Take-Home Assessment Cost in 2026: Hidden Review Hours

The completed-submission number: $85 to $145.

A take-home is the deceptively cheap interview stage. The per-submission cost is the most quoted number, and it is the one that hides the most. Per completed submission, the math lands at $85 to $145 in 2026, driven by reviewer time at engineering loaded rate, platform fees amortised, and a small recruiter coordination overhead. That number assumes the candidate returns the work.

The reality of take-home funnels is that 30 to 60 percent of assigned take-homes never come back. The platform fee for assignment is still paid. The reviewer time is not consumed, but the coordination time still was (the candidate was scheduled, the email was sent, the follow-up was written). Allocating those costs to the candidates who did return pushes the true per-completed cost to $120 to $200, materially above the headline figure most calculators publish.

Sources for completion rate: Greenhouse Hiring Maturity reports and aggregated Triplebyte legacy benchmarks consistently put the drop-off rate in the 30 to 60 percent range, with the spread driven by take-home length, candidate seniority, and whether the take-home is paid. As of May 2026.

The line items behind a completed take-home.

Default scenario: 2-hour take-home for a mid-level engineer, reviewer is a senior engineer at $122 per hour loaded, platform is one of the major coding assessment tools on a mid-tier seat.

Component	Minutes	Loaded $/hr	Cost
Recruiter coordination (assign, follow-up)	15	$45	$11.25
Platform fee (amortised per use)			$8 to $25
Reviewer first-pass scoring	25	$122	$50.83
Reviewer written debrief and ATS notes	12	$122	$24.40
Subtotal, per completed submission	52 + ext		$94 to $112
Plus 1.3x context-switch (reviewer)			+$15 to $23
Plus drop-off allocation (40% rate)			+$25 to $40
True per-completed cost			$130 to $175

The drop-off allocation line is the one that separates honest take-home accounting from the marketing claims of "take-homes save you 50 percent on technical interview cost". They do, until you count the candidates who never returned and the coordination time you spent on them anyway.

Drop-off rates by take-home length.

The single biggest controllable variable in take-home cost is length. Drop-off climbs sharply with stated effort, particularly above 3 hours. Aggregated benchmarks across multiple published sources point to the rough shape below.

Stated take-home length	Typical completion rate	Per-assigned $	Per-completed $
30 to 60 minutes	75 to 85%	$60 to $80	$80 to $105
60 to 120 minutes	55 to 70%	$85 to $115	$130 to $175
2 to 3 hours	40 to 55%	$110 to $145	$210 to $285
3 to 5 hours	25 to 40%	$135 to $175	$340 to $560
5+ hours	10 to 25%	$160 to $210	$650 to $1,400

A 5-hour take-home with a 15 percent completion rate is a $900 per-completed cost. At that price point, a 60-minute synchronous technical screen at $185 is dramatically cheaper per qualified signal generated. Many teams default to long take-homes for the perceived signal depth and never run the per-completed math. The math does not support the long take-home for almost any role.

Paid versus unpaid take-homes.

Paid take-homes pay candidates $75 to $200 per submission, typically through a third- party gift card or a contractor invoice. The case for paid take-homes is the asymmetry-of-effort argument: senior candidates with options resent unpaid work, and paid take-homes signal a serious offer process. Buffer, GitLab, and a few others have published case studies showing completion rates climbing 15 to 30 percentage points when the take-home is paid.

Per-hire net cost can shake out lower despite the per-submission outlay, because completion rates rise and reviewer time is no longer wasted on coordination for candidates who never return. The math depends on the candidate pool. For senior and staff hires sourced through cold outreach, paid take-homes are often cheaper net. For inbound mid-level hires who completed an application already, paid take-homes add cost without much marginal benefit.

A useful framing: pay the take-home if you would feel embarrassed asking a senior engineer at a competitor to do it unpaid. That heuristic captures most of the asymmetry argument and tracks the published completion-rate impact reasonably well.

