Home/Chapter 5

Chapter 5. Interview Process

pp. 68–97

**Interview loop**The full interview cycle from preparation to debrief and next-step ownership.

**Interview process map**A visual structure for keeping the interview process consistent across roles and interviewers.

How to use this chapter

Working scenario. A candidate has completed four interviews. At the debrief, the first interviewer says the candidate

is "strong on communication", the second says "there wasn't enough depth", the third asked almost the same questions as the first,

and the fourth wasn't even assessing the right criterion. The team has plenty of opinions but few facts. The candidate is tired,

the hiring manager is uncertain, and the recruiter has nothing to put towards the decision beyond impressions.

What actually went wrong here: the interviews were meetings, but not a system for gathering assessment facts. To

make interviews support decision-making, each stage must check its own criterion, each interviewer must

understand their area, and the debrief must discuss facts rather than likeability.

This chapter does not repeat all the theory on candidate assessment from Chapter 6. That chapter covers professional

skills, behavioural skills, motivation, signal reliability, scorecards and scale anchors in detail. The focus here is different:

how to design an interview cycle, assign areas between interviewers, prepare behavioural and

situational questions, run a panel interview without voting on impressions, conduct a debrief and

honestly pass assessment facts to the decision-maker.

Use this chapter as a self-contained working module: it contains a process map, role tables, templates,

prompts, SOP snippets, case studies and a maturity model. If you need a quick implementation, use sections 2, 4, 8, 13 and

the desk reference. If you need to train interviewers, use sections 3, 5, 6, 7 and the case studies.

Remember in one phrase: a good interview does not start with a question to the candidate, but with the criterion that

question is designed to assess.

If your task is	Go to section	What you will get
Explain the interview process	1	Boundaries and purpose of the interview cycle
Design the stages	2	Interview cycle map and stage objectives
Prepare a structured interview	3	Criterion-question-assessment fact model
Use behavioural questions	4	STAR without mechanical interrogation
Use situational questions	5	Role-realistic scenarios and scales
Run a panel interview	6	Rules for no impression-based voting
Assign interviewer roles	7	Interviewer roles and context handover
Implement the process	8	Implementation plan
Measure quality	9	Interview quality metrics
Configure the ATS workflow	10	Neutral HarmonyATS example
Work through complex cases	11	Case study tables as mini-guides
Use AI	12	Prompts with human oversight
Create an SOP	13	Procedure snippets
Assess maturity	14	Maturity model

Minimum starting point

If interviews are already running but quality is inconsistent, start not by training everyone, but by introducing one standardised

cycle.

Step	What to do	Artifact
1	Assign each interview one assessment area	Interview map
2	Prepare 3 core questions and 2 follow-ups per criterion	Interview kit
3	Ask interviewers to complete notes before the debrief	Feedback deadline
4	Run the debrief by criteria, not by overall impression	Debrief agenda

1. What is the interview process

The interview process is a sequence of meetings where each meeting has a purpose, defined criteria, prepared interviewers, rules for note-taking and context handover for decision-making. The core principle is simple: if a stage does not add new assessment facts, it should be changed or removed. An interview should not prove whether someone is a "good person". It should answer: are there sufficient facts related to the role's tasks to progress, pause, clarify, or make an offer? Ethical and operational discipline align here: the more structured the process, the less the team depends on impressions, charisma, fatigue, similarity bias and late-recorded recollections.

Interview type	What it checks	When to use it	What it must not do
Recruiter interview	Motivation, expectations, basic fit, process risks	After screening or as the first structured stage	Duplicate the technical interview
Hiring manager interview	Role outcomes, responsibilities, priority skills, motivation fit	Almost always	Become a general chat without a scorecard
Technical / functional interview	Professional skills, domain maturity, decision-making, trade-offs	When critical professional skills are required	Assess personal likeability
Behavioural interview	Past behaviour in relevant situations	For behavioural skills, responsibility, communication, conflict	Accept stories without follow-up questions
Situational interview	Reasoning in a realistic future scenario	When past experience is limited or the role context is unique	Ask fantastical riddles
Case / work sample review	Work output and thinking process	For skills with a high impact on outcomes	Request free work from the candidate
Panel interview	Collaboration in a work-related context	For cross-functional or leadership roles	Vote "one of us / not one of us"
Final alignment	Offer risks, mutual expectations, unanswered questions	Before the offer	Replace missing assessment facts with selling the role

2. Designing the interview cycle: stages, objectives, inputs and outputs

The interview cycle does not start with the calendar, but with the criteria. First the team identifies which risks and skills need to be checked. Then it selects the minimum set of meetings where each stage provides a new signal.

Recruiter interview Motivation, expectations, basic After screening or as the first Duplicate the technical

fit, process risks structured stage interview

Hiring manager Role outcomes, responsibilities, Almost always Become a general chat without

interview priority skills, motivation fit a scorecard

Technical / functional interview Professional skills, When critical professional skills Assess personal likeability

domain maturity, trade-offs

Behavioural interview Past behaviour in relevant For behavioural skills, Accept stories without follow-up

situations responsibility, communication, questions

conflict

Situational interview Reasoning in realistic future When past experience is limited Ask fantastical riddles

scenario or role context is unique

Case / work sample review Work output and thinking process For skills with a high impact on Request free work from the

outcomes candidate

Panel interview Collaboration in a work-related For cross-functional or Vote "one of us / not one of

context leadership roles us"

Final alignment Offer risks, mutual expectations, Before the offer Replace missing assessment facts

unanswered questions with selling the role

2. Designing the interview cycle: stages, objectives, inputs and

outputs

The interview cycle does not start with the calendar, but with the criteria. First the team identifies which risks and skills

need to be checked. Then it selects the minimum set of meetings where each stage provides a new signal.

Design step	Question	Output
1. Criteria review	Which must-have criteria have already been approved?	5–7 criteria from the hiring plan / scorecard
2. Risk mapping	Where is the cost of error highest?	Must-check risks
3. Assessment fact mapping	Which method will yield the best signal?	Interview / case / portfolio / reference check
4. Stage design	Which stages are needed to avoid duplicating signals?	Interview cycle
5. Role assignment	Who checks which criterion?	Interviewer brief
6. Question design	Which behavioural/situational questions and follow-ups?	Question bank
7. Scorecard setup	How do you record rating, assessment facts, confidence?	Scorecard template
8. Debrief rules	How is a decision reached?	Decision process and owner
Example cycle for Head of Customer Implementation:
Stage	Objective	Interviewer	Assessment facts
Recruiter structured screen	Motivation, compensation, availability, role realism	Recruiter	Motivation snapshot and process risks
Hiring manager interview	Responsibility, implementation leadership, stakeholder management	VP Customer	Behavioural assessment facts and role fit
Functional case discussion	Project planning, risk diagnostics, customer escalation	Implementation lead	Scenario reasoning, trade-offs
Panel interview	Cross-functional collaboration and conflict maturity	CS, Product, Support representatives	Observed behaviour against criteria
Debrief	Assessment fact synthesis and decision	Hiring owner + recruiter	Decision log

Component	How it works	Example
Probe	Clarifies personal contribution and depth	"What exactly did you do? What trade-offs did you make? What did you change after the conflict?"
Strong signal	Describes expected assessment facts	Candidate maps stakeholders, identifies interests and builds a path to a solution
Weak signal	Describes risk	Candidate says "we agreed" without explaining their role, actions or outcome
Scale anchor	1/3/5 from Chapter 6 or local scorecard	1 = no structure; 3 = handles obvious stakeholders; 5 = manages conflict and decision systems
Confidence	How reliable is the signal	High if detailed and consistent; low if the story is generic

Structured interviews are fairer when interviewers do not invent criteria after meeting the candidate. New information may create a next step, but not a hidden requirement. If an interviewer discovers a genuinely missing criterion, the process owner must recalibrate the role and decide whether to apply the criterion only to future candidates.

4. Behavioural questions: past behaviour as assessment facts

Behavioural questions ask about past behaviour in relevant situations. They work because real examples reveal context, actions, trade-offs, responsibility and outcomes. But they fail when the interviewer accepts a polished story without asking follow-up questions.

