Sample report. A made-up candidate and made-up answers, so you can see what you get after an interview.

Management Consulting Expert

Full interview · Mercor style · Real conditions · 18 September 2026

62out of 100Nearly ready

You structure answers well and are honest about your role, but your results rarely come with a number. Your strongest skill was keeping it short, and being specific held the score back most.

Answered 5 of 5Average answer 1:36Over time 2

Your score by skill

Being specific2.6/5
The saving is described as "significant" with no figure, baseline or timeframe.
Owning your work3.6/5
Mostly "we": your own decisions are hard to pick out.
Depth3/5
No control group, baseline adjustment or tracking of where the cost went.
Keeping it short3.8/5
The point arrives after a minute of background on the client.

AI-generated practice scores, not a prediction of hiring outcomes.

Your turn

Get a report like this

Pick a real Mercor or micro1 role and practise its interview. The free one is 10 minutes and 3 questions.

Start free practice interview See pricing

Your words
Went well
Fix next
Stronger answer
Your plan

Fix next

Three things to fix first

1
Put a number on every result
Most answers ended with "significant" or "fine" instead of a figure.
Try this: Write your three best results as one sentence each, with a number, and open with them.
2
Say "I decided"
In the cost programme answer, your own decisions were lost in "we".
Try this: For each story, write the one decision that was yours and why you made it.
3
Explain how you checked
The follow-up on the savings had no method behind it.
Try this: For each result, add one sentence on how you knew it was real.

Quick wins

Open with the answer, then one concrete number. If you cannot name a number, name the method and the result.
For every method you name, add how you did it and how you checked it worked.
Say “I did X”, then what the team did. Stating the limit of your role honestly scores better than claiming all of it.

Went well

What you did well

Clear structure
Your case answer and introduction were signposted and easy to follow.
Credible about your role
You separated your own work from the team's and the partner's.

Your plan

Your next 7 days

Day 1
Write your three best results as one sentence each, with a number.
Day 2
Re-record the cost programme answer, opening with the result.
Day 3
For each story, write the decision that was yours and why.
Day 4
Practise two case openings that start with a hypothesis.
Day 5
Add one sentence on how you checked each result.
Day 6
Time five answers; cut any setup longer than one sentence.
Day 7
Take a full mock interview and compare the scores with this one.

For the week: one shape for every answer, then how to fill it for each kind of question.

› The shape of every answer

1Answer the question in your first sentence.
2Give three signposted points.
3Include one real number or named specific.
4Say where it breaks: “Where this breaks is… so I would also…”
5Close decisively, then stop.

› How to fill it, by question type

Knowledge or process question

method · tools · metrics

"How do you reconcile two ledgers that disagree?" Method: match by document number, then by amount and date. Tools: the ledger export and a pivot on the differences. Metrics: unmatched count and unmatched value, with a threshold for escalation.

Case or sizing question

structure · what you would test first · decision rule

"Should the client enter the Lisbon scooter market?" Structure: riders, trips per rider, price per trip. Test first: the trips-per-rider assumption, from two comparable cities. Decision rule: enter if payback is under 18 months at the lower estimate.

Story question (behavioral, resume)

what you did · result · what would have killed it

"Tell me about a time you pushed back on a number." What I did: rebuilt the churn figure from the raw exports and found a 12 percent gap. Result: the board pack was corrected before it went out. What would have killed it: if the exports had been overwritten, I would have had no trail.

Critique or evaluation question

ranked issues · concrete fixes · what you would keep

"What is wrong with this memo?" Ranked issues: the total does not tie to the table (a 400 difference), the recommendation comes last, two charts lack sources. Fixes: correct the sum, move the recommendation to the top, add source lines. Keep: the risk section, which is specific.

› Frameworks, with worked examples

Practitioner knowledge

“What are the best practices for…”, “How do you…”, with no scenario

Approach → method → tools → metrics → pitfall. Name the approach, how you run it, what with, how you know it worked, and where it fails. Put one lived example in the middle.

"What are best practices for month-end close?" Approach: a close calendar with owners. Method: reconciliations before adjustments. Tools: a checklist tied to the ledger. Metrics: days to close, late adjustments. Pitfall: reconciling to a figure that is itself an estimate.

Process or mechanism

“How do you handle…”, “Walk me through your approach”

Mechanism → why it works → how it breaks, and the mitigation. The third beat is the one most people skip, and it is what the follow-up tests.

