Read Time:
Quality of Hire: How to Measure It Across Agencies
Quality of hire is how well a new employee performs, ramps up and stays, compared with what you expected when you opened the role. To measure it across agencies, score every placement on three things (performance, ramp-up and 12-month retention), pull the data from managers, your ATS and replacement records, and put every agency on one scorecard.
| Measure | What it tells you | Data source | When to check |
|---|---|---|---|
| Manager performance score | Did the hire deliver what the brief promised? | Hiring manager rating | At the end of the review period you set for the role |
| Ramp-up success | Share of hires reaching the agreed productivity target on time | Manager plus onboarding plan | At the ramp-up date set before the search |
| One-year retention | Share of hires still in the seat after 12 months | ATS and HRIS records | 12 months after joining |
| Interview-to-offer ratio | Whether shortlists are the right quality before the hire | ATS stage data | Every search |
| Replacement rate | How often a hire fails inside the guarantee window | Replacement and refund log | Guarantee window, typically 30-90 days on CBREX |
| Time from opening to hire | Speed, shown beside quality and never instead of it | ATS dates | Every search, with country notes |
Quality of hire is the value a new employee brings, shown through performance, ramp-up speed and how long they stay. It compares what the person actually delivers after joining with what the hiring brief expected. It is measured after the hire, which is why most agency reports skip it.
The metric is hard to pin down. Wikipedia's entry on recruiting metrics describes quality of hire as a measure of the value a new employee adds, reflecting performance, contributions, adaptability and personality fit, and notes that it is difficult to define and assess. That difficulty is exactly why teams fall back on easy numbers.
The easy numbers are submissions, interviews booked and days to fill. They are cheap to count and they reward activity. An agency that sends 40 CVs to hit a target looks better on a volume report than one that sends three right ones.
If you hire through several agencies, the gap widens. Each one reports in its own format, and none of them sees what happens after the offer is signed. You hold the only complete data, so the scorecard has to be yours.
Speed and cost are inputs. Quality is the result. A fast, cheap hire who leaves in four months is the most expensive outcome on your sheet, because you pay the fee, lose the ramp-up time and start the search again.
For the cost side, see our breakdown of what it really costs to fill a role. Here we stay on the quality side.
You cannot score a hire against expectations nobody wrote down. Before you brief any agency, put three measurable outcomes for the role in writing, and share the same page with every agency on the search.
A usable brief names the following:
The most common mistake we see from India-based GCC heads in a first hiring brief is not benchmarking their level of hiring against market salaries. They want very high quality talent but do not pay at that level. No agency can fix that gap.
When quality scores come back low on such a search, the agency is not always the cause. Check the brief before you mark down the supplier.
On fees, the mistake we see most often per overseas hire is calibrating the fee to the difficulty of the hire instead of to years of experience. Agency fees on overseas hires depend on the country you hire in. In our own European placements the range we typically see is 15-30%.
Fee level and quality are separate questions. Record the fee on each scorecard line anyway, so you can see whether a higher fee bought a better result over time.
Use a simple average of three components, each shown as a percentage. It is easy to explain to a finance head, easy to audit, and hard to game. A common formula averages the performance rating, the share of hires reaching acceptable productivity in a set time, and one-year retention.
The formula in full:
Quality of hire % = (performance score % + ramp-up success % + one-year retention %) ÷ 3
Ask the hiring manager to rate the hire against the three outcomes from the brief, on a five-point scale. Convert the average to a percentage (4.0 out of 5 is 80%). Collect it at the review point you set, not in a rushed annual review.
Count the hires who reached the agreed productivity target by the ramp-up date, and divide by total hires from that agency. A senior engineer and a digital account manager will need different ramp-up periods, so set them per role family and keep them constant across agencies.
Count hires still in role 12 months after joining, and divide by hires from that agency. On CBREX, more than 90% of our local hires stayed at least 12 months. Use your own figure as the baseline for your scorecard, since role mix and country change the number.
Suppose two agencies each placed 10 hires for your engineering teams. The numbers below are invented to show the arithmetic, not results from any real client.
| Component | Agency A | Agency B |
|---|---|---|
| Average manager rating | 4.0 of 5 = 80% | 4.25 of 5 = 85% |
| Hires at ramp-up target on time | 7 of 10 = 70% | 6 of 10 = 60% |
| Still employed at 12 months | 8 of 10 = 80% | 6 of 10 = 60% |
| Quality of hire | (80 + 70 + 80) ÷ 3 = 76.7% | (85 + 60 + 60) ÷ 3 = 68.3% |
Agency B has the higher manager rating, which is what a quick review would reward. Once ramp-up and retention are counted, Agency A is clearly ahead. Without the second and third columns, you would send more roles to the wrong agency.
For the money side of the same comparison, our guide to recruitment ROI shows how to turn a retention gap like this into a cost figure.
No single system holds everything you need. Manager feedback holds performance and ramp-up. Your ATS holds stage data and retention dates. Replacement records show early failures. Join the three by agency and role.
Send a short form at the review point. Keep it to the three brief outcomes plus one question: would you hire this person again? Make it a required step in your process, because voluntary feedback skews to the extremes.
The pass mark is the hire meeting every outcome from the brief at the level the manager expected. The walk-away signal is two or more placements from one agency where the manager would not rehire.