Block	Questions
Situation	"In what situation did this happen? Why was the task important?"
Task	"What were you specifically responsible for?"
Action	"What decisions did you make? What alternatives were available?"
Result	"How did it end? How did you measure the outcome?"
Learning	"What would you do differently now?"
Reliability	"Who can confirm this context? What artifacts were produced?"
Examples:
Criterion	Behavioural question	Follow-up questions	Assessment facts to capture
Accountability	"Describe a task where the outcome depended on you but you lacked full authority."	"How did you gain alignment? What did you escalate? Where did you go wrong?"	Actions, responsible behaviour, outcome, limits
Communication	"Give an example of when you had to explain bad news to a stakeholder."	"How did you frame the impact? What did you propose? How did you verify understanding?"	Clarity, risk communication, next steps
Ability to learn	"A skill you had to pick up quickly for a role."	"How did you identify the gap? How did you practise? How did you measure progress?"	Learning loop, feedback use

Behavioural questions ask about past behaviour in relevant situations. They work because

real examples reveal context, actions, trade-offs, responsibility and outcomes. But they fail

when the interviewer accepts a polished story without asking follow-up questions.

Block Questions

Situation "In what situation did this happen? Why was the task important?"

Task "What were you specifically responsible for?"

Action "What decisions did you make? What alternatives were available?"

Result "How did it end? How did you measure the outcome?"

Learning "What would you do differently now?"

Reliability "Who can confirm this context? What artifacts were produced?"

Examples:

Criterion Behavioural question Follow-up Assessment facts to

questions capture

Accountability "Describe a task where the "How did you gain alignment? Actions, responsible

outcome depended on you but What did you escalate? Where behaviour, outcome,

you lacked full authority." did you go wrong?" limits

Communication "Give an example of when you "How did you frame the Clarity, risk communication,

had to explain bad news to a impact? What did you propose? next steps

stakeholder." How did you verify understanding?"

Ability to learn "A skill you had to pick up "How did you identify the gap? Learning loop, feedback

quickly for a role." How did you practise? How use

did you measure progress?"

Criterion	Behavioural question	Follow-up questions	Assessment facts to capture
Conflict maturity	"An occasion where you disagreed with a colleague or manager at work."	"What was the disagreement about? How did you preserve the relationship? What decision was made?"	Work-related disagreement handling
Customer decision maturity	"A situation where a client was unhappy, but the cause was not obvious."	"How did you diagnose the risk? What did you do first? What changed?"	Diagnosis, prioritisation, follow-through

Example:

Scenario Strong signal Weak signal Follow-up

Sales candidate receives an Qualifies the deal, refines the "I'll push it over the "What needs to happen for the

enterprise deal with high ARR, but buying process, doesn't the deal to stay in forecast?"

the buyer won't provide access to forecast without a next step

the economic decision-maker

Finance candidate spots a variance Assesses materiality, reports "I'll double-check everything" "Who do you inform and when?"

in monthly close before the board risk, sets controls and a without prioritisation

meeting correction plan

Marketing candidate sees a drop in Formulates hypotheses, Tests quality of traffic without "What data do you need in the

conversion after a redesign segments, tests data, a plan for data first 2 hours?"

rollback or experiment

Recruiter candidate receives 60 Sorts by must-have criteria, "I'll forward all CVs" "How do you protect the

inbound CVs and the hiring flags unclear profiles, sets candidate's experience?"

manager asks them to review all feedback SLA

Answers to situational questions should be documented as assessment facts about thinking, not as proof of

a past result. If the role risk is high, combine the scenario with behavioural facts, a work

sample or reference checks.

6. Panel interviews without impression-based

voting

Panel interviews are only useful where the work genuinely requires a joint outcome: cross-functional

delivery, leadership, mentoring, client escalations, product decisions or complex context transfer.

This is not a popularity contest.

A panel interview should observe work behaviour:

Can be observed Must not be used as a criterion

Clarifying questions "Similar / not similar to us"

Ability to explain trade-offs Age, appearance, accent, family status

Conflict and disagreement handling Personal taste, hobbies, social similarity

Collaboration and listening "Energy" without behavioural definition

Learning and reflection Culture fit as a vague label

Respectful challenge Comfort with interviewer personality

Team observation card:

Criterion What to observe Notes with Assessment Confidence

assessment facts rating

Collaboration Invites input,

clarifies roles, builds on

others' facts

Conflict maturity Disagrees with reasoning, not

status games

Communication Summarises decision,

risks and next steps

Role-specific maturity Applies relevant domain

decision logic

Candidate questions Asks questions that reveal

realistic expectations

Facilitation rules:

1. Explain the purpose and format of the panel meeting to the candidate.

2. Give the team the criteria before the meeting.

3. Assign one facilitator; avoid chaotic panel interrogation.

4. Require notes before group discussion.

5. During the debrief, ban "vibe" unless translated into observable work behaviour.

6. If team feedback is not supported by facts, record this as unconfirmed and do not

use it as an assessment fact for rejection.

7. Interviewer roles and context handover

Every interviewer needs a role. Without one, interviewers duplicate each other or assess their favourite topic. The recruiter

owner coordinates, but the hiring manager is responsible for role criteria and the decision.

Role Responsibility Before interview After interview

Recruiter Process, candidate experience, note Sends brief, criteria, logistics Collects scorecards, updates for

owner discipline, debrief facilitation the candidate, decision log

Hiring Business need, final decision, Confirms criteria and stage Makes the decision using

owner role realism objectives assessment facts

Functional Professional skills / domain Reviews rubric and questions Submits assessment facts and

interviewer maturity confidence

Behavioural Behavioural skills and role Selects behavioural questions Records examples and scale

interviewer behaviour anchors

Team Collaboration in a realistic team Reads observation card Submits notes before group

interviewer context debrief

Bar raiser / calibration Process consistency for high- Reviews scorecard and Checks unsubstantiated

interviewer stakes roles bias risks assertions

Coordinator Planning, reminders, candidate Sends calendar and agenda Tracks SLA and missing

logistics scorecards

Context handover pack after each interview:

Field Mandatory content

Stage objective What this interview should have checked

Criteria assessed 2–4 criteria, not the whole person

Assessment facts Concrete examples, quotes only where necessary, behaviour,

outcome

Rating 1–5 or local scale with scale anchors

Confidence High / medium / low and why

Risks Work-related concerns

No assessment facts What remains unknown

Recommendation Progress / reject / pause / next action

Interviewer training checklist

Before bringing an interviewer into the process, it is important not just to send them the questions, but to ensure

they understand their role within the assessment system.

Interviewer skill What to check Minimum standard

Understands the stage objective Can explain which 2–4 criteria Does not duplicate other

they are specifically assessing interviewers' questions

Prepares questions in advance Uses the question bank and Follow-up questions, not

against criteria follow-ups improvisation based on CV

Interviewer skill	What to check	Minimum standard
Takes notes based on facts	Records behaviour, context, outcome and constraints	Does not write generic impressions like "pleasant" or "weak"
Uses the scale	Assigns a rating using 1/3/5 anchors	Does not average intuition and likeability
Stays within boundaries	Does not ask personal, discriminatory or irrelevant questions	All questions relate to the role
Respects the candidate	Maintains timing, explains the format, leaves time for questions	Does not turn the interview into a purposeless stress test
Submits assessment on time	Completes the scorecard before the debrief	Does not influence others before their own rating is recorded
Can argue in the debrief	Separates facts, interpretations and recommendations	Does not defend a "feeling" without assessment facts
8. How to implement a structured interview process
Week	Actions	Owner	Output
1	Select 2 pilot roles; list criteria and risks	Recruitment lead + hiring manager	Pilot scorecard
1	Map the current interview cycle and remove duplicate stages	Recruiter owner	Cycle map
2	Assign interviewer roles and stage objectives	Hiring owner	Interviewer brief
2	Write 3–5 questions per stage with follow-ups and scale anchors	Recruiter + interviewers	Question bank
3	Run interviewer training on assessment fact notes, assessment fairness and debrief rules	Recruitment lead	Training notes
3	Launch scorecard completion before the debrief	Recruiter owner	Scorecard discipline
4	Review pass-through, scorecard completion, feedback delays, candidate drop-offs	Recruitment lead	Pilot report
4	Approve SOP, QA sample review and monthly calibration	HRD / lead	Interview process SOP

