What are the key metrics to measure recruitment success?
While evaluating results is crucial to improve them, it’s hard to find the time to check how well we are hiring. Here you’ll find a selection of KPIs to make the most of our (limited) measuring time.
Before getting started, here’s a brief disclaimer: this selection of metrics is made knowing, first, that hiring is only successful when new hires fulfil their objectives and, second, that there's huge value in ensuring we recruit right. For people focused on other aspects, such as how fast or costly your talent acquisition function is, there are many lists of metrics out there that cover them thoroughly.
“Only about a third of U.S. companies report that they monitor whether their hiring practices lead to good employees (...) Imagine if the CEO asked how an advertising campaign had gone, and the response was «We have a good idea how long it took to roll out and what it cost, but we haven’t looked to see whether we’re selling more.»” Peter Cappelli, Your Approach to Hiring Is All Wrong
This surprising lack of quality monitoring, which Peter Capelli describes in such a compelling way for the U.S., seems to happen all around the world, and there are a couple of likely reasons. One of the key ones is that evaluating if new team members are good, bad or somewhere in between is not as straightforward as measuring the speed or the cost of the process. It’s impossible to find fully objective metrics to do it. However, there are key performance indicators (KPIs) that serve as proxies (indirect measures) of it. Here, we’ve classified them in three categories:
Turnover indicators
Definition: also called attrition (or even failure) indicators, they simply measure what percentage of the hires leave — or stay, and then they are typically called retention or tenure rates.
Pros-and-cons: while these metrics don’t offer a lot of nuance, only flagging when we hire bad enough for people to leave the organisation, they still signal really important flaws and they are the easiest KPIs for recruitment success. High turnover rates typically indicate a mismatch between new hires’ expectations and what they actually find (e.g. the job is oversold), or a mismatch between the team's expectations and the candidates that make it through the process (e.g. skills are poorly assessed or there is a mismatch between actual job requirements and skills sought for during the process).
ℹ️ Even when the main reason for our poor turnover rates seems to be external to the recruitment process (e.g. onboarding processes that don’t set people up for success or high levels of on-the-job pressure that make people leave), it’s still useful to consider adjusting the process while the organisation changes. A typical example is how fast-growing startups need to hire self-learning, highly-motivated candidates while they figure out how to arrange better training and reduce the workload per team member. Talent acquisition should bring the right people for the current context, not for an ideal context that doesn’t match reality. Getting started: make a list of the people you’ve hired recently and simply count how many have left from the total number you hired. This must be adapted to the type of organisation: large organisations can measure this every month, based on the hires that have been already in for at least 6 and 12 months; in contrast, smaller fast-scaling ones may only have resources to measure this every 6 months for hires that were made at least 6 months ago.
Opinion-based indicators
Definition: they build on the opinion of the team to measure the quality of hiring. Types of opinion-based indicators include:
Ratings: the indicator comes from a simple calculation (typically the average) based on the score given by one or more people among the following options: a key individual (e.g. hiring manager rating or performance review scoring), a variety of team members (called 360 ratings), and even the hire themselves (self rating).
Questionnaires: the indicator is the result of processing more complex surveys, such as satisfaction, engagement, belonging, etc.
Feedback: some authors mention feedback forms as a way to establish recruitment metrics. While we aren’t convinced that free text answers can be processed with guarantees into a single indicator, we thought we should mention — please get in touch if you have seen this working well in practice.
Pros-and-cons: sadly, opinions are easily biased. For example, if we are friends with a colleague, we are more likely to have a positive opinion on their work. Plus, if surveyors or respondents don’t take the importance of an objective response seriously, answers may be easily nudged or even become irrelevant — we’ve all seen those 1-to-10 satisfaction surveys with numbers 9 and 10 shown in grass-green vs 1 to 6 shown in blood-red (how angry do you need to be to click on the red?) However, if the opinion collection is managed well, team members are trained to disrupt biases, and they buy in and constructively participate, opinion surveys are fairly simple and inexpensive to run, becoming a realistic way of estimating recruitment success in a more nuanced way than turnover metrics.
ℹ️ These indicators can be made with one single or multiple opinions. Considering more relevant opinions makes the indicator more robust and less biased, but also more complex and expensive to collect.Getting started: if you already have an evaluation set up that you can extract ratings from, e.g. performance reviews, you can take the average of the evaluations of your recent hires as a recruitment KPI. If you don’t have one, a great option is a rating called Net Hiring Score, which uses a single NPS-like question — you’ll find great examples online.
Performance indicators
Definition: they measure the productivity of the new hires to evaluate the quality of hiring.
On-the-job metrics: depending on the nature of the role, the indicators will be calculated using the volume of sales, completed calls, project milestones, tickets solved, positions filled, items manufactured, errors detected, successful deliveries, positive client reviews… to name a few. Organisations may use either the absolute value of the metric (e.g. 300 calls), the achievement vs an objective (75% of the sales quota), or even temporal metrics like ramp up time, or how long it takes hires to reach a certain level of productivity (72 days to expected performance).
Key objective accomplishment: if the organisation uses OKRs, 30-60-90 day plans, or any other system to manage work objectives, indicators can be built based on the level of achievement of the hires within that system. Because most of these goal setting systems were not designed for evaluation (but to make people walk the extra mile), a low scoring is not necessarily a sign of bad recruitment practice.
Other quantifiable signals of performance: achievement of bonuses, awards, promotions, salary increases, etc. We needed to mention them, but we don’t recommend them due to three frequent hard-to-avoid issues: taking too long to be measurable, being too vulnerable to biases, and tending to be strongly influenced by non-hiring related aspects. For example, how many people get a promotion soon that is not because of a random event, like someone else leaving — are people that don’t get promoted a bad hire?
Pros-and-cons: performance indicators also offer more nuance than turnover indicators and, in contrast to opinion-based indicators, they try to mitigate the subjectivity by establishing neutral measures of productivity. Of course, no indicator completely removes personal subjectivity without bringing along new challenges. For example, unless you consider more than one metric to build your indicator, they offer a very monolithic view on performance (a single metric is rarely considered a good source of truth), e.g. sales volumes or completed calls typically need to go together with a measure of customer satisfaction, to avoid creating an army of robots that generate many unsatisfied customers.
ℹ️ Performance indicators are trickier than they seem and need to be cautiously designed for the entire organisation, considering the characteristics of the roles and teams: a 6-month hire on a complex role will probably not be as productive as the average of a team with a tenure of 3 years; it might be easier to measure the performance of customer support than learning and development teams; etc.Getting started: start with what you have, whether that’s one or two performance metrics, OKR accomplishment rates, etc. However, honestly, you might be better off starting with a simpler turnover and/or opinion-based KPI.
No excuses - start measuring!
We get it: measuring how well we're hiring is not as easy as measuring how fast a car is driving. Nevertheless, there's no excuse that justifies not having at least one recruitment KPI in place based on one of the metrics above. Once you’re there, you’ll need to make sense of the chosen metric(s) as organisation-wide KPI(s) and figure out how to build on that to improve or maintain what's usually called Quality of Hire (QoH). That's precisely what we will cover in our next article. If you want even more, please subscribe and stay tuned! Feedback, ideas or requests? Please get in touch!