Key talent acquisition metrics that drive recruitment ROI
The hiring metrics that prove recruitment ROI follow one candidate from source to hire. Which to track, in what order, and what Côte and wagamama found.
The talent acquisition metrics that drive recruitment ROI are the ones that follow a single candidate from the channel that found them to the hire recorded in your ATS: cost per application by source, application completion, the share of applications actually seen, time to hire and cost per hire. Tracked together and by source, they tell you where your money produces hires and where it simply produces clicks, which is the only basis on which a budget can be defended or moved.
Most teams already report some of these. The trouble is that they report them from different systems, at different ends of the process, and nobody can join them up.
Why the usual numbers do not prove anything
A job board invoice tells you about clicks and applications started. The ATS tells you about hires. Between the two sits the drop-off, the people who arrived, looked and left, and it belongs to nobody. A channel with cheap clicks and very few finished applications looks like a bargain on the invoice and a loss in the hiring numbers, and without a way to connect the two you cannot say which it is.
So the first question is not which metrics to track. It is whether you can see one candidate’s path at all. If you cannot, every metric below is an estimate.
The five metrics, in the order they matter
1. Cost per application, by source
This is where ROI starts, because it measures what you paid for something useful rather than for attention. A cheap click on a channel whose visitors never finish is more expensive than it looks, and only cost per application, broken down by source, role and location, shows that.
2. Application completion
Of the people who started an application, how many finished? For frontline roles, where most candidates apply on a phone, this is usually the single biggest lever. A long desktop form loses people you have already paid to attract.
3. Applications seen
Of the applications that arrived, how many did someone actually open? In high-volume hiring this number is often lower than anyone would like to admit, and every unread application is spend with no chance of return.
4. Time to hire
How long from application to accepted offer, and where the waiting happens. For hourly roles, a candidate who waits a week has usually taken another job.
5. Cost per hire
The number finance cares about, and the one that only becomes trustworthy once the four above are measured, because they are what it is built from.
What about quality of hire?
Quality of hire and retention matter too, and they are the long-term test. But they are lagging, they are hard to attribute, and a team that cannot yet see its cost per application by source is not ready to argue about them.
What changes when you can see it
Côte is a good example of what happens when the path becomes visible. The talent team had been buying credits on job boards with little idea of return, distributing roles by hand and compiling reports from disconnected systems. After moving to a rebuilt careers site, job distribution paid on completed applications, and one view of the funnel, Côte cut its cost per application by 70% and its cost per hire by 71%, and maintained hiring volumes despite a 78% reduction in its monthly recruitment marketing budget. The 20+ hours a month the team had spent compiling reports came back too. The Côte case study sets out how.
wagamama found the same thing from the media side. With spend following live performance across channels, it reduced its cost per application by 63% year on year, to £4.89 across 23,000+ applications, and cut its recruitment media spend by 27% while staffing its new openings. The wagamama case study has the detail.
In both cases much of the saving came from moving money away from channels that did not deliver, which is only possible once you can prove which ones those are.
How to get the data joined up
The reason this is hard is structural. No ATS sees what happens before the application, and no job board sees what happens after it. The data you need sits either side of the application, in systems that were never designed to talk to each other.
Analytics and attribution closes that gap by capturing the source at the front of the process, on the careers site and in the application, and carrying it through to the hire. Because the careers site, conversational apply and job distribution all run on inploi, the attribution is recorded rather than inferred. Studio shows the funnel from first click to hire, segmented by job, location, brand or date range, and the attribution is sent back to your ATS where technically possible, so the system of record carries it and your own BI tools can use it.
What to do next
Before buying anything, try to answer one question for a single role in a single city: what did the last hire cost, and which channel brought them? If that takes more than a few minutes, or relies on a spreadsheet someone maintains by hand, that is the gap to close first.
The prove hiring ROI page explains how we approach it, and there is more on measurement in our writing on analytics and attribution.
