The Recruitment Metrics That Actually Help You Run an Agency

Recruitment produces a large amount of activity data.

A CRM can count:

• Calls made
• Emails sent
• Candidates sourced
• CVs submitted
• Interviews arranged
• Jobs opened
• Offers made
• Placements completed

Those numbers can create an impression of control. They can also create a busy dashboard that tells managers very little about what should happen next.

A useful metric needs four things:

1. A clear definition
2. A business question it helps answer
3. Someone responsible for acting on it
4. Enough context to interpret it properly

Without those, a metric is simply a number.

The CIPD distinguishes between measures of recruitment efficiency, such as cost per recruit, and measures of effectiveness, such as early new-hire failure or whether new employees perform successfully. It also warns that dashboards containing too many measures can make it harder for managers to identify what actually deserves attention.

A recruitment agency needs both efficiency and effectiveness.

It should know how quickly and economically it operates. It should also know whether it is filling the right jobs, introducing credible candidates, protecting margin and creating outcomes that clients and candidates value.

No metric is useful until it is defined

Two agencies can report the same metric name while calculating entirely different things.

Consider "CV-to-interview ratio."

Does "CV" mean:

• Every document emailed to a client?
• Every candidate submitted through the CRM?
• Every unique candidate submitted?
• Every candidate the client agreed to consider?
• Only submissions to qualified live vacancies?

Does "interview" mean:

• An interview requested?
• An interview booked?
• An interview attended?
• A second-stage interview?
• Any interaction between candidate and client?

A ratio is not comparable until those definitions are agreed.

Every important metric should therefore have a data dictionary containing:

• Name
• Purpose
• Formula
• Numerator
• Denominator
• Start and end points
• Included records
• Excluded records
• Reporting period
• Owner
• Frequency of review

For example:

That is much more useful than:

Interviews divided by CVs.

Use cohorts rather than mixing unfinished work with completed work

Many recruitment reports produce misleading results because they compare events occurring in the same calendar month rather than work belonging to the same group of jobs or candidates.

Suppose an agency reports in June:

• 40 jobs opened
• 10 placements made

That does not necessarily mean the June fill rate was 25%.

Some of the placements may relate to jobs opened in April or May. Many of the June jobs may still be active and could be filled later.

A more defensible fill-rate calculation uses a closed cohort:

Fill rate = jobs filled ÷ jobs closed during the period

The closed jobs include both:

• Jobs closed as filled
• Jobs closed without a placement

Alternatively, an agency can measure the eventual outcome of all jobs opened during a particular month, but it must wait until those jobs have had enough time to conclude.

The same principle applies to candidate conversion.

Do not divide all June interviews by all June submissions when the interviews arose from submissions made in earlier months unless that is the specific operational question being asked.

Cohort reporting takes more discipline, but it prevents the denominator changing underneath the metric.

Segment before drawing conclusions

An overall agency ratio can hide meaningful differences.

Conversion and speed are affected by:

• Permanent versus temporary recruitment
• Contingent versus retained assignments
• Seniority
• Sector
• Geography
• Salary or rate
• Client
• Hiring manager
• Consultant
• Source
• Exclusivity
• Working arrangement
• Length of relationship
• Difficulty of the role

A 20% submission-to-interview rate may be poor for an exclusive, well-briefed role with an established client. It may be reasonable for a speculative introduction into a highly competitive market.

Likewise, an executive search lasting 90 days should not automatically be treated as less effective than a temporary placement completed in 24 hours.

Report the overall number, but allow it to be broken down into comparable groups.

Begin with commercial outcomes

Recruitment agencies exist to create profitable placements, not merely recruitment activity.

The most important commercial measures normally include:

• Revenue
• Gross profit or contribution
• Cash collected
• Average permanent fee
• Temporary or contract margin
• Placements
• Repeat business
• Client concentration
• Rebates, replacements and credit notes
• Cost to serve

These are lagging indicators: by the time they change, much of the work has already happened.

They are still essential because they show whether the agency's activity is producing a sustainable commercial result.

Revenue per consultant needs context

Revenue per consultant is often presented as the clearest productivity metric.

It can be useful, but it has limitations.

For permanent recruitment, invoiced fees may be reasonably close to the revenue created by the consultant, subject to rebates, shared credit and collection.

For temporary recruitment, headline revenue can include substantial amounts passed through to workers, umbrellas and employment costs. Gross profit or contribution may therefore provide a more useful view than turnover alone.