"Walk me through how you handle a disputed invoice." Mechanism: hold payment, gather the PO, receipt and invoice, compare line by line. Why it works: three documents rarely agree by accident. How it breaks: partial deliveries; the mitigation is paying the matched lines only.

Case or scenario

A client, a number and a decision to make

Clarify the objective → three non-overlapping drivers → which you would test first and with what data → decision rule → main risk.

"A retailer wants to cut delivery cost by 15 percent." Objective: cost per order, not cost per route. Three drivers: route density, vehicle fill, failed deliveries. Test first: failed deliveries, from last month's logs. Decision rule: act on whichever driver moves cost per order most per euro spent.

Market sizing

“Estimate the market for…”

Structure before numbers: base → segment → penetration → frequency → price. Then a bottom-up cross-check and one sanity anchor. The structure and the check are what score.

"Estimate the number of dentists in Spain." Base: 48 million people. Segment: one dentist per 1,500 to 2,000 people in comparable countries, so 24,000 to 32,000. Cross-check: about 40 dental schools graduating 1,500 a year over a 35-year career gives around 50,000 licences, of which perhaps 60 percent practise. Anchor: about 30,000.

Behavioral

“Tell me about a time…”

STAR, with 60% on the action. One-sentence situation → your steps in order, in “I” language → result with a number → one line of reflection. No villains.

"Tell me about a time you missed a deadline." Situation in one sentence. Action, in "I" language and in order: I told the sponsor two days early, cut the scope to the two decisions that mattered, and delivered those on the date. Result: the decisions were taken on time; the rest followed a week later. Learning: I now agree the cut list at kick-off.

Resume deep-dive

Quotes a number or claim from your resume

Claim → your specific role → method → outcome → what would have killed it. Define every metric you cite before they ask.

"Your resume says you saved 300,000. How?" Claim: 300,000 a year in freight. My role: I ran the analysis and the tender, not the negotiation. Method: lane-level cost per pallet against three quotes. Outcome: 310,000 in the first year, verified in the ledger. What would have killed it: volume falling below the tender minimums.

Document critique

“What is wrong with this…”

Verdict → issues ranked by severity (wrong numbers, then structure, then polish) → a fix for each, not just a flag → one strength → top-three close. Check the base and size of every headline number first.

"What is wrong with this slide?" Verdict: not ready for a client. Issues by severity: the growth figure contradicts the appendix, the title states a finding the chart does not show, the axis starts at 40. Fixes: correct the figure, rewrite the title as the chart's message, start the axis at zero. Strength: the segmentation is the right cut.

Deliverable quality check

“How do you know it is good enough?”

Three passes: storyline (answer first, action titles, no overlaps or gaps) → numbers (tie out, trace to source, recompute the ones that matter) → audience (right level, risks, so-what, a sceptical reader’s view).

"How do you know a deliverable is good enough?" Storyline pass: answer first, action titles, no gaps. Numbers pass: every figure ties to a source and the totals add. Polish pass: consistent units, dates, names. It ships when the numbers pass is clean; storyline issues get one more edit.

AI evaluation

“What makes an answer good?”, “Rate this response”

Six dimensions: problem structure, reasoning validity, factual accuracy, prioritisation and so-what, actionability, calibration. Good rubric criteria: 4 to 6, independent, observable, each with anchors.

"Rate this model answer." Structure: does it answer first? Reasoning: are the steps valid? Facts: are the numbers right and sourced? Prioritisation: does it say what matters most? Actionability: could a reader act on it? Tone: fit for the audience? Score each, then explain the lowest two.

Writing feedback

“How would you give feedback on…”

Verdict → issues ranked, each quoting the text → rewrite the worst passage → one strength → scores justified against the rubric.

"Give feedback on this paragraph." Verdict in one line. Issues ranked, quoting the text. Rewrite the worst passage. One strength. Scores justified against the issues, not against a feeling.

Governance or prioritisation

“Who decides?”, “What if they tie?”

Method → artifact → one accountable owner. Tie-breakers: regulatory need → unblocks others → cost of delay → can it be staffed.

"Two teams both want the budget. Who decides?" Method: a scoring sheet on regulatory need, unblocking others, and cost of delay. Artifact: the sheet with both bids. Owner: the sponsor signs, once. Tie-breaker: regulatory need first.

Prompt design

“Write a prompt that tests…”

Self-contained, a realistic constraint, one named deliverable, numbers that can be checked, a distractor, and a rubric written alongside. To test reasoning rather than recall, supply all the facts and make them conflict.

