Data-Driven Hiring: How to Use Recruitment Analytics to Make Better Decisions

The Real Cost of Gut-Feel Hiring

Every hiring decision involves uncertainty. You cannot know with certainty how a candidate will perform once they are in the role, how they will integrate with the team or whether the skills they demonstrated in an interview will translate to on-the-job output. But there is a significant difference between informed uncertainty — where decisions are anchored in structured data about the candidate and the role — and gut-feel hiring, where decisions are made primarily on subjective impressions and pattern-matching against past experience.

The cost of a bad hire is well-documented. The US Department of Labor estimates it at 30% of the employee's first-year earnings. For a mid-level role at €50,000 per year, that is a €15,000 direct cost — before accounting for the time spent managing performance issues, the impact on team morale, and the cost of restarting the recruitment process. Research by the Harvard Business Review found that 80% of employee turnover is due to poor hiring decisions, with structured interviews and skills assessments cited as the most reliable predictors of on-the-job performance.

Data-driven hiring does not eliminate uncertainty. It reduces it systematically, by replacing subjective impressions with structured data at every stage: structured interview scores rather than general impressions, skills assessments rather than assumptions based on job titles, source tracking rather than guessing where good candidates come from, and funnel metrics that identify where your process is losing strong applicants unnecessarily.

The Metrics That Actually Matter

Every ATS generates metrics. Most teams track far too many, understand far too few and act on almost none. Effective recruitment analytics starts with selecting a small number of high-signal metrics and tracking them consistently over time rather than attempting to monitor every possible data point.

Time-to-hire is your most immediate operational indicator. Long time-to-hire loses candidates to competitors who move faster, increases cost-per-hire through extended advertising and recruiter time, and signals process inefficiency that is usually concentrated at one or two specific stages. Benchmarks vary significantly by industry and role type, but in competitive European markets, top candidates typically make decisions within five to ten days of first contact. If your average time-to-hire for technical roles exceeds 30 days, you are almost certainly losing the best candidates before you extend an offer.

Offer acceptance rate is a leading indicator of several different problems, each requiring a different response. If your offer acceptance rate drops below 80%, the root cause is likely one of: compensation out of alignment with market rates, a candidate experience that damaged enthusiasm between application and offer, a competitor moving faster and extending an offer first, or expectations set during the interview process that the offer did not meet. Tracking offer acceptance rate alongside time-to-hire and candidate source tells you which of these is most likely.

Source Effectiveness Analysis: Where Good Hires Actually Come From

Most recruitment teams dramatically over-invest in some candidate sources and under-invest in others, simply because they do not have reliable data on where their successful hires originate. LinkedIn advertising is expensive and visible. Employee referrals are free and produce candidates who arrive with existing social proof. niche job boards for specific sectors often produce higher-quality applicants than large general boards at lower cost. Without source tracking in your ATS, you are allocating job board budget on habit rather than evidence.

Source effectiveness analysis goes beyond counting applications by source. The metric that matters is quality-per-source: what percentage of applications from each source progress past CV screening, what percentage reach interview, and what percentage result in a hire. A source that generates 200 applications but produces zero hires is not a good source regardless of its apparent reach. A source that generates 15 applications but produces four hires is an exceptional source that deserves more investment.

Employee referrals consistently outperform other sources on quality-of-hire metrics. Referred candidates typically have a 55% higher retention rate at 24 months compared to candidates sourced via job boards, according to research by LinkedIn Talent Solutions. They also tend to progress through interview stages at higher rates, reducing the time your team spends screening applicants who are not a strong fit. Tracking referral hire rates in your ATS allows you to calculate the ROI of a formal referral programme and build a business case for investing in referral incentives.

Funnel Conversion Analysis: Finding Where You Lose Good Candidates

Your recruitment funnel has a specific shape. For a typical mid-level role, you might receive 150 applications, advance 30 to CV review, invite 10 to first interview, progress 5 to final interview and extend 2 offers. Each of these transitions is a conversion — and each conversion rate is a signal about a different aspect of your process.

A very low application-to-CV-review rate suggests your job description is either attracting the wrong candidates (misaligned expectations) or your screening criteria are too strict (rejecting viable candidates early). A high first-interview-to-final-interview conversion but low final-interview-to-offer conversion suggests your final-stage process is not generating the confidence needed to make an offer decision — often a sign that interview formats are not effectively assessing the right competencies. A low offer acceptance rate, as discussed, has several possible causes that need to be distinguished through additional data.

Funnel analysis over time is more powerful than a single-point snapshot. A conversion rate that was 40% six months ago and is now 25% at the same stage signals a change — in the candidate market, in your employer brand, in the quality of your job descriptions or in a specific hiring manager's standards. Spotting this trend early allows you to investigate and intervene before you have a prolonged period of poor hiring performance.

Measuring Quality of Hire: The Hardest and Most Important Metric

Time-to-hire and cost-per-hire are process efficiency metrics. They tell you how well your recruitment machine runs. Quality of hire tells you whether it is producing the right output. It is the metric that connects recruitment performance to business performance — and it is the metric most teams neglect entirely because it requires follow-through after the hire is made.

