.png)
Cost per qualified applicant is your total recruitment spend divided by the number of applicants who met every pre-agreed qualification criterion for the role.
That's the whole calculation. The difficulty sits in the inputs.
"Qualified" does the heavy lifting in CPQA, and it has to point at something a second person could verify without you in the room.
Appcast's recruitment glossary is blunt about who decides: "The definition of qualified is up to the discretion of each hiring organization. Most companies select a step in the ATS workflow as the designation such as Recruiter Accepted, Submitted to Hiring Manager or Interviewed."
Pick that step once. Write it down. Apply it to every campaign you intend to compare.
Timing matters as much as the choice itself. Set the standard before the campaign starts, then leave it alone. Move it halfway through and you lose the ability to compare anything at all, not across channels, not across quarters, not across roles.
CPQA = total spend ÷ number of qualified applicants
Recruitics runs the same arithmetic under a slightly different name, cost per quality applicant: "take your total spend on a particular job or set of jobs and divide it by the total number of quality applicants received on the same job(s)."
Total spend covers every cost you can attribute to the campaign or job group you're measuring. Media across all channels, platform fees, agency fees, any line item pointed at sourcing candidates for those roles. Assign them the same way every time. Count programmatic spend against a logistics role and you have to count it against the retail role you want to compare it to.
Number of qualified applicants counts the candidates who finished the full screening flow and passed every criterion agreed before launch. Someone who abandoned halfway doesn't count. Someone who finished and failed one criterion doesn't count either.
Both rules protect the denominator. Loosen either one and your CPQA will look better than your hiring actually is.
You run a campaign for 20 warehouse operative roles in Rotterdam. Total spend across all channels is €8,000. These figures are illustrative, so read them as £8,000 if you budget in sterling.
At the agreed ATS step, Recruiter Accepted, 80 candidates have met every criterion.
€8,000 ÷ 80 = €100 per qualified applicant.
Now put €8,000 into each of two channels.
Channel A returns 400 applicants. 80 pass. CPA is €20 and CPQA is €100.
Channel B returns 200 applicants. 160 pass. CPA is €40 and CPQA is €50.
On CPA, Channel A is the obvious buy at half the price. On CPQA it costs twice as much as Channel B.
Channel A also hands your recruiters 320 applications to reject. Channel B hands them 40. At twenty minutes of screening each, that gap is around 93 hours of recruiter time sitting inside the cheaper-looking number, and none of it shows up in the CPA.
Channel A qualifies one applicant in five. Channel B qualifies four in five.
CPA measures how efficiently an ad generates volume, and nothing else. It won't tell you whether a single one of those applicants meets the role criteria, so the screening cost lands on you after you've already paid for the click.
CPQA ties spend to something your recruiters can act on this week. It answers the question a Head of Talent Acquisition actually asks about a live campaign: what did we pay for each candidate who was ready to interview? What it still can't tell you is whether that candidate accepts the offer or stays past month three. Those risks remain yours.
CPH carries everything, including the two risks CPQA leaves open. But you can only calculate it once someone has been hired, which lands weeks or months after the campaign closed. By then the budget is spent and the channel attribution has usually gone fuzzy. CPQA gives you a read within days of launch, while there is still something left to change.
Run both. Use CPQA to steer while the campaign is live, then CPH afterwards to confirm the full funnel held up.
For high-volume frontline hiring, the criteria that hold up are binary. Pass or fail. No scoring, no judgment call, no room for a recruiter to wave through someone who's close enough.
Categories that work across logistics, manufacturing, retail and care in Benelux and the wider European market:
Every one of those has a yes or no answer taken straight from the screening responses. That's the test. If you can't answer a criterion pass or fail from what the candidate submitted, it doesn't belong in your CPQA definition.
Which rules out "strong communicator" and "culture fit". Both are real things recruiters assess later. Neither can be verified in a screening flow, and both will hand you a different CPQA depending on who did the screening that day.
Wonderkind's product splits into three modules, Attract, Qualify and Deliver. Each one moves a different part of the CPQA calculation.
Attract runs always-on social job ads pulled from your job feed and distributes them where candidates already spend their time: Instagram, Facebook, TikTok and the rest. This stage sets how many people enter the funnel at all.
Qualify applies your criteria on mobile, before anyone reaches the ATS. Every candidate answers the same questions against the same agreed criteria, and the ones who fail don't proceed. This stage controls the denominator. Raise the share of screened candidates who pass and your CPQA falls without you spending a euro less.
Deliver is where billing and counting meet, and Wonderkind's terms set the bar in four parts. A candidate counts once they complete the application flow, pass every knock-out question you configured, meet the job-description criteria you configured, and land in your ATS. All four, or it isn't billable. If you think an outcome was counted wrongly, you have ten business days from the dashboard report to dispute it.
Which answers the denominator question contractually rather than conversationally. An abandoned application never satisfies condition one, so it can never reach your invoice.
That's why CPQA reads earlier than cost per hire. The first delivered cohort gives you a number within days of launch, early enough to shift budget between channels or tighten a criterion while the campaign is still running.
Changing the rule mid-campaign. Redefine qualified after launch and every comparison to earlier data is void. Treat any change as the start of a fresh measurement period.
Counting partial applications. Someone who started the flow and left didn't qualify. Including them pads the denominator and flatters the result.
Attributing spend inconsistently. Agency fees inside one channel's total and outside another's makes the two CPQAs incomparable. This bites hardest on European campaigns running across programmatic social, job boards and third-party aggregators, each billed on its own terms.
Pooling roles with different criteria. A warehouse operative role in Amsterdam and a night shift packing role in Antwerp rarely share a qualification bar. Average them into one CPQA and you've averaged two different standards. Measure them separately, or confirm the criteria match before you combine them.
None of these produce an obviously wrong number. They produce a number that looks right, which is the more expensive failure.
No, and you should be suspicious of anyone who offers you one. CPQA moves with the role, the labour market, the location, how tight your criteria are and which channels you bought. A logistics operator hiring forklift drivers in a competitive Dutch industrial region will not see the number a retailer hiring sales assistants in a quiet market sees. Build your own baseline by role family, then improve against that.
At the single ATS step you agreed before launch. Recruiter Accepted, Submitted to Hiring Manager and Interview Scheduled are the common choices. Whichever you pick, it has to represent a completed screening flow checked against every agreed criterion, and it can't be a step a recruiter advances people through on instinct.
You're buying volume that doesn't meet the bar, and paying twice for it: once in media, once in screening time. Tighten the audience targeting first. Then check whether the job ad states the hard requirements plainly enough for unsuitable candidates to rule themselves out before they apply. Most of the time the ad is the cheaper fix.
No. They sit at different points in the funnel and answer different questions. CPQA tells you what you spent per interview-ready candidate while you can still change something. CPH tells you what the finished hire cost. You need both.
Write the qualification criteria down before it launches. Make each one binary. Agree the single ATS step where a candidate gets counted. Then calculate at channel level as well as campaign level, so you can see which channels produce interview-ready candidates and which produce screening work.
Most teams discover their CPQA after the money is gone. Wonderkind Deliver sets it first, as a fixed price per qualified candidate delivered into your ATS.