9. How to measure interview quality and

assessment fairness

Metric Formula What it shows Risk of misuse

Interview pass-through Candidates passing on / candidates Low pass-through does not always

passing the stage per stage Stage strictness and fit mean poor quality; the stage may be

intentionally selective

Scorecard completion Completed scorecards / completed Assessment fact discipline Filling in the scorecard formally

interviews after the decision

Assessment fact completeness Scorecards with assessment facts Quality of notes Counting long notes as good

for each rated criterion / all for relevance

scorecards

Feedback latency Date feedback submitted Candidate experience and Update status without useful

• interview date feedback

Duplicate questions share Repeated same criterion across Cycle design quality Remove repetition that is

stages / all criteria intentionally confirming high-risk

criteria

Unconfirmed feedback share Vague comments without Assessment fairness and Punishing interviewers instead

assessment facts / selective decision quality of training

comments

Candidate drop-off after Candidate departures after Process length, communication, Blaming the candidate instead of

interviews interview / candidates after role realism fixing the process

interview

Rating variance by Average rating variance by Need for calibration Compare across different roles

interviewer interviewer and role without context

Time to decision Decision date – final interview Debrief discipline Speed without quality

Qualitative QA matters. Review a sample of scorecards and ask:

Did the interviewer assess the assigned criteria?

Are the assessment facts concrete and work-related?

Have unsubstantiated phrases been removed or rewritten?

Did the debrief happen after individual notes?

Were candidate updates sent within SLA?

Are adverse patterns visible by source, stage or interviewer?

10. In HarmonyATS

In HarmonyATS: interview stages, scorecards, notes and decision logs can sit on the candidate

card. This helps the debrief use assessment facts instead of recollections and keeps stage progression

linked to rationale.

In HarmonyATS: the Funnel with detail, Rejections and SLA reports show where interviews create

bottlenecks: long feedback delays, high drop-off after a weak previous stage or

candidates stuck waiting for interviewer notes.

In HarmonyATS: AI Studio can draft questions from approved criteria, and

AI-Matching can suggest hypotheses for review. Interviewers remain the owners of questions, assessment

facts, ratings and decisions.

In HarmonyATS: the candidate movement report can serve as a factual meeting summary: who

moved, who is waiting, which interviewers owe feedback, and whether the process meets SLA.

These reports alone do not prove interview quality. They help run the debrief, provided there are

sample scorecards, candidate feedback and calibration with the hiring manager alongside them.

End-to-end example: interview cycle

Stage BDM in B2B SaaS payments Backend Engineer

Recruiter Must-have criteria: baseline level, motivation, Must-have criteria: baseline level, motivation,

screening compensation constraints compensation constraints

Hiring manager / Commercial responsibility, B2B SaaS context, Backend fundamentals, work examples,

technical screen segment thinking

Case / work sample Discovery role play: problem, urgency, purchase Code / work sample review: code quality, tests,

process, next step reasoning

Domain / system Payment scenario: payment flow, financial or System design: recurring payment retry, webhook,

deep dive legal stakeholder, integration risk reconciliation, observability

Cross-functional Communication with product, finance and legal Collaboration with product/ops/security and working

interview with trade-offs

Final debrief Scorecard decision, risks, offer readiness Scorecard decision, level, offer readiness

11. Case studies: typical situations and actions

Situation Risk What to check Recommended How to document Metric / next

action action

Interviewer writes "not our vibe" Bias, unconfirmed Ask the interviewer to "Unconfirmed comment Share of unconfirmed

rejection; which affected job translate into observable removed; no criterion-level assessment fact in feedback

performance; which criterion; what assessment facts in decision

assessment facts if none, remove or rewritten

All interviewers ask the Candidate fatigue, Interview plan, role Redesign cycle; assign "Duplicate responsibility Share of duplicate

same question no new signals assignment, scorecard criteria to stage; ownership moved to

panel interview observes

HM only; overlapping

collaboration"

Panel interview turns into Similarity bias, Observation card, Facilitate criterion-based "Team feedback limited to Team scorecard

popularity vote assessment fairness individual notes before discussion; no voting collaboration, conflict

discussion without assessment facts maturity, communication

assessment facts"

Candidate gives polished Charisma bias Authorship, scale, Use follow-up "Assessment facts low First-stage false

stories without detail outcome, artefacts, questions; lower confidence: no concrete positives

consistency role/outcome after follow-up role/outcome; ask

questions targeted next action

Situation	Risk	What to check	Recommended action	How to document	Metric / next action
Hiring manager changes criteria after interview	Process unfairness, changing criteria after the start	Was the criterion approved in advance; have business parameters changed; which candidates were affected	Pause and recalibrate; decide whether to apply the change only to new candidates; record the parameter change	"New criterion added after parameter change; not applied retrospectively to past decisions without review"	Criteria changes after launch
Interviewer delays scorecard 5 days	Memory decay, candidate experience risk	Calendar load, reminders, SLA, owner	Require scorecard before debrief; escalate repeated delays	"Feedback overdue by 3 days; candidate update sent after deadline"	Feedback latency, SLA breach
Candidate needs accommodation or alternative format	Assessment fairness and accessibility risk	Company policy, work-related requirements, assessment alternatives	Offer reasonable process alternative through policy owner	"Alternative interview format used; criteria unchanged"	Candidate drop-off, accommodation handling
AI suggests interview questions about personal life	Legal and assessment fairness risk	Prompt inputs, criteria, AI governance	Reject questions; update prompt restrictions; human approves question bank	"AI question removed as non-work-related; approved criteria-only bank"	AI prompt QA issues
Agency client gives vague feedback after interview	No learning, poor shortlist calibration	Client criteria, notes, rationale, assessment facts	Require feedback in "criterion – assessment fact – decision" format; pause new shortlists if feedback is unusable	"Client feedback requested against scorecard criteria; vague comment not accepted as final input"	Client feedback completeness

12. SOP and procedure snippets

SOP: structured interview process

Field Rule

Purpose Gather comparable facts related to role tasks through

structured interviews and honest context handover for

decision-making.

Scope All roles with two or more interview stages; for roles with one

interview a simplified version is used.

Field	Rule
Owners	The hiring manager owns the criteria and decision. The recruiter owns the process, candidate experience and debrief discipline. Interviewers own notes and assessment facts for their assigned criteria.
Inputs	Approved criteria, scorecard, interview cycle, interviewer brief, candidate materials as permitted by policy, privacy/AI rules.
Steps	1. Link criteria to stages. 2. Assign interviewer roles. 3. Prepare questions and follow-ups. 4. Brief the candidate and interviewers. 5. Conduct interviews. 6. Complete the scorecard before the debrief. 7. Run the debrief on assessment facts. 8. Record the decision and update the candidate.
Mandatory fields	Stage objective, criteria assessed, assessment facts, rating, confidence, risks, missing data, recommendation, next action.
Prohibited fields	Vague feedback about "vibe", protected characteristics, stereotypes, unsubstantiated personality labels, new hidden criteria discovered after the interview.
SLA	Scorecard completed by end of the next working day or before the debrief, whichever is earlier. Candidate update follows the process SLA.
Outputs	Progress, reject, pause for clarification, prepare the offer, or recalibrate the role/process.
Exceptions	Executive search, regulated roles or accessibility adaptations may require a different format, but criteria and documentation must remain work-related.
Metrics	Scorecard completion, feedback delay, unconfirmed feedback share, duplicate question share, stage pass-through, candidate drop-off.
Procedure snippet for the corporate knowledge base