The metric can also become misleading when comparing:

• A new starter with an established consultant
• A delivery consultant with a full-desk recruiter
• A manager with individual contributors
• Permanent and temporary consultants
• Consultants working different markets
• A person building a new desk with someone managing inherited accounts

A better productivity view may include:

• Gross profit per fee-earning consultant
• Gross profit per full-time equivalent
• Gross profit against salary and direct cost
• Placements per consultant
• Average fee or contribution per placement
• Repeat-client gross profit
• Gross profit collected rather than merely invoiced

The metric should support resource decisions without pretending that every role in the agency has the same commercial function.

Measure gross profit, not just billings

High billings do not always mean high profitability.

A consultant may generate substantial revenue while also:

• Working low-margin vacancies
• Giving frequent fee discounts
• Producing replacements
• Incurring high advertising costs
• Requiring extensive delivery support
• Serving clients that pay slowly
• Sharing significant commission credit
• Working temporary placements with high pass-through costs

A useful profitability view distinguishes between:

• Revenue invoiced
• Gross profit
• Direct delivery costs
• Credit notes
• Rebates
• Bad debt
• Cash collected

Not every agency will calculate placement-level profitability in the same way. The important point is to agree a consistent definition and avoid using top-line revenue as a substitute for economic value.

Fill rate: are you converting the jobs you accept?

Fill rate answers:

Of the vacancies we agreed to work, how many produced a placement?

A basic formula is:

Fill rate = jobs filled ÷ jobs closed

But the result depends heavily on what counts as a job.

Exclude:

• Duplicate vacancies
• Test records
• Jobs created only for speculative marketing
• Unqualified leads incorrectly recorded as live jobs
• Multiple records representing the same requirement

Decide how to treat:

• Multi-hire vacancies
• Jobs placed by another agency
• Roles withdrawn by the client
• Roles filled internally
• Hiring freezes
• Roles closed because the salary was unworkable
• Retained searches that changed substantially
• Temporary bookings with several workers

The reasons for non-fill are often more useful than the headline rate.

A low fill rate may reflect:

• Weak delivery
• Poor client qualification
• Unrealistic briefs
• Non-exclusive work
• Slow client feedback
• Low salaries
• Repeated withdrawal
• Too many speculative jobs
• Inadequate candidate supply
• Consultants retaining dead vacancies in the pipeline

The action depends on the reason.

A delivery problem requires a different response from a client-selection problem.

Submission-to-interview conversion

This metric asks:

Of the candidates formally presented to clients, how many progressed to interview?

A low ratio can indicate:

• Weak understanding of the brief
• Poor candidate qualification
• CVs sent too early
• Inadequate explanation of relevance
• A changing client requirement
• An unrealistic vacancy
• Client disengagement
• Duplicate or speculative submissions
• A mismatch between the consultant and hiring manager's interpretation

It does not prove that the recruiter selected poor candidates.

For example, a client that changes the criteria after receiving submissions can reduce the conversion rate even when the original submissions were reasonable.

The metric should therefore be reviewed by:

• Client
• Hiring manager
• Job
• Consultant
• Vacancy type
• Source
• Reason for rejection

Count unique candidates rather than every version of a CV or every resubmission.

Interview-to-offer conversion

This measures how often candidates who interview receive an offer.

A low conversion rate may point to:

• Candidates being submitted before full qualification
• Weak interview preparation
• Poorly defined criteria
• Inconsistent assessment
• A client interviewing candidates it was never likely to hire
• Salary or working-arrangement problems emerging late
• Candidate information being represented inaccurately
• A change in the brief
• Too many interview stages
• Strong competing candidates

Again, the rate alone does not identify the cause.

Review the documented reasons candidates were rejected after interview. If the reasons are mostly blank, vague or inconsistent, the first problem is the quality of the data.

Offer acceptance rate

A basic formula is:

Offer acceptance rate = accepted offers ÷ formal offers made

This metric can reveal:

• Poor management of expectations
• Uncompetitive compensation
• Delays between final interview and offer
• Misalignment over flexibility
• Weak understanding of candidate motivation
• Counter-offer risk
• Changes to the package
• Poor candidate experience
• Competing processes not being managed

It should also distinguish between:

• Verbal offers
• Written offers
• Conditional offers
• Accepted offers
• Withdrawn offers
• Offers declined
• Offers allowed to expire

An informal statement that the client "would like to proceed" should not necessarily be counted as a formal offer.

