The decision: screening at volume
Four thousand applicants. One bar. No extra recruiters.
Screening at volume is the decision of how to cut a large applicant pool to an interviewable shortlist without interviewing everyone. It is answered by putting the same measurement in front of every applicant — English, role ability and workplace judgement — scored automatically on submission, so recruiter hours go only to the people above the bar.
At a glance
Screening Applicants at Volume: the facts, with their units
| The decision | Reduce a large applicant pool to an interviewable shortlist without interviewing everyone. |
|---|---|
| Cost per applicant | US$7.50–13 — a screening set runs 15–26 credits at US$0.50 a credit. Volume packs discount up to 13%. |
| Recruiter time per applicant | None at the screening stage. Every attempt is scored on submission and ranked automatically; recruiters open only the shortlist. |
| Delivery | One share link or QR code per wave. No candidate account, no card, any phone; the player is resumable on a weak connection. |
| Time to a ranked list | As fast as candidates submit — the ranking builds live during a walk-in day rather than after it. |
| Cost to pilot | 250 free credits on signup — 9 to 16 real candidates screened end to end before you pay anything. |
The funnel is limited by recruiter hours, not by candidate quality
A team that can run 40 phone screens a day cannot honestly evaluate 4,000 applicants, so it stops evaluating and starts rationing — first-come, nearest-to-the-site, best-formatted CV. The good candidate at position 3,200 never gets reached. Measurement removes the rationing step: everyone is assessed, and the ranking decides who gets the hour.
Speed pressure quietly lowers the bar
When a wave has to be filled by Friday, the bar bends. Because the bar lives in the score threshold rather than in a tired recruiter's judgement at 6 p.m., it stays where you put it — and when you do lower it to fill seats, that is a visible, deliberate decision with a number attached rather than something that happened to you.
Volume screens measure the easiest thing, not the deciding thing
Typing speed and a language score are cheap to measure, so they become the whole screen — and then the attrition shows up in week three, on behaviours nobody measured. Judgement under pressure, ownership and de-escalation are what separate agents who survive nesting; they belong in the screen, not in the post-mortem.
What you get
Built for screening applicants at volume
One sitting, several measures
Spoken English and listening scored by AI speech engines, written English and comprehension, typing speed and accuracy, and situational judgement — combined into one assessment set so a candidate sits once and you get several independent signals.
Share-link and QR delivery
A wave is a link. Candidates need no account and no card, sit on their own phone, and resume where they left off if the connection drops — which is what makes a job-fair queue or a provincial pool practical to screen at all.
Proctoring with evidence you can inspect
Identity-baseline photo, face-presence and additional-face detection, tab-switch and screen-capture monitoring, and a High/Medium/Low integrity band on every attempt — with the underlying evidence attached rather than summarised away.
Ranked results the moment they submit
Scores, per-section breakdowns, an AI-written summary of strengths and risks, and a live position against everyone else in the drive. Shortlisting is reading the top of a list, not opening 4,000 files.
Spoken role-play where the customer answers back
For voice roles, an adaptive mock call: a simulated irate customer that reacts to what the candidate actually says and hardens or softens accordingly, scored on ownership, de-escalation, resolution and spoken clarity, with the full transcript in the report.
Per-candidate pricing, no platform fee
Credits at US$0.50, screening sets at 15–26 credits, packs discounting up to 13%, and no seat licence or annual minimum — so a 4,000-applicant wave costs what those 4,000 sittings cost and nothing else.
How the decision is made
The sequence, in order
- 1
Draw the line before you open the funnel
Decide the pass bar per role first — spoken English band, typing floor, judgement score — and write it down. A bar set after you see the scores is not a bar; it is a rationalisation, and it is the single most common way volume screening quietly turns into gut feel at scale.
- 2
Put the same measurement in front of everyone
One sitting, one link: AI-scored spoken English and listening, written English and typing where the process needs them, customer or process judgement scenarios, and proctoring with an identity photo. Same items, same scoring, same conditions — which is what makes the comparison between two applicants mean anything.
- 3
Let the ranking build itself
Scoring happens on submission, so the ranked list exists while the drive is still running. Nobody batches CVs overnight, and no candidate waits a week to hear whether they passed a test a machine graded in seconds.
- 4
Spend the recruiter hours below the line
Interview the people the measurement ranks highest, and open the proctoring evidence on the finalists — the identity photo, the face and tab-switch flags, the integrity band — before an offer, not after a doubt.
