How scoring works

Every score, explained.

What each Smuker score measures, how to read it, and how we show you where you’re lacking.

1

The short version

Three kinds of score

Code counts what can be counted. AI judges what needs judgement. Neither is allowed to invent a number about you.

Answer scores · 1 to 10

How good the content of an answer is, judged by an AI coach against a fixed, written rubric.

Delivery · counted

How an answer sounded: pace, filler words, pauses. Counted from your recording, so the same recording always gives the same result.

Resume & fit · 0 to 100

How your resume lines up with a saved job. AI estimates against the job description.

2

Which score, where

Which score you get, and where

A “tell me about a time” story and a “why this role?” answer need different yardsticks.

Where you practiseHow it’s scored
Practice Q&A by Role (questions from a saved job)STAR, 1–10 per part + overall
Practice Q&A by Topic: Behavioral, Recruiter, Entry Level5-point rubric, 1–10 per dimension + overall
“Practice this” on a Role match requirementSTAR or the 5-point rubric (you choose) + the interviewer’s take
Mock interviewInterviewer’s score per question and overall, plus a hire verdict
Any answer you speakVoice delivery feedback
Spoken mock interviewsDelivery composure, 0–100
Saved jobs and resumesRole match, ATS and Recruiter scores, 0–100
3

STAR score

The STAR score

For behavioral stories. Each part is scored 1–10, plus a separate check that you answered the question asked.

Situation

Is the context clear? Scale, timeline, and what was at stake.

Task

What you were responsible for, and the constraints you were under.

Action

The specific steps you personally took, and why.

Result

The outcome, ideally measured.

ScoreWhat it means
8–10Clear, specific, compelling. Would impress an interviewer.
5–7Present, but vague or underdeveloped.
1–4Missing or very weak.

On-topic match is scored on its own: a well-told story about the wrong thing still scores low. The overall is the coach’s judgement of the whole answer, not an average of the parts. We also check whether you said “I” or “we”, and whether your answer ended on a measurable result.

Example

“Tell me about a time you improved a process.”

Answer A: “Our release process kept slipping… The team agreed something had to change. We started holding more stand-ups and things got a lot better after that.”

All “we” — name what you didAdd a number to your result

Answer B: “…As the delivery lead, my job was to get us back on schedule. I mapped every hand-off and found code reviews waited three days. I set a 24-hour review rule… Within two months, we hit five of the next six sprint dates.”

Closed with a number
4

5-point rubric

The 5-point rubric

For questions that aren’t a single story, and for the whole topic bank. Five dimensions, each 1–10, plus an overall.

1 · Relevance

Did you answer this question, including every part it asks?

2 · Clarity

Can an interviewer grasp your main point quickly?

3 · Evidence

Did you back up your claims with specific examples, facts or numbers?

4 · Impact

Did you show real value: a result, a measurable outcome, or clear growth?

5 · Differentiation

Did you give the interviewer a reason to remember you?

Overall

The coach’s judgement of the whole answer, not an average of the five.

Each question is judged against what that particular question is looking for. After you answer, the question list shows your best score:

Good · 8 and upAvg · 5–7Poor · below 5
Real example

“Describe a project where you realized the problem statement you were given was actually the wrong one. How did you figure that out?”

Vague answer

“Yeah, so on one project we were asked to build a new dashboard for the sales team. We started building it and then realized it wasn’t really what they needed. We talked to some people and figured out the actual problem was different…”

Relevance4
Clarity5
Evidence2
Impact3
Differentiation2

Overall 3/10

Coaching: The biggest fix is specificity: name the original dashboard request, the actual evidence that surfaced the real problem, and the design you built instead.

Specific answer

“Last year our VP of Sales asked my team to build a pipeline dashboard because forecasts kept missing by 20 to 30 percent. Before building anything, I sat in on three forecast calls and pulled six months of deal history… Forecast error in that region dropped from 27 percent to 9 percent.”

Relevance9
Clarity9
Evidence9
Impact9
Differentiation8

Overall 9/10

Coaching: To push it from strong to exceptional, open with a one-line thesis naming the mismatch before the story, and add a sentence on the business-dollar or confidence impact of the fix.

5

Where you’re lacking

How we tell you where you’re lacking

A score without a fix is just a grade. Every dimension comes with feedback.

Your scoreWhat the feedback does
8–10Says what worked, then gives one refinement to reach a 10.
5–7Names the exact gap, then gives a concrete instruction.
1–4States what is missing and shows what a strong version would include.

The overall coaching leads with your single most important fix. At the end of a practice session, a summary shows your strongest area and up to three things to work on next.

6

Voice delivery

Voice delivery

When you speak an answer, we look at how it sounded as well as what you said.

Pace

How fast you speak, labelled slow, conversational, brisk or fast.

Filler words and hedging

“Um”-type fillers, and uncertain phrases like “I think” or “maybe”, counted separately.

Pauses

Long silences mid-answer, and where they happened: reaching for the story, or running out of material.

Length

Aim for 1–2 minutes per answer. Past about two minutes, interviewers start losing the thread.

The coach only raises a delivery problem when the measurements show one. If your delivery was clean, it says so.

7

Delivery composure

Delivery composure (0–100)

One number for how a spoken mock interview was delivered. It measures habits a listener can hear, not your tone of voice or how you felt.

Fluency

Filler words and false starts.

Steadiness

Whether your pace held, and how often you stalled.

Decisiveness

Hedging, and how long you took to start.

Concision

Answer length and run-on sentences.

Ownership

“I” vs “we”, and whether answers ended on a result.

Why

One weak habit holds the whole score back, because that’s how a listener experiences it. The headline names the habit to work on first.

8

Mock interview

The mock interview debrief

Scored the way a real interviewer thinks: once, at the end, across the whole conversation.

Each answer gets a 1–10 score and a note, and the interview gets an overall score, which is the interviewer’s judgement rather than an average. It ends on one of four verdicts, defined the same way every time:

VerdictWhat it means
Strong hireThey'd push to advance you without reservation. Your answers were specific, well-structured, and showed clear evidence of your own impact.
Lean hireThey'd advance you, with a caveat in the notes. The substance is there, but at least one answer needed prompting before it got concrete.
BorderlineThey'd want another conversation before deciding. Your answers were plausible but thin — short on specifics, or unclear on what you personally did.
Not yetThey wouldn't advance you off this conversation. Answers were vague, drifted off the question, or didn't show your own contribution.

Answer a question again and the debrief compares both attempts. Across mocks, it checks whether you fixed what it flagged last time.

9

Resume & fit

Resume & fit scores (0–100)

Three gauges on every saved job, each an AI estimate against that job description.

ScoreWhat it estimates
Role matchHow much of the job’s requirements your resume shows, requirement by requirement.
ATS scoreHow well your resume’s keywords and structure match the job for applicant tracking systems.
Recruiter scoreHow a busy recruiter would rate your resume after a quick skim.
Strong fit · 80 and upGood fit · 60–79Needs work · below 60
10

Honest limits

Honest limits

What these scores are, and what they aren’t.

Can the same answer score differently?

Answer scores are an AI judgement, so they can move a point either way. Delivery measurements never change for the same recording.

Is a 6 a bad score?

It means there’s a real gap, and the feedback tells you exactly what it is.

Do you judge my tone or nerves?

No. We score habits a listener can hear, never your tone of voice.

Is my answer rewritten?

No. Scores and feedback are about what you actually said. Model answers are shown separately and labelled.

How many answers you can get scored depends on your plan. See plans.

See your own scores

Answer one question and get the full breakdown.

Start free