Cheap applications can make a campaign look successful at first glance while the recruitment team wastes time on unsuitable people. To measure the real contribution of advertising, link the spend to a specific position and track the candidate through qualification, a completed interview and hiring.
The cost of a qualified candidate is the acquisition spend divided by the number of unique applications that meet the predefined requirements for the position. The cost of a completed interview uses the first interviews actually held as the denominator. Calculate both metrics for the same group of applications and state clearly which costs you include.
What does "qualified candidate" mean?
It's a candidate verified to meet the essential requirements of the specific role. Sending a CV is not qualification in itself. Define the criteria before the advertising starts: required experience, relevant skills, location or the option of remote work, working schedule and a realistic match of expectations.
Use only job-related criteria. Record the reason for rejection with consistent categories, for example "missing a required skill" or "incompatible schedule". If recruiters apply different rules, the report measures the differences in assessment together with channel quality.
Four metrics that shouldn't be mixed
| Metric | Formula | What question it answers |
|---|---|---|
| Cost per application | Ad spend / unique applications | How much does entry into the funnel cost? |
| Cost per qualified candidate | Ad spend / qualified candidates | How much does a suitable entry cost? |
| Cost per completed interview | Ad spend / candidates with a completed first interview | How much does a real meeting cost? |
| Advertising cost per hire | Ad spend / hired candidates | What is the contribution up to hiring? |
The last metric is not the full cost of recruitment. For that you also need to include the team's work, tools, external partners and other applicable costs. With zero hires, the cost per hire is not zero: the metric can't be calculated yet.
Example: more applications don't mean a better campaign
The following numbers are a hypothetical example, not a DigitaL+ client result. Both channels advertise the same position; the application groups have had enough time for a first interview.
| Result | Channel A | Channel B |
|---|---|---|
| Ad spend | 1 200 EUR | 1 200 EUR |
| Unique applications | 120 | 60 |
| Qualified candidates | 24 | 30 |
| First interviews held | 12 | 24 |
| Cost per application | 10 EUR | 20 EUR |
| Cost per qualified candidate | 50 EUR | 40 EUR |
| Cost per completed interview | 100 EUR | 50 EUR |
Channel B costs twice as much per application but half as much per completed interview. This is a reason for an additional test, not automatic proof that it should get the whole budget. Check the sample size, the quality of the interviews and whether the team handled both channels equally.
How to connect advertising with the recruitment system
- Create a separate identifier for the position and record the source, campaign and application date.
- Use an internal candidate identifier and a separate record for the application to a specific position.
- Keep the original source; don't overwrite it on a later visit.
- Record the dates of qualification, scheduling, the completed interview, rejection and hiring.
- Remove duplicate applications to the same role by a predefined rule. Don't automatically delete a valid application for another position.
This data belongs in the recruitment system. Don't send CVs, names or interview content in analytics events. The general link between the advertising input and the business result is covered in the material on inquiry quality and tracking; here we measure the candidate funnel specifically.
Compare cohorts, not incompatible monthly numbers
If you divide October's spend by October's interviews, some of those interviews may be with candidates from September. Group people by application date and track their progress. Mark the groups that haven't finished the process yet.
Separate roles, regions and campaigns with different requirements. Also track time to first contact: a channel with suitable candidates may look weak if contact comes too late. The difference between scheduled and completed interviews points to a separate problem in confirmation or organization.
What to change based on the result
- Many applications, few qualified: check the promise in the ad, the role description and the screening questions.
- Good qualification, few interviews: check the contact speed, the times and the confirmation.
- Many interviews, few hires: review the expectations, the offer and the criteria with the recruitment team.
For recruitment and employer branding marketing DigitaL+ can connect performance campaigns with measuring the candidate funnel. A useful starting document is a "position → source → qualification → interview" table, not a report with clicks only.
Frequently asked questions
What is a good cost per qualified candidate?
There's no universal value. Evaluate it against the difficulty of the role, the market, the quality of candidates and the likelihood of hiring. Your own comparable campaigns are more useful than someone else's average price.
Should we count a second interview as another result?
For the cost of a completed first interview, you count the candidate once. Later meetings are separate stages; mixing them artificially increases the denominator.
Can the ad platform show all hires?
Not necessarily. Compare the platform report with the recruitment system and note the tracking limitations. An unknown source should stay unknown instead of being distributed arbitrarily.