13. Interview process maturity levels

Level What it looks like Data Key risk Next

step

Chaos Every interviewer asks what they Bias, repeated questions, slow Create a scorecard

likes; decision by memory from decisions

chat calendar events and

impressions

Managed There is a cycle, roles, scorecards Stage notes, ratings, decisions Scorecards filled formally Train notes with

and debrief assessment facts and debrief facilitation

Level	What it looks like	Data	Key risk	Next step
Data-driven	Team reviews pass-through, latency, unconfirmed feedback, rating variance	ATS reports and QA samples	Metrics used without context	Monthly calibration and sample review
Candidate-oriented	Process is short, clear, respectful, with timely updates	Candidate feedback and drop-out	Candidate experience treated separately from assessment	Link communication SLA to decision quality
Fair and calibrated	Criteria are stable, interviewers are trained, vague feedback removed	Calibration notes, scorecard QA	Over-standardisation without role nuance	Role-specific scenarios and scales and periodic review
AI-assisted operations	AI drafts questions and summaries, humans verify	AI reviews notes, unsubstantiated assertions removed	Automation bias	AI governance and audit trail
What to take from this chapter / where to go next
What you need	Where to find it
Terms and short definitions	Appendix B. Mini-glossary
Ready-made procedures and SLAs	Appendix D. Ready-made procedures and processes
Working templates	Appendix E. Template library
End-to-end BDM and Backend examples	Appendix F. End-to-end case studies for BDM and Backend
Full AI prompts for the chapter	Appendix H. AI prompt library, section H4. Role, criteria and assessment

What to take from this chapter / where to go next

What you need Where to find it

Terms and short definitions Appendix B. Mini-glossary

Ready-made procedures and SLAs Appendix D. Ready-made procedures and processes

Working templates Appendix E. Template library

End-to-end BDM and Backend examples Appendix F. End-to-end case studies for BDM and Backend

Full AI prompts for the chapter Appendix H. AI prompt library, section H4. Role, cri-

teria and assessment

In short: this chapter covers the interview process. If you need a ready-made artifact, take it from the appendices, and

keep the application logic in the chapter.

Chapter 6. What exactly do we assess

in a candidate

How to use this chapter

Working scenario. At the BDM debrief everyone quickly agrees: the candidate is confident, energetic, "definitely knows

how to sell". Ten minutes later it emerges that nobody checked the key thing: can they actually run discovery in B2B

SaaS payments, distinguish the budget holder from an ordinary user, and avoid starting a presentation before understanding

the client's pain. In the neighbouring vacancy, a Backend Engineer speaks quietly and doesn't sell themselves, but calmly

explains where the system needs retries, idempotency and monitoring. If the assessment rests on impressions, the team

may choose charisma over signal and miss strong engineering maturity.

What actually went wrong here: the team discusses people using words that sound confident, but do not lead

to a decision. "Strong", "not senior", "lacks energy", "not our vibe" do not explain what the person actually did, what

outcome they created and how much that can be trusted.

This chapter helps turn assessment from an argument about impressions into a clear chain: criterion → verification

method → assessment fact → confidence → decision. After reading it, it should be easier to train interviewers, defend

decisions, reduce bias, use AI carefully and see where the process produces strong signals versus where the team

is simply believing a polished story.

Remember in one phrase: an assessment fact is only stronger than an impression when it is linked to the work.

If your task is	Go to section	What you will get
Quickly explain what we assess	1	Model: professional skills / behavioural skills / motivation / signal reliability
Create criteria for a role	2	Worksheet: from business outcome to scorecard
Choose an assessment method	3	Methods table and applicability
Describe professional skills	4	Domain matrices and scales
Assess behavioural skills	5	Twelve detailed skill cards
Implement a scorecard	6–7	Templates and SOP
Work through a contentious case	8	Decision tables
Use AI	9–10	Prompts and human oversight

How to read a long chapter. This is the most reference-heavy chapter in the book. For a quick assessment setup, read sections 1, 2, 6 and 7, then move on to the minimum starting point and Appendix E. If you are training interviewers or building skill matrices, use sections 4–5 as an example library rather than as mandatory linear reading.

Quick chapter map

What we assess Key question Where to check What to record

Professional skills Can the person do the work Work sample, portfolio, Technical/functional Assessment facts, difficulty

at the required level? interview, case level, trade-offs,

outcome

Behavioural skills Will the person work reliably, Structured interview, Behaviour, repeatability,

with people and within the with others and within the role's behavioural questions, context, risks

conditions? panel case, references

Motivation Why does the candidate want Screening, deep-dive interview, Motivations, constraints, risks,

this role and will they stay? pre-offer condition check critical conditions

Signal reliability How real is the evidence? Follow-up questions, personal Contribution checks, Confidence, gaps, contradictions,

contribution questions, reference checks, missing assessment facts

case defence

Minimum starting point

Do not start with a large competency library. Start with a minimal scorecard.

Step	What to do	Artifact
1	Choose 4–6 criteria without which the role decision is impossible	Minimum scorecard
2	For each criterion, define 3 levels: unconfirmed, sufficient, strong	Scale anchors
3	Separate rating from confidence: a high score without reliable facts does not equal an offer	Rating + confidence
4	At the debrief, bring assessment facts, not a final "for / against" verdict	Assessment facts table

1. Assessment model: professional skills, behavioural skills, motivation and signal reliability

Candidate assessment fails not when the interviewer asks the "wrong" question. More often it fails earlier: the team has not agreed on what actually constitutes a strong candidate. One hiring manager says "we need a strong senior", another hears "an autonomous expert", the recruiter translates this as "five years of experience", and the interviewer in the meeting assesses confidence in speech. The result looks like competency assessment but actually collects impressions, habits and personal preferences. The simplest working model looks like this:

1. Criterion. What matters for the role and why is it linked to the work?

2. Verification method. Where can we see this criterion: screening, interview, case, work sample, portfolio, references?

3. Assessment fact. What did the candidate actually do, say, decide or demonstrate?

4. Confidence. How much can this fact be trusted?

5. Decision. What does the team do next: progress, pause, clarify, change level, or prepare the offer?

If even one link drops out, the team reverts to impressions. A polished story without a fact does not become evidence. A score without confidence does not become a decision. A criterion with no link to the work should not enter the assessment. The working assessment model must separate four layers:

1. Professional skills — abilities without which the person cannot perform the role's key tasks.

This is not a list of tools from the vacancy, but the ability to produce the required work outcome: running discovery, designing architecture, closing enterprise deals, building a financial model, writing working code, managing implementation, analysing the funnel. A good criterion describes the action, difficulty level, quality of outcome and the conditions in which the skill must be demonstrated.

2. Behavioural skills — observable work behaviour that affects a person's reliability in a specific environment. This is not "pleasant", "adequate" or "team-oriented", but behavioural patterns: how the candidate reports risks, argues a point, keeps commitments, clarifies expectations, handles conflict, accepts feedback, aligns stakeholders. A behavioural skill cannot be assessed by likeability. It must be translated into situations, actions and consequences.

3. Motivation — the reasons a person chooses this role and the risks that may cause them to quickly lose interest or fail to cope with the context. Motivation does not mean being "passionate". For hiring, what matters more is: do the candidate's expectations match the actual work, pace, level of uncertainty, task type, management style, compensation model and growth horizon? A strong candidate with misaligned motivation can become an expensive mistake.

reliable facts do not equal an offer

4 At the debrief, bring assessment facts, not a final Assessment facts table

"for / against" verdict

1. Assessment model: professional skills,

behavioural skills, motivation and signal

reliability

Candidate assessment fails not when the interviewer asks the "wrong" question. More often it fails earlier: the team

has not agreed on what actually constitutes a strong candidate. One hiring manager says "we need

a strong senior", another hears "an autonomous expert", the recruiter translates this as "five years of experience", and

the interviewer in the meeting assesses confidence in speech. The result looks like competency assessment but

actually collects impressions, habits and personal preferences.

The simplest working model looks like this:

1. Criterion. What matters for the role and why is it linked to the work?

2. Verification method. Where can we see this criterion: screening, interview, case, work sample, portfolio,

references?