"Write a prompt that tests reconciliation skill." Self-contained brief, a realistic constraint (a missing month), one named deliverable (a reconciliation table), numbers that can be checked, one distractor (a superseded extract), and a rubric written alongside so the grader and the writer agree.

Detail

Question by question

Open a question to see what you said, what worked, and a stronger answer.

Background

Walk me through your background and why you want this role.

8/10

Strong

A clear, well-ordered introduction that lands on why you fit the role. It would be stronger with one number from your consulting work.

Being specific3/5
Names the firm type and sectors, but no result or number backs up the experience.
You said: “mostly retail and consumer goods, some financial services”
Missing: One measurable result from a past engagement.
Try: Add: "My last programme took 4% out of a retailer's store costs in a year."
Owning your work4/5
Says what you led yourself and where a partner owned the client relationship.
You said: “I ran the day-to-day workstream; the partner owned the client”
Depth3/5
Mentions issue trees and modelling without saying how you used them.
Keeping it short5/5
Answers in the first sentence and closes on why this role, inside 90 seconds.

Went well

Clear opening. Your first sentence said who you are and what you do.
Honest about your role. You separated your work from the partner's, which reads as credible.

Do better

Add one result. Pick your best engagement and give its outcome as a number.

› What you said · spoken, 1:18

“I am a management consultant with about six years at a mid-sized strategy firm, mostly in retail and consumer goods, some financial services. I ran the day-to-day workstream; the partner owned the client. I use issue trees and a fair amount of modelling. I want this role because it is about judging whether an answer is structured and useful, which is what I do every day.”

› What a strong answer sounds like

I am a management consultant with six years at a mid-sized strategy firm, mostly in retail and consumer goods. I run workstreams of three to five people: the last one took 4% out of a 400-store retailer's operating costs within a year. What I enjoy most is turning a messy question into a clear structure and checking the numbers hold up. That is what this role asks for: judging whether an answer is structured, correct and useful, so I would be doing the part of my job I am best at.

Say it better

I have done a lot of cost work across sectors.I have led three cost programmes; the largest saved a retailer $12m a year.

Be ready for: “Which of those engagements are you proudest of, and why?”

Your CV

Your CV says you led a cost programme for a retailer. What did you personally decide, and what was the result?

5/10

Getting there

You described the programme well, but most of the answer was about what the team did. The interviewer asked what you decided and what changed.

Being specific2/5
The saving is described as "significant" with no figure, baseline or timeframe.
You said: “we delivered significant savings for the client”
Missing: The size of the saving, against what baseline, over what period.
What good looks like: States the result in the first sentence: "$12m a year, about 4% of store costs."
Try: Lead with the number, then explain how you got there.
Owning your work3/5
Mostly "we": your own decisions are hard to pick out.
You said: “we looked at all the cost lines and we prioritised labour”
Missing: The one call you made yourself and why.
Try: Say "I decided to start with labour scheduling because it was 60% of store cost."
Depth3/5
Names benchmarking and a labour model without saying how either worked.
Missing: How the labour model set staffing levels, and what data it used.
Keeping it short3/5
The point arrives after a minute of background on the client.
Try: Open with the decision and the result; keep the background to one sentence.

How it sounded

Long run-up: about 60 seconds before the first point.

Went well

Good structure. You took the programme in a sensible order: diagnose, prioritise, deliver.

Do better

Say what you decided. Swap "we" for "I" where the decision was yours, and name it.
Give the result as a number. An interviewer cannot score "significant".
How you measured the saving

› What you said · spoken, 2:11

“So the client was a large retailer with around four hundred stores and margins had been falling for a few years. We were brought in to look at costs. We looked at all the cost lines and we prioritised labour, then we built a labour model and benchmarked the stores against each other. We rolled it out over about a year and we delivered significant savings for the client.”

› What a strong answer sounds like

I led the store-cost workstream for a 400-store retailer and the programme saved $12m a year, about 4% of store operating cost. My main decision was where to start: I chose labour scheduling over property and energy, because labour was 60% of store cost and we could measure it weekly. I built the staffing model from twelve months of till data, tested it in twenty pilot stores for eight weeks, and only rolled it out once the pilots held service scores. The partner owned the client relationship; the model, the pilot design and the roll-out plan were mine.

Say it better

We delivered significant savings for the client.I set the target at 4% of store costs and we hit it in eleven months: $12m a year.
We looked at all the cost lines.I started with labour scheduling, because it was 60% of store cost and the easiest to measure.