- 5
Check the bar against what actually happened
After the cohort has been through nesting or probation, compare who cleared the bar with who is still there and performing. A screening bar that never gets checked against outcomes drifts; a bar you re-cut once a quarter is the difference between a screen and a superstition.
Compare the approaches
Four ways teams screen at volume, compared
Approach-level, not vendor-level: most volume-screening decisions are actually a choice between these four shapes. Every row carries the case for choosing it instead, because most of the time one of them genuinely is the right answer.
External facts verified at source on .
| Approach | What it measures | What it costs | When to choose it instead |
|---|---|---|---|
| CV and résumé screening | What the applicant claims, plus how well they format a document. Nothing is demonstrated. | Recruiter hours. At two minutes a CV, 4,000 applicants is roughly 133 recruiter-hours before a single conversation. | Choose it when the requirement is genuinely documentary — a licence, a specific degree, a visa status — because those are facts about paperwork and a test cannot check paperwork. |
| Phone or video screen first | Conversation and availability, for as many people as you have hours to call. | About 15 minutes per candidate of recruiter time, so the pool is capped by the calendar rather than by the applicants. | Choose it when the pool is small enough to talk to properly — under about a hundred — or the role is senior enough that the conversation is the assessment. |
| A language test on its own | English proficiency, precisely. Pearson's Versant range is the long-standing benchmark in Philippine contact centres: a 17-minute speaking and listening test, a 30-minute four-skills test and a 35-minute writing test, AI-scored with results returned within minutes (Pearson's published product pages, checked 30 August 2026; pricing is quoted on request, not published). | Published on request by the vendor. We do not restate a price we cannot verify. | Choose it when a language score alone is the requirement — a client contract that names a recognised English benchmark, or a process where English is the only variable in doubt. A dedicated language instrument is a serious instrument and this page does not pretend otherwise. |
| One combined assessment screen (AssessAll) | Spoken and written English, role ability, and workplace judgement in a single proctored sitting, with an integrity band and evidence attached to every attempt. | US$7.50–13 per candidate (15–26 credits at US$0.50), up to 13% less on volume packs. 4,000 applicants is US$30,000–52,000 at list, with no recruiter hours consumed at the screening stage. | Not the answer when your bottleneck is sourcing rather than selection. If 4,000 applicants is the problem, measure them; if 40 applicants is the problem, spend the money on reaching more people and interview the ones you have. |
Facts about other companies' products are taken from their own published pages on the date shown and are re-checked monthly; where a vendor does not publish a figure, this table says so rather than estimating one. AssessAll does not claim any accreditation, certification or regulatory approval, and a screening bar is your decision to defend — monitoring selection rates across groups, for which the four-fifths rule of thumb in the EEOC Uniform Guidelines is the usual starting check, is the employer's responsibility and not something any platform can certify for you.
Sources: Pearson — Versant corporate language testing (Philippines) · EEOC Uniform Guidelines on Employee Selection Procedures (29 CFR 1607)
The worked example: the Philippines
What screening at volume looks like in a Philippine hiring wave
The decision is the same everywhere; the constraints are not. These are the four that shape a Philippine contact-centre wave, and they are the reason a volume screen there has to run on a phone, in a queue, and on more than a language score.
Walk-in days and job fairs
A drive can put several hundred people in a queue in one morning. A QR code on the registration desk turns that queue into completed, scored, ranked sittings by lunchtime — candidates use their own phones, create no account, and recruiters interview from a list that is already ordered rather than from a stack of forms nobody has read.
Provincial and work-from-home pools
Hiring in Cebu, Clark, Davao or Iloilo, or fully home-based, means selecting people you will never meet before their start date. Proctoring evidence — the identity photo, face-presence and tab-switch flags, the integrity band — is what makes a remote result worth acting on, and the resumable player is what stops a dropped connection from destroying an attempt.
2026: fewer entry seats, and a higher bar on each one
IBPAP cut the industry's 2028 outlook in July 2026 to between US$43.3 billion and US$50.5 billion in revenue and 1.85–2.14 million workers, down from a previous target of US$59 billion and 2.5 million full-time employees, while holding 2026 at about US$42.3 billion and 1.96 million (BusinessWorld, 15 July 2026). Fewer routine entry seats does not mean less screening — it means each remaining seat carries more judgement, so the screen has to measure judgement and not only English.