The reasons for decline are more actionable than the percentage alone.

Placement-start rate

An accepted offer is not always a successful placement.

A useful additional measure is:

Placement-start rate = candidates who start ÷ offers accepted

This reveals losses during:

• Notice periods
• Referencing
• Right-to-work checks
• Counter-offers
• Pre-employment screening
• Contract negotiation
• Communication between acceptance and start
• Changes in the client's circumstances

The agency should not assume every failed start is within the recruiter's control. It should still understand what happened.

Replacement, rebate and early-failure rates

A placement fee can be invoiced and still fail to create lasting value.

Track:

• Replacements requested
• Rebates issued
• Credit notes
• Candidates leaving within defined periods
• Clients terminating the hire early
• Temporary workers leaving assignments prematurely
• Reasons for early failure

These are not perfect measures of recruitment quality.

A strong hire may leave because of poor management, redundancy, a changed role or circumstances the agency could not reasonably have predicted.

SHRM notes that retention, performance reviews and hiring-manager satisfaction are useful proxies for quality of hire, but none is a complete measure by itself. It also warns that employee retention can be influenced by management and workplace conditions after the recruitment process ends.

Use early-failure data to investigate patterns rather than automatically assign blame.

Quality of hire is not one universal number

"Quality of hire" sounds like the ideal recruitment metric, but there is no agreed universal formula.

Possible components include:

• Performance against agreed outcomes
• Time to productivity
• Hiring-manager satisfaction
• New-hire satisfaction
• Retention
• Promotion
• Successful completion of probation
• Team contribution
• Client satisfaction
• Assignment completion

Different organisations define success differently. SHRM describes quality of hire as one of the most difficult recruitment metrics to define and recommends agreeing in advance what successful performance means.

For an agency, measurement is further complicated because post-hire outcomes are partly controlled by the client.

A practical agency approach may include:

• Client check-in after 30, 60 or 90 days
• Candidate check-in
• Probation completion, where the client shares it
• Replacement or rebate activity
• Repeat work from the hiring manager
• Whether the client would use the agency again
• Whether the candidate would work with the agency again

Do not present a composite "quality score" as objective unless the inputs, weighting and limitations are clear.

Time-to-fill and time-to-hire are not interchangeable

The terminology is used inconsistently across the industry.

SHRM currently describes time to hire as the period from identifying a candidate to offer acceptance. It recommends treating the measure as a process-health indicator rather than assuming that faster is always better.

"Time to fill" is commonly used for the longer period from opening or approving a vacancy to filling it.

An agency should define its own start and end events precisely.

For example:

Time to first qualified submission

Job accepted to first candidate formally submitted who meets the agreed essential criteria.

This is largely within the agency's influence.

Time to first interview

Job accepted to first completed client interview.

This depends on both agency delivery and client availability.

Time to offer

Job accepted to formal offer.

Time to acceptance

Job accepted to candidate acceptance.

Time to start

Job accepted to the candidate's first working day.

Each measure answers a different question.

A role can have a fast shortlist but a slow final outcome because the client delays interviews. One overall time-to-fill figure would hide that distinction.

Measure time in stage, not only total duration

Total time tells you that a process was slow. Time-in-stage data helps show where.

Useful stage measures include:

• Job accepted to first sourcing activity
• Sourcing to submission
• Submission to client response
• Interview request to interview completion
• Interview to feedback
• Final interview to offer
• Offer to acceptance
• Acceptance to start

This creates accountability without assigning every delay to the recruiter.

For example:

• A long sourcing stage may indicate a difficult market or insufficient consultant focus.
• A long submission-to-feedback stage may indicate client delay.
• A long offer stage may indicate unclear authority or package approval.
• A long acceptance-to-start period may be normal for a senior candidate with a three-month notice period.

The dashboard should show both the total and the stage causing it.

Job ageing helps identify wasted effort

An ageing report should show:

• Days since the job opened
• Days since the last client contact
• Days since the last submission
• Days since the last stage movement
• Number of candidates submitted
• Number interviewed
• Next agreed action
• Probability the vacancy remains live

An old vacancy is not automatically dead.

Executive searches and specialist roles may take time. But a job with no client communication, no stage movement and no next action is not meaningfully active merely because nobody has closed it.

Use ageing to prompt requalification:

• Is the vacancy still approved?
• Has the brief changed?
• Is the salary still realistic?
• Is the client still engaging?
• Is another supplier already at offer?
• Should the agency continue investing time?
• Should the job be paused or closed?

