket shows missing information that blocked resolution. Ask for a reply of 3-6 chat messages.

Writing rubric (0-4 per dimension, 20 per prompt)

1. Empathy and tone
2. Clarity and structure
3. Information gathering (asks for what's missing, only what's necessary)
4. Policy alignment and judgment (firm but fair, offers approved alternatives)
5. Ownership and next steps (timeline, what you'll do, what they should do)

Anchors:

• 4, excellent: validates feelings, concise steps, correct policy framing, clear timeline, proactive follow-up.
• 3, good: mostly clear and polite; minor gaps like an unstated timeline.
• 2, risk: generic or slightly defensive; unclear next steps; misses a key question.
• 1, poor: blunt denial, overpromises, or no structure.
• 0, disqualifying: shares restricted info, violates policy, rude or biased language.

Scoring and hiring benchmarks

Weighting (100 points)

• Scenario questions: 60 points. Most-aligned choice 6, acceptable alternative 3-4, risky 1-2, misaligned 0.
• Written work samples: 40 points (20 per prompt).

Pass ranges and review flags

Suggested starting bar: 70/100 for entry-level, 78/100 for experienced roles. Calibrate against your own 90-day outcomes.

Review flags, even when the total is high:

• Policy and compliance dimensions averaging under 2/4
• De-escalation items consistently chosen poorly
• Written tone showing defensiveness, blame, or sarcasm

What the tiers mean for your decision

• 90-100, lead-ready: calm under pressure, strong judgment on policy edge cases. A candidate you can grow into QA or queue-lead work.
• 75-89, high-performing hire: strong judgment and clear writing. Move to offer conversation quickly; these candidates go fast.
• 60-74, job-ready with gaps: solid on routine cases, shakier on exceptions and escalations. Hire with a coaching plan for the flagged areas.
• 0-59, not yet: friendly intent, inconsistent execution. The risks (policy mistakes, repeat contacts) land on your customers.

Integrity guardrails

• Rotate the scenario bank quarterly if you hire at volume.
• Timebox items (60-90 seconds per scenario on average).
• Add a short "why did you choose that?" field on 2-3 questions. The reasoning is often more revealing than the pick.

Structured interview follow-ups by weak area

• Low policy judgment: "Tell me about a time you had to say no to a customer. How did you keep the relationship intact?"
• Low de-escalation: "Walk me through your exact phrasing when a customer becomes abusive."
• Low troubleshooting: "How do you isolate the cause when you can't reproduce the issue?"
• Low ownership: "How do you handle cases that need another team when the customer expects you to own it?"

Compliance and fairness notes

• Job-relatedness: every item above maps to real support tasks (resolution, documentation, policy adherence). Keep that mapping if you customize.
• Bias reduction: avoid culture-loaded slang, unnecessary idioms, and trick questions.
• Accessibility: plain language, readable formatting, screen-reader compatibility, and extra time where appropriate.
• Privacy: minimize personal data, state a retention period, and comply with GDPR and CCPA where applicable.
• Consistency: same rubric and thresholds per role. Document exceptions.

Run this screen automatically with Truffle

This kit works. Running it manually on 150 applicants doesn't. Someone has to send the test, chase completions, grade two writing samples per candidate, and keep the rubric consistent from applicant 3 to applicant 130.

Truffle is a candidate screening platform that combines talent assessments with resume screening and one-way video interviews. Load the scenario bank as an assessment, use the written prompts as application questions, and add a one-way interview to hear de-escalation tone before any live call. Truffle scores every response against your criteria and hands you a ranked shortlist with the reasoning shown. AI surfaces the evidence. You make the call.

The 7-day free trial includes 30 credits, no card required. Start free trial