The night shift and the account, not just the language
A US or ANZ account is a graveyard-shift job with an account-specific vocabulary, and both are predictable sources of week-three attrition. Shift-work adaptability and account-type readiness screens sit alongside the English and judgement components in the same sitting, so the wave is filtered on fit to the actual roster and the actual account rather than on accent alone.
Common use cases
- Contact-centre hiring waves for voice, chat and back-office processes
- Walk-in days and job fairs where hundreds arrive in a morning
- Work-from-home and provincial pools you will never meet in person
- Seasonal and ramp hiring against a fixed start date
- Campus and fresher drives with no work history to read
- Staffing-agency shortlists that have to arrive at the client pre-evidenced
Pricing for screening at volume
Volume screening is priced per candidate in US dollars: 1 credit = US$0.50, a screening set runs 15–26 credits (US$7.50–13), and credit packs (250 / 1,000 / 5,000 at US$125 / 465 / 2,175) discount volume by up to 13%. There is no platform fee, no seat licence and no annual contract, so the cost of a wave is exactly the cost of the sittings in it. New organisations get 250 free credits — enough to screen a first small batch end to end, read the reports, and set your bar on real scores before you spend anything.
Frequently asked questions
How do you screen thousands of applicants fairly?+
Give every applicant the same measurement under the same conditions, set the pass bar before you see the results, and let scoring rank the pool automatically. In practice that is one assessment sitting per applicant — spoken and written English, role-relevant ability, and workplace judgement — delivered by a share link or QR code, scored on submission, with proctoring evidence attached to each attempt. Recruiter time then goes only to the shortlist the measurement produced, which is both faster and more consistent than reading four thousand CVs at two minutes each.
What does it cost to screen 1,000 applicants?+
On AssessAll, US$7,500 to US$13,000 at list: credits are US$0.50 and a screening set runs 15–26 credits per candidate depending on how many components it includes, with volume packs discounting up to 13%. There is no platform fee, no seat licence and no annual minimum, so 1,000 applicants costs exactly 1,000 sittings. New organisations get 250 free credits to run a pilot batch first.
How many applicants can one recruiter handle with assessments?+
The number stops being set by recruiter hours. Every attempt is scored automatically on submission and ranked live, so the recruiter's work moves from evaluating the pool to interviewing the top of it — the practical limit becomes how many links you can distribute and how many interviews you want to run, not how many CVs a person can read in a day.
Won't a screening test reject good candidates?+
Any bar rejects someone, including an unwritten one — the difference is that a measured bar is visible and checkable. Three practices keep it honest: measure several independent things rather than one, so a single weak module does not sink a good candidate; set the threshold before you see the distribution; and re-check the bar against what happened after nesting or probation, so it tracks the job rather than drifting. Employers should also monitor selection rates across groups; the four-fifths rule of thumb in the EEOC's Uniform Guidelines is the usual starting check in the US, and no assessment platform can certify your process for you.
Can candidates cheat a high-volume online screen?+
Unproctored screens, yes — routinely, now that capable AI assistants are universal. Proctored sittings change the economics: an identity-baseline photo is captured at the start, face presence and additional faces are monitored, tab switches and screen captures are recorded, and every attempt carries a High, Medium or Low integrity band with the evidence attached for you to inspect. Spoken tasks add a second layer, because a recorded voice answering an adaptive call cannot be pasted from a chatbot.
Do candidates need an account, a computer, or a good connection?+
None of the three. A wave is delivered as a share link or QR code, candidates sit on their own phone with no account and no card, and the player is resumable — a dropped connection pauses the attempt instead of ending it. That is what makes walk-in queues, provincial pools and work-from-home applicants screenable on the same terms as everyone else.
Related reading
BPO Hiring in the Philippines
The same decision with the Philippine market's screening bundles, prices and delivery patterns spelled out — voice, chat and back-office screens per candidate in USD.
How BPO companies screen candidates in the Philippines
What the incumbent screening stack actually measures, where the mock-call layer changes the shortlist, and what a wave looks like end to end.
Reducing early attrition in Philippine BPOs
Why week-three attrition is a screening problem before it is a training problem — and which measures at screening predict who is still there after nesting.
How to compare skills assessment platforms
The checklist for evaluating any assessment vendor: what to ask about scoring, proctoring evidence, delivery and price, and which claims to verify before you buy.