3. Assessment fact. What did the candidate actually do, say, decide or demonstrate?

4. Confidence. How much can this fact be trusted?

5. Decision. What does the team do next: progress, pause, clarify, change level, or prepare

the offer?

If even one link drops out, the team reverts to impressions. A polished story without a fact

does not become evidence. A score without confidence does not become a decision. A criterion with no link to the work

should not enter the assessment.

The working assessment model must separate four layers:

1. Professional skills — abilities without which the person cannot perform the role's key tasks.

This is not a list of tools from the vacancy, but the ability to produce the required work outcome: running

discovery, designing architecture, closing enterprise deals, building a financial model, writing

working code, managing implementation, analysing the funnel. A good criterion describes the action,

difficulty level, quality of outcome and the conditions in which the skill must be demonstrated.

2. Behavioural skills — observable work behaviour that affects a person's reliability in

a specific environment. This is not "pleasant", "adequate" or "team-oriented", but behavioural patterns: how the candidate

reports risks, argues a point, keeps commitments, clarifies expectations, handles conflict,

accepts feedback, aligns stakeholders. A behavioural skill cannot be assessed

by likeability. It must be translated into situations, actions and consequences.

3. Motivation — the reasons a person chooses this role and the risks that may cause them to quickly

lose interest or fail to cope with the context. Motivation does not mean "passionate". For hiring, what

matters more is: do the candidate's expectations match the actual work, pace, level of uncertainty,

task type, management style, compensation model and growth horizon? A strong candidate with misaligned

motivation can become an expensive mistake.

4. Signal reliability — how much trust the received signal deserves. The same answer can be

strong evidence, weak evidence or simply unsuitable for a decision.

Consider the context: did the candidate do the work themselves or were they part of a strong team; was the outcome

measurable or described in vague terms; was the task similar to your role or merely sounds similar; did the interviewer

check the details or accept a polished story at face value.

Assessment facts are not the interviewer's opinion and not a restatement of the CV. They are verifiable observations

about the candidate's work.

A good assessment fact answers: what happened, what the task was, what role the candidate played, which

decisions they made, what they did personally, what constraints existed in the context, how the work ended and what

can be verified through a follow-up question, case, portfolio or reference. The higher the cost of error in

hiring, the less the team should rely on single impressions and the more they should rely on several independent

signals.

Vague labels are dangerous because they sound like assessment but do not lead to an actionable decision. "Strong", "senior",

"proactive", "not our vibe" do not explain what the person must actually do in the role and which behaviour the

team considers sufficient. Such words give interviewers too much freedom: each assesses their own thing,

there is nothing to debate, nothing to calibrate, it is impossible to give the candidate honest feedback, and bias is

easily disguised as "professional intuition".

Practical rule: any label must be translated into observable behaviour before the interview starts.

Vague formulation Why it is dangerous How to translate into a What assessment facts are

criterion needed

"Strong candidate" Unclear where the strength lies Professional skill / Specific example, outcome,

behavioural skill / difficulty, candidate's role

motivation / level

"Not senior" Could mean anything Complexity, What decisions they made, how

autonomy, influence, they affected the outcome, what

quality of decisions they did without oversight

"Proactive" Often substitutes for responsibility, or a Situation where the candidate

initiative, early risk reporting, action themselves identified a problem and drove

without reminders it to completion

"Not our vibe" May hide bias Communication style, Observed behaviour and link

collaboration, conflict to role tasks

maturity, pace fit

Translation examples:

Instead of "strong", ask: "In which tasks does the role demand a strong outcome: speed, quality of decisions,

depth of expertise, influence on others, resilience to uncertainty?" Then define the criterion:

"independently runs discovery with enterprise clients, identifies decision criteria, captures

implementation risks and translates them into an action plan".

Instead of "senior", break down the level along an axis: task complexity, autonomy, influence, quality of decisions,

ability to improve the system around them. Seniority for a given role is not age or years of experience, but the

ability to consistently make decisions at the right level without constant oversight.

Instead of "proactive", clarify what the business actually needs: early risk detection, independently

proposing options, driving outcomes without reminders, knowing when to escalate in time. If

this is not linked to role tasks, the criterion is better removed.

Instead of "not our vibe", record the work risk: communication pace, conflict, decision-making

style, maturity in feedback, ability to work with uncertainty. If the risk cannot be

linked to role tasks, it cannot be used as a hiring criterion.

Mini-template for calibration:

Label Observable behaviour Where we check What counts as a strong

assessment fact

Senior Makes decisions in complex Situational interview, Specific decision, alternatives,

conditions and explains trade-offs case, portfolio review consequences, candidate's role

Accountability Owns the outcome, risks and Behavioural interview, An example where the candidate

communication reference check spotted a problem, chose an

action and drove it to completion

Communication Helps others understand status, Structured interview, Clear answer structure, audience

risks and decisions role-play scenario, references adaptation, no hidden surprises

Pace fit Works at the required pace without Case, manager interview, Examples of deadlines, prioritisation,

damaging quality reference check trade-offs and recovery after

setbacks

2. From business task to measurable criteria

Assessment criteria must not start with a competency list. They must be derived from the business task of the role. Otherwise the team quickly falls into generic wishes: "analytical", "communicative", "autonomous", "learns quickly". Such words may be true, but they do not help choose between two candidates.

The correct sequence goes from outcome to verification method:

1. Business outcome. First, define what capacity or result the role must create. Not "hire a CSM" but "accelerate enterprise onboarding without losing quality", "reduce manual billing errors", "create a predictable outbound funnel", "reduce dependency on founder-led sales".

2. Key tasks. Then list the tasks that create this outcome. The important thing here is not to describe the job in its entirety, but to select 5–8 tasks where a hiring mistake would be expensive.

3. Professional skills. For each key task, define what the person must be able to do in practice. The skill must be verifiable: not "knows CRM" but "builds a funnel hygiene process and identifies the causes of conversion drops".

4. Behavioural skills. Define the work behaviour without which professional skills will not produce results in your environment. For example, with many stakeholders, what matters more than "sociability" is structured communication, expectation management and conflict maturity.

5. Motivation risks. Record why even a competent candidate may fail to stay or quickly lose effectiveness: too much uncertainty, too few established processes, a high proportion of client escalations, a slow enterprise cycle, no familiar team, a different balance of strategic versus operational work.

6. Signal reliability. For each important criterion, decide in advance what might be overstated and how to verify it: authorship of the outcome, depth of decisions, scale of the task, autonomy, data quality, genuine contribution to a team project.

7. Assessment methods. Choose methods to match the criteria, not the other way around. A professional skill can be checked with a work sample, case or technical interview; behaviour with a behavioural or structured interview; motivation with a deep-dive and pre-offer condition check; a questionable signal with a follow-up action, portfolio defence or reference check.

8. Scorecard. Assemble the criteria into a scorecard: 5–7 criteria, weights, 1/3/5 scale anchors, notes with assessment facts and confidence. The scorecard should help reach a decision, not be a formal appendix to the interview.