Quality of hire is typically measured at two points: 90 days post-hire and 12 months post-hire. The inputs vary by company, but commonly include: manager satisfaction score (collected via a structured survey), performance rating in the first formal review, retention at the measurement date, and whether the hire has been promoted or given expanded responsibility. Recording these data points in your ATS against the original hire closes the loop between recruitment process and business outcome.

Once you have quality-of-hire data linked to source, interview score and hiring manager, the analysis becomes genuinely powerful. If candidates hired via employee referral consistently score higher on 90-day quality assessments than candidates sourced via job boards, you have a quantified case for investing in your referral programme. If candidates who scored above a certain threshold on structured interview assessments retain at significantly higher rates, you have evidence for standardising that assessment across all roles of that type.

Building a Data Culture in Your Recruitment Team

Collecting recruitment data is the easy part. Turning it into decisions is the hard part, and it requires deliberate effort to build a team culture where metrics are consulted habitually rather than assembled reactively when someone asks for a report.

Start with a regular cadence of data review. A monthly 30-minute analytics review with your recruitment team — looking at the five core metrics and any significant changes from the prior month — builds the habit of data consultation without creating reporting overhead. Over time, this review generates the institutional knowledge to distinguish signal from noise: the time-to-hire spike that reflected a one-off role, versus the sustained increase that indicates a structural problem requiring intervention.

Involve hiring managers in the data conversation. Most hiring managers are deeply interested in how their roles are performing in the market — how long it is taking to fill their open positions, how their roles' offer acceptance rates compare to others, and what sources produced their last successful hires. Sharing relevant data with hiring managers at the start of each new role opens a productive conversation about process and expectations, and creates shared ownership of the outcome rather than a dynamic where recruitment is the bottleneck and hiring managers are passive consumers.

AI and Predictive Analytics in Recruitment

Traditional recruitment analytics is retrospective — it tells you what happened. AI-powered analytics introduces a predictive dimension: using historical patterns to inform current decisions. At its most practical, this means AI match scoring that evaluates new applicants against the profile of candidates who previously succeeded in similar roles, surfacing strong matches that might otherwise be overlooked in a high-volume application flow.

Treegarden's AI Match Score analyses candidate CVs against role requirements and generates a fit score for each applicant, allowing recruiters to prioritise their review effort on candidates most likely to progress. In high-volume roles where 200 applications arrive in the first 48 hours, this focus significantly reduces the time between application and first contact — which, in competitive markets, is the difference between a candidate who is still available and one who has already accepted an offer elsewhere.

Predictive analytics also supports proactive talent pipeline building. By identifying patterns in which types of roles tend to open at specific points in the business cycle, HR teams can begin sourcing candidates before a role is formally opened — reducing time-to-fill by weeks when the requisition is eventually approved. This requires historical data from your ATS to be meaningful, which is one more reason to invest in data collection discipline from the beginning of your ATS usage rather than treating it as an afterthought.

Frequently Asked Questions

Which recruitment metric should I focus on first if I am starting with analytics?

Start with time-to-hire, broken down by role type and department. This single metric reveals the most about where your process is slow, which roles are hardest to fill and how your speed compares to market benchmarks. Once you understand your time-to-hire baseline and can see which stages absorb the most time, you have a clear optimisation priority. Most teams that start here identify two to three specific interventions that reduce time-to-hire by 20–30% within 90 days — such as reducing CV review time by using AI match scoring, or shortening the gap between first and second interviews by pre-scheduling panel time.

How do we measure the quality of a hire, not just the speed?

Quality of hire is measured retrospectively, typically at 90 days and 12 months post-hire. The three most commonly used indicators are: performance rating in the first formal review, retention at 12 months, and hiring manager satisfaction score collected via a structured survey at 90 days. The most predictive pre-hire indicators of quality tend to be structured interview scores, skills assessment results and reference check outcomes — not CV screening or gut-feel impressions from an unstructured interview. Linking your post-hire quality data back to these pre-hire indicators over time reveals which assessment methods are actually predictive for your specific roles and culture.

Our team is small. Do we actually generate enough data for analytics to be useful?

Yes. Even small teams benefit from recruitment analytics, but with a different focus. Rather than seeking statistical significance across hundreds of hires, small teams use analytics primarily for process visibility: how long is each role taking, where in the funnel are candidates dropping out, which source produced the last five hires, and how quickly did candidates move between stages. These questions are answerable with as few as 10–15 hires and produce actionable insights regardless of company size. The discipline of tracking consistently from the beginning also means that when you do reach the scale where statistical analysis becomes meaningful, you have the historical data to work with.

What is the difference between recruitment metrics and recruitment analytics?

Metrics are individual measurements at a point in time — time-to-hire this quarter, number of applications received last month. Analytics is the practice of combining metrics over time and across dimensions to identify patterns and make predictions. A metric tells you that your average time-to-hire is 32 days. Analytics tells you that your time-to-hire is 32 days on average but 47 days for technical roles, that it has increased by eight days over the past two quarters, and that the delay is concentrated at the technical assessment stage — pointing to a specific intervention. The difference is not the data; it is the questions you ask of it.