Your ATS holds source, stage dates, offer and joining dates, and exit dates. Tag every candidate with the agency that submitted them. If your agencies work through one platform, integration with your ATS keeps this tagging automatic. Our note on managing recruitment vendors across countries on one dashboard covers the setup.
Also track the interview-to-offer ratio per agency. If candidate quality is the issue, that ratio is where it shows first.
Log every replacement request, refund and credit note, with the reason. These show early failures that never reach a performance review. On CBREX, the replacement guarantee covers a placement for 30-90 days, and the remedy per eligible claim is a free replacement or a refund.
Read the log for patterns, not totals. Several early exits for the same reason, such as a role described differently from what the hire found, point to a brief problem rather than an agency problem.
If managers do not return feedback, do not fill the gap with a guess. Score that placement as "not rated" and track the response rate. A response rate you cannot defend means the whole scorecard is weaker than it looks, and the fix is process, not maths.
Retention takes 12 months to show. You need earlier signals, or you will keep sending roles to a weak agency for a year. The best early signal is how many interviews it takes to get an offer.
On recent CBREX searches, clients saw an interview-to-offer ratio of 1:3 to 1:4 after our 3-level screening. That means roughly one offer for every three to four interviews. Those are our per-search figures, so set your own baseline before comparing.
The three levels run in order: the agency pre-screens the candidate, C Screen validates the CV with AI, and the shortlist is stack-ranked. C Screen is our AI screener, and its 98% accuracy rate was measured over more than 10,000 cases.
These are leading indicators. They do not replace the 12-month score, but they tell you which agency to put on watch while you wait for it.
Use one scorecard, with the same three components for every agency, and add a country note column rather than changing the formula per market. A scorecard that changes by country cannot be compared, and the point is to see who delivers quality wherever they work.
| Column | What goes in it |
|---|---|
| Agency and country | Name, market, role family |
| Placements measured | Number of hires in the period; flag samples that are too small to trust |
| Quality of hire % | Average of performance, ramp-up and retention |
| Interview-to-offer | Per search, from the ATS |
| Replacements and refunds | Count and reasons |
| Fee terms | Percentage and what it is calculated on |
| Country note | Notice period, visa step, local versus relocated hire |
Raw days mislead across markets. Notice periods differ, and a visa step can add weeks that no agency controls. The reference points below come from CBREX searches and show why you should record the context beside the days.
| Search | Days, opening to hire | Context |
|---|---|---|
| India GCC or niche search (most recent) | 24 days | We have filled AI, semiconductor, hardware and software engineer roles for GCCs |
| India-based roles, typical | 15-45 days | Typical range on CBREX |
| Germany search | 30-45 days | No visa step, because the candidates were local or EU citizens who could work in Germany |
| Pharma placement, USA (last anonymised) | 17 days | Single placement, not an average |
| Singapore placement, 2021 | 31 days | Single placement from role opening to hire |
Typical European notice periods run 30 days, so a European hire's start date sits at least that far beyond the offer date, whatever the agency does. Our time-to-hire benchmarks give you a wider market view for context.
Speed belongs on the scorecard as context, not as a quality score. Agencies that deliver in 17 or 24 days and then lose the hire in month five have not delivered quality.
Mark every placement as local or relocated. A relocated hire carries visa and settling-in risk that a local hire does not, so retention for the two groups should be tracked separately before you average them.
Rank agencies by quality of hire percentage, but only after you check sample size and read the replacement reasons. Send more roles to agencies that score well over several placements, and put low scorers on a written improvement plan before you cut them.
Use these decision rules:
Something remains unsettled here. Manager ratings are subjective, and a manager who disliked the hire process may mark the person down. We have no perfect fix. Using two raters on senior roles reduces the bias, and a standard form with fixed outcomes reduces it further.
Scoring gets harder as agencies multiply. One client consolidated 54 agency contracts into a single CBREX contract, and their time-to-hire before consolidation had been 60-75 days. Fewer contracts mean fewer formats to reconcile, so one scorecard can cover everything. Our guide on how many agencies to use per role helps you decide the right number.
Guarantees vary by partner, so read the terms before signing. On CBREX, the replacement guarantee covers a placement for 30-90 days, and the remedy per eligible claim is a free replacement or a refund. Check which exits qualify, and when the clock starts.
A marketplace can check quality at several points: how it matches roles to agencies, how it screens CVs and how it tracks outcomes. On CBREX, AI vendor matching routes the role to relevant specialist agencies, and each shortlist passes through a 3-level screen before it reaches you.
Measure performance and ramp-up at the dates you set in the brief, and retention at 12 months. Early signals such as interview-to-offer ratio and replacement requests give you a view in the first weeks, long before the 12-month figure exists.
Yes, but sample sizes are small, so judge on patterns across searches rather than one placement. For senior and niche roles, add a second rater and a longer ramp-up window. Our guide to hiring your first country manager abroad covers those searches in more detail.
Start small. Take your last ten placements, tag each by agency, and calculate the three components in a spreadsheet this week. You will likely learn more about your agencies from that sheet than from a year of submission reports.
If you want the scorecard built into how you hire, book a CBREX demo and we will show how one contract, one platform and ATS-tagged placements make quality of hire measurable across 55 countries. Curious what slow, unmeasured hiring costs you today? Calculate your Hidden Hiring Tax, or talk to our team about your agency mix.
Recruiting firms that want to be measured on outcomes can sign up as a talent supplier on the platform.