Ready-to-use worksheet:
Field	Question	Example completion
Business outcome	What capacity or result must the role create?	Accelerate enterprise onboarding without losing quality
Key tasks	Which tasks create this outcome?	Client diagnostics, implementation plan, stakeholder coordination
Professional skills	What must the person be able to do?	Discovery, project planning, process mapping
Behavioural skills	Which behaviour is critical in this environment?	Communication, reliability, conflict maturity
Motivation risks	Why might the candidate fail to stay?	Not prepared for client escalations or high uncertainty
Signal reliability	What might be overstated?	Genuine contribution to projects, responsibility, depth of decision-making
Methods	How do we verify?	Structured interview, case, reference check
Scorecard	How do we rate?	6 criteria, weights, 1/3/5 scale anchors, confidence
Example completion for the role of Head of Customer Implementation:
Step	Team decision
Business outcome	Reduce chaos in enterprise client implementations and make onboarding predictable
Key tasks	Diagnose the client, gather requirements, build an implementation plan, manage risks, align sales / product / support
Professional skills	Discovery, project planning, stakeholder mapping, process design, basic analytics
Behavioural skills	Structured communication, reliability, conflict maturity, expectation management
Motivation risks	Candidate wants an advisory role without operational responsibility; not prepared for difficult client conversations
Signal reliability	In the previous role may not have been the owner but a member of a strong implementation team
Methods	Structured interview, case interview, portfolio review, reference check
Scorecard	6 criteria: implementation design, client diagnosis, stakeholder management, risk responsibility, communication, motivation fit

Criterion	Weak anchor 1	Working anchor 3	Strong anchor 5
turns diagnostics into an implementation plan
Communication	Gives status irregularly or too broadly	Structures status, risks, decisions and next steps	Adapts communication to the audience, reduces uncertainty and prevents conflict before escalation

3. Assessment methods: what to choose and when

An assessment method must answer the question "what signal do we need?" One method rarely covers the whole role. An interview shows thinking structure and past behaviour well, but is weak at checking actual work execution. A work sample produces a strong work-related signal, but can be costly for the candidate and team. Reference checks help verify behavioural repeatability, but depend on the quality of questions and the source. AI-assisted review speeds up material analysis, but does not remove responsibility from people.

Method	What it checks	When to use it	Weaknesses	Cost / time	Candidate experience	Bias / fairness risks	How to implement
Structured interview	Comparable answers against pre-set criteria, hard/soft signals, reasoning depth	Almost any role, especially when multiple interviewers need a single standard	Poor at checking actual work execution without additional methods	Medium: requires question preparation, scale anchors and interviewer training	Usually perceived as fairer when questions are role-relevant and the interviewer explains the process	Risk shifts to criterion and scale design; poor scale anchors standardise poor assessment	Run job analysis, choose 4–6 criteria, write identical questions, follow-ups, 1/3/5 scale anchors and note-taking rules
Behavioural interview	Past behaviour in similar situations: responsibility, conflict maturity, decision quality, resilience	When the candidate has relevant experience and you need to understand repeatable patterns	Candidate may tell polished stories; past context may not match yours	Low–medium: quick to implement, but requires discipline on follow-up questions	Good experience if questions do not turn into interrogation and relate to the work	Interviewer may assess confidence instead of facts; cultural storytelling styles influence impressions	Use STAR/CARE logic: context, task, action, result, candidate's role, what they would repeat differently
Situational interview	How the candidate reasons in a realistic future scenario or dilemma	For new contexts, industry transitions, roles with high uncertainty	Tests intentions and reasoning, but does not prove past ability to execute	Medium: requires realistic scenarios and response criteria to be written	Can be good if the scenario is short, clear and does not require hidden knowledge	Advantage to candidates familiar with "correct" interview language	Describe a work-related scenario, give identical briefs, assess trade-offs, questions, risks and decision quality
Case interview	Analytics, problem structuring, business maturity, communication of decisions	For consulting, product, strategy, revenue, operations, implementation and leadership roles	Often turns into a puzzle if not linked to real work	Medium–high: case preparation, interviewer, review	Fair experience if the case is bounded and does not require free work for the company	May favour those who practised the format; risk of assessing presentation style over decision quality	Make the case resemble a real task, provide criteria, timebox, identical data and a rating rubric
Work sample	Ability to produce a slice of real work: a work product, solution, quality of execution	When you can honestly simulate a role task without access to internal secrets	Does not cover long-term behaviour, teamwork or motivation	Medium–high: task design, review, sometimes paid	Good experience if the task is short, transparent and does not resemble unpaid labour	Risk of unfair burden on busy candidates; risk of covert use of the result	Limit scope, remove commercially useful output, provide a rubric, deadline, context and the ability to explain decisions
Test assignment	Checking specific skills via a take-home or synchronous task	When an artefact is needed: code, text, analysis, plan, design, financial model	Take-home tasks poorly control conditions and may overburden the candidate	Medium: cheaper than an assessment centre, more expensive than an interview	Often worsens if the task is long or unpaid	Unequal conditions on time, caregiving, access to tools; risk of AI-generated output without understanding	Keep it short, work-related, with a clear time limit; assess not just the answer but the reasoning

Method	What it checks	When to use it	Weaknesses	Cost / time	Candidate experience	Bias / fairness risks	How to implement
behind the answer
Portfolio review	Quality of past work, task complexity, decision-making style, growth in level	For design, content, product, engineering, marketing, sales support and leadership roles with artefacts	Portfolio may be a team effort; outcome may depend on brand, budget or market	Low–medium: requires a prepared interviewer and rubric	Usually a good experience: candidate shows real work	Risk of assessing visual polish or company status instead of contribution	Request 2–3 artefacts, review context, candidate's role, constraints, decisions, outcome and process lessons
Technical interview	Depth of professional knowledge, reasoning, debugging, design decisions	For engineering, data, security, finance, legal, operations and other expert roles	May check trivia instead of work; live stress and whiteboard formats distort the signal	Medium: requires an expert and calibrated rubric	Can be stressful; better when tasks resemble real work	Bias towards thinking-out-loud style, language, format familiarity	Anchor questions to key tasks, define expected signals in advance, allow follow-up questions, record assessment facts
Reference check	Behavioural repeatability, reliability, boundaries, work style, risk management	After strong interviews, before an offer, or to verify contentious signals	Sources may be loyal, cautious or unable to remember details	Low–medium: 20–40 minutes per source plus preparation	Fine for the candidate if the process is transparent and consent is requested	Risk of informal checks without consent and unequal network access	Obtain consent, ask work-related questions, verify specific criteria, separate facts from opinions
Role play / assessment centre	Behaviour in an interactive situation: negotiation, escalation, coaching, stakeholder management	For sales, customer success, support, leadership, recruiting, people management	Expensive; candidate may "play a role" rather than demonstrate a consistent pattern	High: scenarios, observers, assessment, logistics	Can be a strong experience if scenarios are realistic and respectful	Risk of assessing acting confidence, accent, extroversion	Use short realistic scenarios, train observers, assess specific behaviours and outcomes
AI-assisted matching / AI-assisted review	Fast matching, CV summaries, comparing assessment facts against criteria, next-action prompts	To support sourcing, initial material review, preparing interviewer notes	AI does not know the full context, may err, amplify bias or invent confident conclusions	Low per candidate, but requires governance setup	May be invisible to the candidate; transparency is important where required by policy or law	Risk of discrimination, proxy variables, opaque vendor assessments, automation bias	Use AI output only as a hypothesis, keep human decision-making, check work-related criteria, monitor adverse impact
Scorecard / debrief / calibration	Unified decision on assessment facts: criteria, weights, confidence, interviewer disagreements	For all roles with more than one assessor or where the cost of error is high	Does not improve hiring quality if criteria are poor or the debrief turns into a status vote	Medium: requires a template, facilitation and note-taking discipline	Candidate does not see it directly, but benefits from a fairer process	Groupthink, HiPPO effect, post-hoc rationalisation, skew towards the loudest interviewer	Complete the scorecard before the debrief, discuss assessment facts by criteria, separate rating from confidence, record decision rationale

4. Professional skills: how to describe,

verify and assess

Professional skills are work abilities, not a keyword list from the vacancy. "Knows Python", "worked with

HubSpot", "5 years in B2B SaaS" or "has enterprise experience" do not in themselves answer the key question: can the

person do the required work at the required level of difficulty and quality.

A good professional skill criterion describes four things:

1. Work output. What work result must the person produce: a debug plan, an architectural solution,

positioning, a financial model, qualification notes, an interview rubric.

2. Complexity. Under what conditions the skill must operate: ambiguous requirements, stakeholder

conflict, incomplete data, limited budget, high cost of error, tight deadlines.

3. Assessment facts. How we verify the skill: a past artefact, case review, work sample, technical

interview, portfolio defence, reference check.

4. Acceptable performance level. What counts as "good enough" for this role: not a perfect answer, but

a minimum level of autonomy, reasoning quality, accuracy and applicability of the solution.

Years of experience are a weak proxy for assessment facts. They may suggest where to look for examples, but they do not prove level.