The shape behind it: Result, decision, method, check

1Result first (10s)
2Your decision (20s)
3How (30s)
4Your part versus the team (15s)

Be ready for: “How did you check the savings were real and not just moved to another budget?”

Your CVFollow-up

How did you check the savings were real and not just moved to another budget?

4/10

Needs work

The question asked how you checked the saving. The answer said the client agreed with it, which is not a check.

Being specific2/5
No method, data source or comparison is named.
You said: “the finance team signed it off at the end”
Owning your work3/5
Clear that finance signed off, but not what you did to make the number trustworthy.
Depth2/5
No control group, baseline adjustment or tracking of where the cost went.
Missing: A comparison against stores that did not change, or a tracked baseline.
What good looks like: Names the control stores and how volume changes were taken out.
Keeping it short4/5
Short and direct, but short on substance.

Went well

Direct. You answered quickly and did not pad the answer.

Do better

Name the check. Say how you compared pilot stores with similar stores that did not change.
A control group or baseline
Whether cost moved to overtime or agency staff

› What you said · spoken, 0:34

“We tracked it with the client every month and the finance team signed it off at the end, so they were comfortable the numbers were right.”

› What a strong answer sounds like

Two ways. First, I compared the twenty pilot stores with twenty matched stores of similar size and sales, so the saving was measured against what would have happened anyway, after adjusting for volume. Second, I tracked overtime and agency spend in the same stores every week, because that is where labour savings usually leak back. Overtime rose slightly in month one and then settled, so the net saving was about 10% lower than the gross figure, and that net figure is the one finance signed off.

Say it better

The finance team signed it off.I compared the twenty pilot stores with twenty matched stores, adjusted for sales volume, and tracked overtime so cost could not hide there.

Case

A regional bank's branch profits fell 15% in two years. How would you structure the problem?

7/10

Strong

A clean structure with stated assumptions. It needed one hypothesis to test first and a sense of which branch matters most.

Being specific3/5
The tree is sound, but no number or hypothesis gives it a starting point.
Owning your work4/5
You said what you would look at first and why, in the first person.
Depth4/5
Revenue split into volume and margin, costs into fixed and variable, with the data you would ask for.
Keeping it short4/5
Signposted three branches and closed with next steps.
You said: “Three areas. First, revenue per branch”

Went well

Signposted. "Three areas. First..." made the answer easy to follow.
Asked for the right data. Branch-level revenue and cost, split by product, is what you would need.

Do better

Lead with a hypothesis. Say which branch you suspect first, for example that customers moved to mobile banking.
A first hypothesis to test

› What you said · spoken, 1:36

“Three areas. First, revenue per branch: volume of transactions and products, and the margin on each. Second, costs: staff, property and everything else, split into fixed and variable. Third, the market: are competitors seeing the same thing. I would ask for branch-level revenue by product and the cost base for the last two years, then look at which branches fell most.”

› What a strong answer sounds like

Profit is revenue minus cost, so I would split the 15% fall into those two first. My first guess is falling branch volume as customers move to the app, which would show up as fewer transactions per branch with a largely fixed cost base. I would ask for two years of branch-level revenue by product and branch costs split into staff, property and other. If volume is the driver, the options are fewer branches, smaller formats or moving staff to advice, and I would size each before recommending one.

Say it better

I would look at revenue, costs and the market.My first guess is falling transaction volume as customers move to the app, so I would test that first.

Be ready for: “If volume is the problem, would you close branches? What would make you say no?”

Behavioural

Tell me about a time a client disagreed with your recommendation.

6/10

Getting there

A good story with a real disagreement. The ending was vague: say what the client decided and what happened next.

Being specific3/5
The situation is clear, but the outcome has no detail.
You said: “in the end it worked out fine”
Owning your work4/5
Your part in reworking the analysis is clear.
Depth3/5
Mentions re-running the analysis without saying what changed in it.
Keeping it short3/5
Runs past two minutes; the setup could be one sentence.

How it sounded

Over time: 2:20 against a 2:00 target.

Went well

A real conflict. You chose an example where the client pushed back on substance, not style.

Do better

Finish the story. End with what the client chose and one result.

› What you said · spoken, 2:20

“We recommended closing fifteen underperforming stores and the client's operations director disagreed strongly, because some of those stores were in towns where they had been for decades. I went back and re-ran the analysis store by store with his team, and we found some of the stores had costs allocated to them that were really head-office costs. We changed the recommendation and in the end it worked out fine.”

Be ready for: “What would you do differently if it happened again?”