Data-driven recruitment means that hiring decisions rest on measurements of what actually happens, rather than on habit or gut feeling. It covers the whole process: where the role is marketed, who sees it, how applicants are assessed and how long the hire takes.
The term covers a lot of ground. Some people mean metrics such as time to hire and cost per hire, others mean tools that sort applications. This page is about the first step, reaching the right candidates, and about the data that makes it possible: how candidates behave on job ads.
In short: measure what candidates do, and let that decide the next move.
Which data is used in recruitment?
It helps to separate four kinds of data, because they answer different questions.
Process data: Time to hire, cost per hire and how many qualified applicants each role attracts. It shows whether hiring is working, but only once the applications are in.
Channel data: Where the applicants and the best candidates came from. It shows which channels are worth the money for a particular type of role.
Behavioral data: How candidates behave on job ads, such as how long they read, how far they scroll and whether they go looking for similar roles afterwards. It shows interest before anyone has applied.
Demographic data: Age, location and job title. It says something about who a person is, but little about what they are interested in right now.
Where does data make the biggest difference?
Many teams start measuring when the applications arrive. But the decisions that determine who sees the role are taken before that: which channels are used, who the ad is shown to, how much budget each channel gets, and how long the campaign runs. If those decisions are made out of habit, there is little the application data can repair afterwards.
That makes the marketing of the role a natural place to start. Read how it works in the guide to job marketing, and how to find out where candidates drop off in too few applicants.
Behavioral data: what candidates actually do
Behavioral data in recruitment means measurements of how candidates behave on job ads. At Frantz, measurement points on the job ads register scrolling, reading time and intent to apply. Every interaction that passes a threshold becomes a data point, tagged by role type, industry and geography.
Over time the data points show which audiences engage with which types of roles. A visitor who leaves the ad straight away produces very different data from one who reads the whole description and stays for several minutes. Clicks and impressions do not capture that distinction.
Read more in the article about data points and behavioral data, and about how impact is measured in the article on marketing impact.
Why 90 days?
Frantz builds every campaign on behavioral data from the past 90 days. That gives a current picture of who has shown interest, rather than profile data that may be out of date. The same 90 days are used to pick channels and distribute budget according to what has produced the best engagement for similar roles.
Digital twins: more candidates who resemble them
The candidates who have already engaged with similar roles are only the starting point. Frantz builds an audience from them, and the channel algorithms, Meta for instance, find people who resemble them: digital twins. That is how the ad also reaches candidates who are not applying for anything, but who look like the ones who have shown interest.
Read how it works in the guide to digital twins in recruitment. It is particularly useful for reaching passive candidates, who rarely look for work on job boards.
Privacy and consent
Behavioral data that can be linked to a person is personal data and has to be handled under the privacy rules. At Frantz the tracking script splits consent into categories, and marketing pixels from the ad platforms are only activated after explicit consent. Consent is stored with a timestamp and a policy version, and primary data is stored within the EEA.
Read more about consent and data storage in the FAQ.
How to get started
Decide what a good result looks like: Qualified applicants and engagement on the ad say more than impressions and clicks.
Measure behavior on the ad: Reading time and scrolling show whether the right candidates are taking the role seriously.
Compare channels on quality: Look at which channels deliver engaged candidates for that type of role. The most clicks is not the same as the most good candidates.
Carry the experience into the next role: Every campaign produces data that makes the next choice of channel, audience and budget better.
With Frantz AI you enter a job title and a location and get a campaign strategy built on behavioral data from similar roles over the past 90 days, with a benchmark for what a good result looks like.
Read more
See how the technology works, or start with the guide to job marketing.
Frequently asked questions
Data-driven recruitment is hiring where the decisions are made on the basis of data about what works, such as which channels deliver qualified applicants and how candidates behave on job ads, rather than on habit.
Data-driven recruitment covers the whole hiring process, from marketing to the assessment of applicants. Job marketing is the part that is about reaching the right candidates, and that is where behavioral data from job ads is most useful.
Behavioral data in recruitment means measurements of how candidates behave on job ads, such as reading time, scrolling and intent to apply. It shows who is genuinely interested in a type of role, even before anyone has applied.
Digital twins are people who resemble the candidates who have already engaged with similar roles. Frantz builds an audience of the interested candidates, and the channel algorithm, Meta for instance, finds the twins.