One candidate in two years may have run complex tasks autonomously; another in seven years may have repeated a narrow

slice of the process without responsibility. Therefore, the scorecard should record not "5+ years of experience" but

"independently diagnoses problems, chooses a verification method, explains trade-offs and demonstrates

outcomes on data".

IT / Product / Engineering

Skill Simple Assessment facts Strong signal Weak signal Question / case 1/3/5 scale Rating error

definition anchors

Debugging Finds the cause of a Incident review, Builds hypotheses, Tries random "The service started 1: guesses Chaotic evaluation, assessing

technical or product code review, isolates permutations, variables, blames started giving errors randomly. 3: confidence or specific stack

problem and drives technical interview, tests logs and data, external factors, proposes a reason instead of quality of diagnosis

to a verified root cause postmortem cannot explain cannot explain the order of reasonable plan, diagnosis

corrective action preventive action causes, captures priority checks but misses plan, but the root but misses the systematic

the root cause within the first 30 diagnosis. 5: diagnosis itself

fix and prevention minutes?" systematically isolates the

problem, assesses

impact, proposes a

fix and prevention

System design Designs a system to Architecture review, Clearly describes Requirements, Draws trendy "Design a 1: a set of technologies Confusing maturity level with

handle load, constraints, design case, past limits, data flows, components with no notification for a without an architecture. quantity of named

reliability and scale schemes, next action for rejection scenarios link to the task, B2B product using 3: working schema for technologies

trade-offs failure modes, does not discuss failure resilience and SLA" basic scenarios. 5:

operational risks and escalation path, mitigation options justified design with

compromises observability trade-offs, risks,

observability and

migration path

Product discovery Uncovers real user Discovery notes, Separates problem Problem, solution, "Customers want a 1: collects wishes as Easuring hype and "product

problems and translates user interviews, from solution, tests backlog. 3: a new dashboard. language" instead of quality

them into product opportunity tree, assumptions, uncovers guiding How do we know it's assessment facts

solutions PRD, product case, segments what's actually needed? really needed? What

validates willingness to builds?" 5: validates

users, links insights to change, the problem, assesses

decisions formulates next experiment impact, selects

Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Technical trade-off analysis	Compares solution options against cost, risk, speed and future support	Design doc, RFC, incident decision log, case interview	Names alternatives, selection criteria, reversibility, short-term and long-term consequences	Defends one favourite option, ignores ownership cost or team constraints	"We need to ship an integration quickly: buy an off-the-shelf solution or build in-house?"	1: chooses by preference. 3: compares obvious pros and cons. 5: links choice to business risk, deadlines, support and review plan	Taking a complex answer for a strong one if there are no clear criteria and decision logic
Marketing
Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Audience research	Understands the audience through real segments, purchase context and proven pains	Research notes, customer interviews, survey design, sales call analysis	Identifies segments, jobs, triggers, objections, buying committee and fact sources	Describes the audience by demographics or vague terms, does not distinguish user and buyer	"We need to understand why SMB clients drop off before demo. How would you research this?"	1: builds a persona from assumptions. 3: collects data and interviews. 5: links insights to messaging, channels and funnel decisions	Believing attractive personas without checking data and sales
Positioning	Defines who the product is for, what problem it solves and why it should be trusted	Positioning doc, landing copy, sales deck, win-loss analysis	Links ICP, category, alternatives, differentiation, proof and buying trigger	Writes generic value propositions, copies competitor language, does not show proof	"The product looks like several competitors. How would you formulate positioning for a new segment?"	1: a set of trendy words. 3: a clear formulation for a segment. 5: positioning differentiates the product, links to proof and applies in sales	Assessing writing style instead of strategic accuracy
Performance analysis	Finds causes of changes in marketing metrics and selects actions	Funnel dashboard, campaign analysis, cohort report, case interview	Separates volume, conversion, mix, attribution, lag, seasonality and data quality	Looks only at top-line CAC or leads, draws conclusions without base rate and segments	"Leads grew by 40%, the funnel did not grow. How would you investigate?"	1: draws one conclusion from one metric. 3: analyses funnel stages. 5: checks segments, data quality, attribution and next actions	Confusing knowledge of a dashboard tool with analytical thinking
Experiment design	Designs hypothesis testing so the result is interpretable	Experiment brief, A/B test review, campaign plan, growth case	Formulates hypothesis, metric, sample size, duration, constraints and decision rule	Launches many activities without control logic and stop criteria	"We want to test a new outbound offer. How would you design the experiment?"	1: suggests just launching. 3: sets a hypothesis and metric. 5: describes design, risks, decision rule and next action	Counting any change as an experiment
Sales / Customer Success
Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Discovery	Uncovers the client's business problem, decision criteria, stakeholders and urgency	Call recording, role-play scenario, account notes, deal review	Asks about pain, impact, current process, buying process, risks and next step	Presents the product before understanding the client, accepts the first answer as truth	"The client says: we need report automation. Run the first 10 minutes of discovery."	1: pitches the product. 3: uncovers the problem and basic context. 5: reveals impact, stakeholders, criteria and next step	Assessing talk time and friendliness instead of diagnostic depth
Qualification	Determines whether to continue a deal or implementation and which risks are critical	CRM notes, funnel review, MEDDICC/BANT adaptation, role-play scenario	Checks fit, budget logic, authority, timeline, pain severity, blockers and mutual plan	Updates statuses formally, does not distinguish interest from buying intent	"The funnel has many active deals with no next step. How would you qualify priorities?"	1: treats all opportunities as equal. 3: identifies obvious fit. 5: honestly disqualifies weak deals and explains risk-adjusted forecast	Rewarding optimism bias for a "big funnel"

Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Negotiation	Reaches agreement without destroying value, margin and relationships	Deal review, pricing exception notes, role-play scenario, reference check	Discovers interests, anchors value, manages concessions, captures mutual commitments	Quickly gives a discount, argues positionally, does not understand BATNA and alignment process	"The client demands a 25% discount before the quarter ends. How do you respond?"	1: concedes or pressures. 3: trades with partial logic. 5: defends value, seeks trade-offs and preserves the next step	Taking aggression for negotiation skill
Client risk diagnostics	Early warning of churn risk, failed onboarding or dissatisfaction; translates it into a plan	Health score review, escalation case, QBR notes, implementation retrospective	Separates adoption, value realisation, stakeholder change, support issues and commercial risk	Reacts only after a complaint, confuses client mood with churn risk	"The client rarely logs into the product, but on calls says everything is fine. What would you check?"	1: waits for renewal. 3: checks usage and contact. 5: diagnoses risk causes, builds a recovery plan and ownership	Assessing empathy without the ability to change outcomes
Finance / Operations
Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Financial modelling	Builds a model that supports a decision, not just one that calculates prettily	Model review, case, board pack, forecast variance analysis	Makes clear assumptions, drivers, scenarios, sensitivity and checks	Complex spreadsheet with no logic, hardcoded numbers, no error checks	"We need to assess the payback period for a new sales team. How would you build the model?"	1: a static spreadsheet. 3: model with drivers and a baseline forecast. 5: a model ready for a decision, with scenarios, risks and sensitivity analysis	Assessing Excel speed instead of quality of management insight
Risk analysis	Sees financial, operational and compliance risks before they become incidents	Risk register, audit notes, control design, case interview	Assesses likelihood, impact, controls, owner, residual risk and triggers	Lists risks without prioritisation or action plan	"The company is entering a new market. Which risks would you check before launch?"	1: a general list of worries. 3: groups risks and actions. 5: prioritises, assigns controls and monitoring	Taking pessimism for mature risk management
Process control	Makes the process stable, measurable and protected against repeated errors	SOP, control checklist, SLA dashboard, incident review	Defines inputs, context handover, controls, exceptions, metrics and escalation path	Creates instructions nobody uses, does not see failure points	"Billing regularly has manual errors. How would you stabilise the process?"	1: asks people to be more careful. 3: describes the process and checks. 5: changes control points, ownership, metrics and the feedback loop	Assessing the existence of a procedure instead of actual error reduction
Operational prioritisation	Selects which operational problems to solve first under limited resources	Ops roadmap, incident backlog, prioritisation case, metrics review	Compares impact, urgency, effort, reversibility, dependencies and customer/business risk	Takes the loudest requests or personal stakeholder preferences	"There are 20 process problems and one ops manager. How would you pick the first 3?"	1: sorts by noise. 3: uses impact/effort. 5: considers risk, dependencies, capacity and review cadence	Confusing busyness and reaction speed with prioritisation
Recruitment / HR
Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Intake translation into criteria	Translates a hiring manager's request into work-related criteria, a scorecard and search strategy	Intake notes, scorecard draft, calibration doc, role kickoff case	Clarifies business outcome, key tasks, assessment facts for must-have criteria, trainable gaps and stop factors	Records wishes as a keyword list, does not separate nice-to-have criteria from must-have criteria	"The hiring manager says: we need a strong senior marketer. How would you run the intake?"	1: accepts the brief as-is. 3: clarifies tasks and profile. 5: translates the request into criteria, assessment facts and calibration	Counting a long list of requirements as a good intake