Measure client feedback speed

Client responsiveness directly affects recruitment performance.

Track:

• Average time to review submissions
• Average time to provide interview feedback
• Time to approve offers
• Vacancies without client activity
• Interviews awaiting feedback
• Offers awaiting approval

This helps prevent consultant metrics from absorbing delays caused elsewhere.

It also creates useful client advice:

Your average interview-feedback time is currently five working days. Three candidates have withdrawn during that stage over the last quarter.

That is stronger than telling a client vaguely that it needs to move faster.

Source effectiveness is more than source volume

A job board may produce many candidates and few placements. A referral source may produce fewer candidates but stronger conversion.

For each source, consider:

• Candidates identified
• Candidates contacted
• Replies
• Candidates qualified
• Candidates submitted
• Interviews
• Offers
• Placements
• Gross profit
• Cost
• Time required
• Early failures

SHRM recommends looking beyond raw source-of-hire and examining source effectiveness and cost. It also suggests that source of interview can help show recruiter influence before hiring-manager decisions affect the final outcome.

The source attached to a candidate also needs a consistent rule.

Was the candidate sourced from:

• The original place they entered the CRM?
• The channel used for this vacancy?
• The consultant who re-engaged them?
• A referral?
• A marketing campaign?
• A job application?

Without an agreed attribution model, source reports become arguments rather than analysis.

Pipeline coverage is useful only when the pipeline is credible

Pipeline coverage compares the expected value of open opportunities with a target.

A simple version might be:

Pipeline coverage = expected gross profit in pipeline ÷ target

But adding the full potential value of every job produces an inflated number.

A better forecast applies stage probabilities:

Weighted pipeline = potential gross profit × probability of placement

Those probabilities should be based on historic conversion where possible, not optimism.

For example:

• Unqualified opportunity: 5%
• Qualified vacancy: 20%
• Candidate interviewing: 40%
• Final interview: 65%
• Offer made: 85%

These are only illustrations, not recommended universal values.

The agency should calculate probabilities from its own history and segment them where sample sizes allow.

A retained assignment, exclusive job and speculative vacancy should not automatically receive the same probability.

Forecast accuracy should also be measured:

How close was the forecast made at the start of the month to the result eventually achieved?

A forecast that is consistently overoptimistic needs recalibration even if its pipeline coverage looks healthy.

Track repeat business, but define it properly

Repeat business can mean:

• A second placement
• Another vacancy
• Work from another hiring manager
• Revenue within 12 months
• A retained or exclusive assignment
• Renewal of a temporary placement
• Expansion into another division

Choose a definition aligned with the agency's model.

Useful measures may include:

• Percentage of revenue from existing clients
• Percentage of clients producing repeat work
• Average placements per active client
• Gross profit by client tenure
• Number of active hiring managers per client
• Client concentration
• Time since last assignment
• Revenue retained year over year

High repeat revenue is not automatically healthy if it depends on one client or produces poor margins.

Retention and concentration should be reviewed together.

Candidate-experience metrics belong on the dashboard

Candidate experience is not fully captured by a satisfaction survey.

Operational measures can include:

• Applications acknowledged
• Candidates awaiting an update
• Time from interview to communication
• Interviews rescheduled
• Candidate withdrawals
• Withdrawal reasons
• Offer declines
• Submission without recorded agreement
• Complaints
• Candidates who would engage with the agency again

These metrics help identify avoidable silence and process friction.

Do not reward consultants for fast closure if the result is a generic or inaccurate rejection sent merely to clear the dashboard.

The metric should support appropriate treatment, not encourage superficial compliance.

Fairness metrics can reveal stage-specific problems

An overall placement rate can hide differences between groups at different stages.

Where an employer or agency lawfully collects equality-monitoring data, it can compare progression through:

• Application
• Screening
• Submission
• Interview
• Offer
• Appointment

Acas advises that equality-monitoring information should be kept separate from applications and CVs, and should not be available to people making individual hiring decisions. It also notes that smaller organisations may be unable to report some data safely because they cannot guarantee confidentiality.

Possible measures include:

Interview selection rate for a group = people in that group interviewed ÷ people in that group who applied

and:

Offer rate for a group = people in that group offered ÷ people in that group interviewed

Differences do not by themselves prove discrimination.

They indicate where further investigation may be needed.

