Mint.Psychometrics

Transparency

How Mint Psychometrics works

A transparent walkthrough of the assessment experience. See what candidates complete, how we score it, and how you receive your results.

The candidate experienceHow we score

For candidates

What candidates experience

Assessments are fully online, unproctored, and designed to be clear and fair. Here is what the experience looks like. We describe the structure openly; we never publish the content itself.

Executive Leadership Assessment

ELA: around 60 minutes online, plus a verbal assessment

The ELA is built for senior director, VP, SVP and C-suite appointments, where the cost of being wrong is high and the evidence available is usually thinnest. Candidates complete an online assessment of roughly 50 to 60 minutes, capped at 60 minutes of active answering time, followed by a structured verbal assessment of up to 60 minutes.

The online form combines judgement scenarios, forced-choice blocks, trade-off ranking tasks, written responses and risk indicators. Candidates can pause and resume through the same link, no proctoring is required, and adjustments such as extended time are handled by pausing the timer rather than shortening the assessment.

After the online assessment, candidates complete a structured verbal assessment of up to 60 minutes with a trained interviewer. The questions are fixed, so every candidate is asked the same things in the same order — that is what makes the comparison fair.

Sales Talent Assessment

STA: 40 minutes, fully online

The STA is built for sales hiring and team design, from SDR and hunter roles through to account management and sales leadership. Candidates complete it entirely online, typically in 36 to 40 minutes, against a single active-time budget of 40 minutes.

Six sections cover judgement scenarios, forced-choice patterns, commercial reasoning, ranking tasks, written responses and risk indicators. Candidates can pause and resume, no proctoring is required, and extended time is provided by pausing the timer. Idle gaps and breaks do not count towards the budget.

The entire assessment is completed online. No proctoring, no video recording. Pause and resume as needed — the timer stops while you are away.

For hiring teams

How we score and review

Every assessment is scored through a combination of automated analysis and human review. Here is how that works, step by step.

Step 1

AI-assisted analysis

Responses are scored against published rubrics and fixed keys held on the server. Structured sections — scenarios, forced-choice blocks, reasoning items, ranking tasks and risk indicators — are scored deterministically: the same responses always produce the same result, and every report records the exact rubric and parameter versions used.

For the ELA, the online assessment is scored automatically and the verbal assessment transcript is scored against fixed criteria. For the STA, every section is scored automatically except the written responses, which are drafted by AI and routed to a person before they count.

Each draft carries a confidence level alongside the score. AI is used here for evidence extraction and efficiency, never to make a decision.

Step 2

Human review and validation

A trained reviewer reads every AI-generated judgement before it reaches a hiring team. They confirm, amend or override it, and any change is recorded with a written rationale against the evidence.

Risk signals receive additional scrutiny and are corroborated against evidence from other sections before they appear in a report. A flag on its own is a prompt to look closer, never a conclusion.

No assessment is released without human sign-off. AI assists; people decide. We never make automated decisions about anyone.

Step 3

Validity checks

Every result carries a validity layer: consistency across methods, effort and engagement, response-pattern checks, and calibration between how confident a candidate was and how they actually performed.

You are told how much confidence a result deserves before you are told what it says. Where the evidence is thin, the report says so instead of presenting a confident-looking number.

Anomalies — unusually fast completion, high inconsistency, long periods away from the window, incomplete sections — are annotated for review. This protects against gaming, distraction and technical problems alike.

The output

What you get

A comprehensive, defensible profile. Not a verdict; a foundation for an informed decision.

ELA: integrated leadership profile

The ELA report integrates the online and verbal assessments, which are scored independently and weighted equally. Where the two methods disagree, that divergence is reported rather than averaged away — it is usually the most interesting thing on the page.

Scores are reported across six leadership expectations on a one-to-five scale, each with a stated confidence level. Diagnostic themes explain how a leader delivers outcomes, not simply whether they do.

Risk signals are flagged for human review and corroborated before they appear. The report closes with recommended lines of enquiry for your interview — the questions worth asking next, not a verdict.

Sample output — ELA

Leadership expectation profile

Fit band: Strong
  • Future focusStrength4.4
  • Performance orientationStrength4.1
  • Talent focusDeveloping3.2
  • Self awarenessSolid3.6
  • Integrity and respectStrength4.6
  • MotivationWatch2.8

Illustrative example. Two independent methods — an online form and a structured verbal assessment — are weighted evenly, then reviewed by a person before anything is released.

STA: sales capability profile

The STA report gives scores across eight sales dimensions on a one-to-five scale, each with a confidence level and the methods that evidenced it. No dimension rests on a single method.

Three risk sub-indices sit alongside as governance signals rather than capability scores. They are read by a person, corroborated against other evidence, and never trigger an automatic rejection.

Reports export as a branded PDF, with benchmarking data to give scores context. They support individual hiring decisions and cohort analysis for team design equally well.

Sample output — STA

Sales dimension spread

Validity: clear
4.5
4.2
3.1
3.9
2.7
3.6
4.0
4.7
Commercial curiosityAchievement driveResilience under refusalRelationship depthProcess disciplineNegotiation poiseLearning agilityIntegrity in selling

Illustrative example. Forty minutes online, eight dimensions, three risk sub-indices and a validity layer that tells you how much confidence the pattern deserves.

Both examples are anonymised, illustrative output. Full example reports are shared on request during an introductory conversation.

Philosophy

Transparent. Rigorous. Fair.

Structured assessment does not replace human judgement, and it is not meant to. It gives human judgement something honest to argue with — evidence gathered the same way for every candidate, so a disagreement between two managers becomes a discussion about the evidence rather than a contest of confidence.

Our approach is deliberately multi-method. No dimension rests on a single item type: scenarios, forced-choice blocks, reasoning, ranking, written responses and, for the ELA, a structured verbal assessment each catch something the others miss. That triangulation is what makes a profile hard to fake and worth defending.

The validity layer comes before the result, not after it. You are told how much confidence a score deserves, and where evidence was thin we say so. A quiet "insufficient evidence" is more useful than a confident number nobody should trust.

Human review is the final gate. AI drafts; people decide, and every override carries a recorded rationale. Every material action — score change, report export, access to a candidate record — is written to an append-only audit trail, so a decision made today can still be explained in two years.

More detail is set out in our trust and governance centre.

Candidate support

We support candidates throughout

Candidates are told what to expect before they start: how long the assessment takes, what each section asks of them, how it will be scored, and who reviews it. During the assessment, clear instructions appear at the start of every section, and progress and remaining time are always visible.

Adjustments are available and straightforward. Extended time is provided by pausing the timer, so nobody is penalised for needing a break, and screen readers, keyboard navigation and tap-based alternatives to drag-and-drop are supported throughout. Ask us before the assessment begins and we will arrange it.

A technical issue button is available on every screen of every assessment, so a candidate can report a problem without leaving the page or losing their answers. Responses autosave continuously.

The assessment is designed to be clear and fair, not deliberately difficult. The items are demanding because the roles are demanding — never because we made the instructions hard to follow.

View candidate support resourcesContact support
Extended timeProvided by pausing the timer, so the active-time cap is never breached and no candidate is disadvantaged.
Assistive technologyScreen reader support, full keyboard navigation, and tap-to-swap alternatives wherever drag-and-drop is used.
Help while you workA technical issue button on every screen, continuous autosave, and a support contact for anything else.

Questions

Frequently asked questions

Still have a question?

Tell us about the role you are hiring for. We will walk you through the process, share an example report, and set out what a pilot would look like.

Talk to us about a role