Skill	Simple definition	Assessment facts	Strong signal	Weak signal	Question / case	1/3/5 scale anchors	Rating error
Structured screening	Runs a first-pass selection against identical work-related criteria and assessment facts	Screening rubric, notes, call recordings, funnel audit	Asks a consistent baseline of questions, records facts, separates gaps from rejections, checks motivation risks	Assesses by impression, changes criteria from candidate to candidate	"For the role there are 80 candidates. How would you structure screening to avoid losing strong ones?"	1: sorts by CV and feelings. 3: applies a basic rubric. 5: preserves comparability, notes, assessment fairness and calibration loop	Confusing screening speed with signal quality
Interview design	Creates interviews that check specific skills, not just fill the calendar	Interview plan, question bank, scale anchors, interviewer briefing	Links questions to criteria, writes follow-up questions, 1/3/5 scale anchors, distributes areas between interviewers	Gives all interviewers one generic questionnaire without scales or assessment fact notes	"We need to assess a Head of Customer Implementation. How would you design the interview cycle?"	1: a set of generic chats. 3: has stages and questions. 5: each stage checks its own risk, has scenarios and scales and debrief rules	Counting the number of interviews as assessment quality
Recruiting analytics interpretation	Reads funnel data and understands where the problem is: process, market or criteria	Funnel report, source analysis, conversion review, hiring SLA dashboard	Separates source quality, stage conversion, interviewer behaviour, compensation, timeline and role clarity	Draws conclusions from a small sample, blames the market without checking bottlenecks	"The funnel is full but offer acceptance is low. How would you find the cause?"	1: looks at one metric. 3: analyses stages. 5: checks segments, sample size, causes, process defects and next actions	Confusing reporting with interpretation

5. Behavioural skills: a detailed assessment

matrix

Behavioural skills are not a matter of personal taste, candidate "likeability" or similarity to the team. In hiring

assessment they must be work-related criteria just like professional skills: linked to the role's tasks, observable

in behaviour and verifiable through assessment facts. If a criterion cannot be described through a work

situation, question, strong signal and weak signal, it cannot be honestly used in a hiring decision.

The same soft skill looks different across environments. In a start-up, what matters is autonomy, decision

maturity and the ability to learn without ready-made instructions. In a regulated process, what matters is

accountability, reliability and ethical behaviour, because an error can create legal, financial or safety

risk. In client-facing work, communication, adaptability and stress tolerance are more prominent: the candidate must

maintain trust when conditions change. In leadership roles, behavioural skills affect not just personal

effectiveness, but the environment, decisions and behaviour of others.

The term "cultural fit" must be translated into work behaviour, not similarity. You must not assess

whether a candidate resembles the team in communication style, background, interests or "vibe". You can assess

how they accept feedback, handle conflict, keep commitments, report risks,

make decisions with incomplete information and act within ethical boundaries. This makes criteria fairer,

more useful and protected from bias.

Communication

Block Content

Simple definition The ability to convey information, expectations, risks and

decisions clearly so that others can act.

Why it matters Poor communication creates rework, mistrust,

missed risks and wrong decisions even with strong

professional skills.

Where it shows up In statuses, meetings, emails, context handover between teams,

feedback, client calls, escalation and debrief.

What to ask "Tell me about a time when you had to explain a complex

problem to someone without your expert context. What

did you say, how did you verify understanding and how did it end?"

Strong signals Structures the message for the audience, separates facts from

conclusions, names risks and next steps, checks understanding.

Weak signals Gives lots of detail without a conclusion, avoids bad news, assumes

"it was obvious", does not adapt style to the recipient.

Red flags Withholds critical information, blames others for not

understanding, uses communication as pressure or manipulation.

Scale 1/3/5 1 = communicates chaotically, without a recipient and action; 3 = clearly

conveys key facts, but does not always capture risks and the

next step; 5 = proactively chooses the format, clearly states the

decision, risk, owner and next step for different audiences.

Rating errors Confusing communication with charisma, extroversion, fast

speech or good English where it is not work-related.

Mini-case Situation: a project is delayed by a week; risk: the team

finds out late; how to verify: ask the candidate to write a short

status for the manager and a neighbouring team; decision:

assess clarity of cause, impact, options and next step.

Accountability

Block Content

Simple definition The ability to take responsibility for one's area, the consequences

of decisions and the timely reporting of risks.

Why it matters An accountable person does not guarantee the absence of errors,

but reduces the likelihood of hidden problems and repeated failures.

Where it shows up In deadlines, promises, escalations, post-mortems, handling

errors, context handover and completing mandatory procedures.

What to ask "Tell me about a work error you were responsible for. When did

you realise the problem, who did you warn, what did you do and what did you

change afterwards?"

Block	Content
Strong signals	Acknowledges their part of the responsibility, quickly reports impact, proposes a fix and prevention, does not hide the problem.
Weak signals	Only talks about external causes, recalls the error without lessons learned, cannot name a preventive action.
Red flags	Shifts blame, distorts facts, hides problems until the last moment, breaks mandatory rules for speed.
Scale 1/3/5	1 = avoids responsibility and explains problems by external factors; 3 = takes responsibility for obvious tasks but escalates risks late; 5 = tracks their own commitments, raises risks early, fixes consequences and changes the process.
Rating errors	Counting someone as accountable merely because they work a lot or are always available, without assessment facts on accountability and quality of follow-through.
Mini-case	Situation: the candidate promised a report by Friday, but the data is incomplete; risk: the decision will be based on faulty data; how to verify: ask for an action plan; decision: a strong answer includes an early warning, options, impact and a new commitment.
Ability to learn
Block	Content
Simple definition	The ability to quickly pick up new skills, verify understanding in practice and adjust approach based on feedback and assessment facts.
Why it matters	In roles with change, the person cannot rely solely on past experience; the value they create comes from the speed of quality growth.
Where it shows up	In onboarding, mastering new tools, changing markets, new tasks, feedback cycles and error reviews.
What to ask	"Give an example of a skill you had to learn quickly for a role. How did you identify what to study, how did you practise and how did you measure progress?"
Strong signals	Identifies a gap, builds a learning plan, seeks feedback, applies knowledge to tasks, demonstrates a change in outcome.
Weak signals	Learns only through passive reading, does not test themselves in practice, cannot explain what changed in their behaviour.
Red flags	Rejects feedback, repeats errors, treats learning as the company's duty without personal accountability.
Scale 1/3/5	1 = does not see gaps and defends the old way of working; 3 = learns when the task is clear and with external support; 5 = independently diagnoses gaps, experiments quickly, asks for feedback and transfers lessons to new situations.
Rating errors	Confusing the ability to learn with the number of courses, certificates or general curiosity without application to the work.
Mini-case	Situation: you need to master a new CRM process within two weeks; risk: errors in funnel data; how to verify: ask for a learning plan; decision: assess gap diagnosis, practice loop, quality checks and willingness to ask for help.