Consider:

• Sample size
• Voluntary non-disclosure
• Role and seniority
• Source
• Geography
• Required qualifications
• Assessment method
• Data accuracy
• Whether groups can be reported without identifying individuals

Do not use equality data collected for monitoring to influence an individual decision unless a specific lawful exception applies.

Activity metrics are not automatically vanity metrics

Calls, emails and candidate searches are inputs.

They can be useful when diagnosing why outcomes changed.

For example:

• Falling placements with stable conversion but fewer qualified conversations may indicate insufficient activity.
• High activity with low conversion may indicate weak targeting.
• Low activity and high conversion may indicate an efficient consultant—or a desk whose pipeline is about to run dry.
• High email volume with low replies may indicate poor data, relevance or deliverability.

The problem is not measuring activity.

It is treating activity as success without examining what it produces.

A useful activity metric should be connected to a funnel:

Contacts attempted → replies → qualified conversations → submissions → interviews → placements

Raw calls in isolation tell you little.

CVs submitted can be either useful or harmful

Submission volume helps show whether a vacancy is receiving attention.

But a target such as "five CVs per job" can create poor behaviour:

• Candidates sent before qualification
• Weak matches included to reach the number
• Duplicate submissions
• Client attention diluted
• Candidate consent overlooked
• Strong candidates buried among irrelevant profiles

A better combination is:

• Qualified submissions per active job
• Submission-to-interview conversion
• Client rejection reasons
• Time to first qualified submission
• Number of candidates required per placement

The aim is not the lowest possible number of submissions.

One consultant placing from four submissions is not automatically better than another placing from ten if the roles, markets and service models differ.

The ratio is a prompt for investigation, not a universal performance ranking.

Database size needs a quality denominator

"100,000 candidates" sounds impressive.

It does not reveal:

• How many records contain usable contact details
• How many addresses bounce
• How many people have opted out
• How recently records were updated
• How many duplicates exist
• How many candidates have current skills and preferences
• How many can be found through search
• How many have been engaged recently
• Whether the agency has a continuing reason to hold the information

More useful database measures include:

• Contactable candidates
• Candidates engaged within a defined period
• Bounce rate
• Duplicate rate
• Profile completeness
• Records with current availability
• Records without a lawful-basis or retention status
• Search-to-contact conversion
• Candidates rediscovered and placed from the database
• Records due for review or deletion

Database size is an inventory count. Database utility is a business metric.

"Jobs on" is meaningful only when jobs are qualified

A consultant can appear to have a large pipeline by retaining vacancies that are:

• Unapproved
• Already filled
• Being worked by many agencies
• Unsupported by client feedback
• Outside the available salary
• Missing a genuine decision-maker
• Paused indefinitely
• Duplicated

Separate statuses such as:

• Lead
• Potential vacancy
• Qualified job
• Live job
• On hold
• Filled
• Closed unfilled
• Withdrawn

Then define the minimum conditions required for a job to be treated as live.

This prevents a large but fictional vacancy count from being mistaken for future revenue.

Leaderboards need careful design

Leaderboards can make performance visible and create friendly competition.

They can also reward whatever is easiest to count.

A leaderboard based only on calls encourages calls. One based only on submissions encourages submissions. One based only on revenue may favour established desks or high-value markets.

Where leaderboards are used, consider balancing:

• Gross profit
• Conversion
• Data quality
• Client retention
• Candidate experience
• Forecast accuracy
• Team contribution
• Activity appropriate to the consultant's role

Do not combine unrelated measures into one unexplained score.

Consultants should understand:

• What is being measured
• Why it matters
• How it is calculated
• Which factors are outside their control
• What action the agency expects

The purpose should be learning and accountability, not public shaming through an oversimplified ranking.

Use averages and distributions together

An average time-to-fill of 35 days can describe several very different realities:

• Almost every vacancy takes around 35 days.
• Half take 10 days and half take 60.
• Most take 20 days but several very old jobs distort the result.
• Different sectors have completely different timelines.

Use:

• Median
• Mean
• Range
• Percentiles
• Number of observations
• Outliers

The median often provides a better picture of a typical experience when a few extreme cases distort the average.

Do not hide the difficult cases by reporting only the statistic that looks best.

Small samples need caution

A consultant who made one placement from one submission has a 100% conversion rate.

That does not yet prove they are the agency's most effective recruiter.

Always display:

• The numerator
• The denominator
• Reporting period
• Minimum sample threshold
• Comparison group

A ratio based on six events should not be interpreted with the same confidence as one based on 600.

