The useful question is not what screening software costs per month. It is what the sift costs you per hire today, since that is what the software has to beat. Most organisations have never calculated it, and so software conversations get stuck on licence fees rather than on the thing the licence is meant to replace.
What follows is a model you can run on your own figures in about twenty minutes. We publish the assumptions so you can disagree with them.
Start With What the Sift Costs Now
Four inputs, all of which you either know or can estimate honestly.
- Applications per role. Use your actual figure. For context, UK graduate vacancies average around 140 applications, and Greenhouse reported 244 applications per job across more than 6,000 companies in 2025, up from 116 in 2022.
- Minutes per application. Be honest rather than aspirational. Published estimates of CV review time vary enormously, from a few seconds for a first-pass scan to well over a minute for a considered read, and no authoritative single figure exists. Use two numbers: a scan time for the whole pile and a proper read time for the subset that survives.
- Loaded hourly cost of the reviewer. Salary plus employer National Insurance, pension and overhead. For an in-house recruiter this is commonly somewhere between £25 and £45 per hour fully loaded, but use yours.
- Roles per year. Filled, not opened.
A Worked Example
Take 200 applications per role, a 30 second first pass on all of them and a 4 minute proper read on the 40 that survive. That is 100 minutes plus 160 minutes, so roughly 4.3 hours per role. At £32 per hour loaded, about £139 per role in direct sift labour. Across 60 roles a year, roughly £8,300.
That figure is almost always lower than people expect, and it is not the real cost. The real cost sits in the two lines nobody puts in the model.
The Two Costs That Dominate
Bad Shortlists
Every unsuitable candidate who reaches interview consumes a hiring manager's time, usually at a higher loaded rate than the recruiter's, and frequently two or three people's time simultaneously. An hour-long panel interview with three managers at £60 per hour loaded is £180 of time for one candidate. Two avoidable interviews per role across 60 roles is £21,600, more than twice the entire sift labour cost.
Wrong Hires
This is the line that changes the arithmetic completely, and it changed again on 1 January 2027, when the unfair dismissal qualifying period fell from two years to six months and the cap on the compensatory award was removed. The window for correcting a hiring mistake cheaply is now roughly two quarters, and the downside beyond it is no longer bounded by a statutory maximum.
We are not going to put a number on a wrong hire, because credible estimates vary so widely that any single figure would be decoration. What is defensible is the structure: recruitment cost, onboarding and training, the manager time absorbed, the output not delivered, the cost of replacing, and a legal exposure that is now materially larger and harder to predict. Model it with your own finance team rather than borrowing a statistic.
What to Compare Against
Once you have your current cost, compare on cost per shortlist rather than cost per licence. The calculation is the annual software cost divided by the number of shortlists produced, set against the labour and rework it displaces.
Ask vendors for pricing expressed in a unit that maps to your volume. Per user per month is convenient for a vendor and tells you nothing about cost at 200 applications versus 2,000. Credit or volume-based pricing at least moves with the thing driving your cost.
Also price the ceiling honestly. Ours is 500 CVs per screening batch with 50 files per upload request, which is ample for most UK campaigns and a genuine constraint if you run larger single batches. Every vendor has a ceiling somewhere; ask where.
The Free Tier Question
Free tiers are useful for evaluation and are rarely a plan. Read the actual limits rather than the headline, because the constraint is usually monthly volume or batch size rather than features, and the point at which a free tier stops working is often lower than a first campaign.
The honest way to use one is as a test: run a role you have already filled, see whether the person you hired appears in the shortlist, and read the rationale for two or three rankings. That tells you whether to pay for it.
What the Model Should Not Claim
Three things worth refusing to assert, including about our own product.
Percentage time savings quoted without a stated baseline are meaningless, since the baseline is where all the variance lives. Cost per hire comparisons across organisations are close to useless, because definitions differ so widely. And no screening tool reduces the cost of a wrong hire to zero, because human judgement remains in the decision and should.
What a screening layer can credibly change is the consistency of the sift across a large pile, the proportion of interview time spent on candidates who can actually do the job, and whether you can explain a ranking afterwards. Those are the numbers to hold a vendor to.
Frequently Asked Questions
What Is a Reasonable Cost Per Shortlist?
It depends entirely on volume and role value. A better test is whether the cost is a small fraction of the interview time it saves, which for most mid-market UK hiring it should comfortably be.
Should We Include Recruiter Salary in Cost Per Hire?
Include it in the sift model, because that is the labour being displaced. Be careful comparing your total cost per hire to a benchmark, because almost nobody defines it the same way.
Is Agency Fee Avoidance a Fair Saving to Claim?
Only where you genuinely stop using an agency for that role. Bringing volume screening in-house can reduce agency reliance, but claiming a saving you have not realised makes the business case brittle.
How Do We Justify This to Finance?
Lead with interview time displaced, which is concrete and calculable, and treat improved hire quality as a stated but unquantified benefit. A business case built on an invented hire-quality figure tends to collapse under questioning.
What If Our Volume Is Low?
Then the labour case is weak and you should say so. At low volume the argument is consistency and defensibility rather than cost, and those may still be worth paying for in regulated hiring.
Run Your Own Numbers
Take one role you filled recently. Count the applications. Time yourself reading twenty of them properly. Multiply out. Add the interviews that should not have happened. That figure, not a licence fee, is what you are deciding about.
If the answer is that your sift costs very little and produces good shortlists, do not buy anything. That is a legitimate outcome of the exercise and we would rather you reached it honestly.
This article sets out a model with stated assumptions rather than benchmark claims. Figures attributed to third parties should be checked against current publications.
Sources
Department for Business and Trade. Implementing the Plan to Make Work Pay and Employment Rights Act: Timeline Update.
https://www.gov.uk/government/publications/implementing-the-plan-to-make-work-pay-and-employment-rights-act/plan-to-make-work-pay-and-employment-rights-act-timeline-update
Department for Business and Trade. Unfair Dismissal Changes: Summary of Stakeholder Roundtables.
https://www.gov.uk/government/publications/summary-of-stakeholder-roundtables-on-unfair-dismissal-changes/unfair-dismissal-changes-summary-of-stakeholder-roundtables
Information Commissioner's Office. Employment Practices: Recruitment and Selection.
https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/employment/recruitment-and-selection/
Greenhouse. Greenhouse Pricing.
https://www.greenhouse.com/pricing
Workable. Workable Pricing.
https://www.workable.com/pricing
InsightCircle