This is particularly important for:

• Individual consultants
• Niche markets
• New client relationships
• Equality monitoring
• Rare recruitment stages
• New sources

Data quality is itself a recruitment metric

A dashboard cannot be better than the records underneath it.

Track:

• Jobs without an owner
• Jobs without a source
• Placements without a fee or margin
• Candidates without stage history
• Interviews without an outcome
• Offers without a result
• Closed jobs without a closure reason
• Duplicate records
• Invalid email addresses
• Overdue tasks
• Required fields left blank
• Records updated outside the expected workflow

When conversion appears poor, first ask whether stage changes were recorded consistently.

An agency may have a process problem, a data problem or both.

Build dashboards around decisions

A dashboard should help answer a question.

Daily operational view

• Which candidates need an update?
• Which interviews are awaiting feedback?
• Which offers are at risk?
• Which jobs have no next action?
• Which placements are due to start?
• Which temporary timesheets or approvals are missing?

Weekly desk view

• Where is the funnel slowing?
• Which jobs are ageing?
• Which clients are delaying?
• Which sources are producing interviews?
• Is the forecast changing?
• Which consultants need support?
• Which records are incomplete?

Monthly commercial view

• Revenue and gross profit
• Cash collected
• Placements
• Fill rate
• Conversion by stage
• Average fee or contribution
• Productivity by comparable role
• Repeat-client activity
• Rebates and replacements
• Forecast accuracy
• Cost to serve

Quarterly strategic view

• Quality and retention indicators
• Client concentration
• Market and sector performance
• Source effectiveness
• Candidate experience
• Equality monitoring
• Database health
• Trends rather than isolated results

Not every person needs every dashboard.

A consultant requires actions. A team leader needs coaching information. A director needs profitability, risk and forecast.

A practical agency scorecard

A balanced scorecard might include:

The agency does not need to display every measure at once.

Choose the smallest set that explains the business and supports action.

ISO 30414:2025 provides an international human-capital reporting framework covering areas including recruitment, costs, productivity, diversity and workforce turnover. Its broad structure reinforces the need to examine several dimensions rather than using a single activity or efficiency measure as a complete account of performance.

Where ATSpro fits

ATSpro's analytics and reporting connects reporting to the underlying recruitment workflow.

This allows an agency to examine measures such as:

• Jobs
• Candidate submissions
• Interviews
• Placements
• Conversion between stages
• Time to fill
• Consultant performance
• Revenue and commercial outcomes
• Pipeline and forecasting

The value of live reporting is not simply that the chart updates without a spreadsheet.

It is that each result can be traced back to the relevant jobs, candidates, clients and activities.

When a conversion rate changes, managers should be able to examine:

• Which roles caused it
• Which clients were involved
• Which rejection reasons were recorded
• Whether stages were updated correctly
• Whether the result is concentrated in one market
• Whether the sample is large enough to matter

ATSpro can also support leaderboards and forecasting, but the agency still needs to decide what behaviour it wants to encourage and how each measure should be defined.

Software can calculate the metric consistently.

It cannot decide whether the agency has chosen the right metric.

The takeaway

The recruitment metrics that matter are not simply the ones closest to revenue.

A useful measurement system connects:

• Activity
• Conversion
• Speed
• Quality
• Candidate and client experience
• Fairness
• Profitability
• Data reliability

Calls, emails, CVs, jobs and database size are not automatically vanity metrics. They become vanity metrics when they are displayed without showing what happened next.

Revenue is not automatically a complete performance metric either. It becomes misleading when margin, cost, quality and sustainability are ignored.

Define every metric. Use consistent cohorts. Segment comparable work. Show the numerator and denominator. Separate delays the agency controls from those it does not. Pair efficiency with effectiveness.

Most importantly, decide what action should follow when the number changes.

A dashboard that produces no decision is reporting.

A dashboard that reveals where the process is failing—and helps someone correct it—is management.

**Sources: CIPD, *Human Capital Metrics and Analytics: Assessing the Evidence*; CIPD, *Recruitment Process Overview*; CIPD, *Inclusive Recruitment: Guide for Employers*; SHRM, *Data-Driven Recruiting Proves Business Impact*; SHRM, *How to Measure Quality of Hire*; Acas, *Following Discrimination Law During Recruitment*; Acas, *Equality Monitoring Forms and Surveys*; ISO, *ISO 30414:2025—Human Capital Reporting and Disclosure*; user per month.*.**