When the take-home replaces a stage versus adds one.

The biggest cost discipline question for any team adopting take-homes is whether the stage replaces something else or stacks on top. If a take-home replaces a 60-minute technical phone screen, the per-candidate cost drops modestly (with the drop-off caveat applied) and signal coverage shifts from synchronous problem-solving to async code quality. Net cost win in many cases.

If the take-home stacks on top of the technical phone screen (some teams use both for mid and senior hires), the per-candidate cost climbs by the full take-home amount and the funnel narrows because more candidates drop. Total per-hire cost rises 15 to 30 percent versus running either alone. Teams adopting take-homes for the first time often stack them by default because dropping the existing synchronous stage feels risky. That risk-aversion costs $1,500 to $3,000 per hire that nobody allocates.

For the broader hiring-channel cost view, see techhiringcost.com. For the interview-process per-stage view, this site is the source.

Reviewer time discipline.

The single biggest cost variance inside the take-home stage is reviewer time per submission. A loose rubric and an open-ended "judge holistically" instruction can push reviewer time from the budgeted 30 minutes to 60 to 90 minutes, doubling per-submission cost without producing better signal. Greenhouse Hiring Maturity studies put rubric discipline as the highest-correlation factor with reviewer time variance.

A structured rubric with explicit pass/fail criteria across 4 to 6 dimensions (correctness, code quality, error handling, scope discipline, documentation, testing) tends to land reviewer time consistently around 25 to 35 minutes per submission. A rubric that names "show your judgement" or "calibrate to the role" without explicit criteria typically runs 45 to 75 minutes per submission, and the variance across reviewers is wider, which feeds the calibration debt downstream.

The per-submission cost saving from rubric discipline is modest in dollar terms ($30 to $80) but the calibration consistency it produces reduces re-runs and re-reviews, which is where the larger downstream savings hide. Most teams underinvest in rubric design relative to the cost it controls.

Run your own numbers.

Drop your take-home length, completion rate, and reviewer loaded rate into the calculator for per-hire cost.

Run the calculator

Frequently asked questions

What does a take-home assessment actually cost?

Per completed submission, $85 to $145 in 2026. Per assigned take-home (which includes the 30 to 60 percent that never come back), the true allocated cost climbs to $120 to $200 because the reviewer time and platform fee allocated to non-returners are sunk. Per-hire, at a funnel volume of 6 to 10 completed reviews, total take-home cost lands around $500 to $1,500 depending on role level.

Why are take-home drop-off rates so high?

Two reasons. First, take-homes signal a low-priority candidate fit (strong candidates with multiple options resent the unpaid work asymmetry, particularly mid-senior level). Second, take-homes that exceed 3 hours of work see drop-off rates above 50 percent because candidates do the cost-benefit calculation and walk. Greenhouse hiring-maturity data and Triplebyte legacy benchmarks both put the inflection point around 2 to 3 hours of effort.

Is a take-home cheaper than a synchronous technical interview?

Per candidate, yes: a take-home runs $85 to $145 versus $145 to $220 for a live screen. Per qualified signal, often no, because drop-off rates eat the per-candidate saving. The true comparison is reviewer time per qualified signal generated. For roles where async work is the main job (backend, data engineering), take-homes are typically cheaper net. For roles where collaboration and live problem-solving matter (frontend with design partners, IC product), synchronous wins.

Should we pay candidates for take-homes?

Paid take-homes ($75 to $200 per submission, typically) increase completion rate dramatically (15 to 30 percentage points in published reports from teams like Buffer, GitLab, and others) and reduce the asymmetry candidates resent. Per-hire net cost is often similar or slightly lower despite the per-submission cash outlay, because completion rates rise and reviewer time is no longer wasted on non-returners. Paid take-homes also defuse the strongest objection to async assessment.

How long should a take-home take?

Cap at 2 hours of estimated effort. Anything longer triggers cost-benefit drop-off from strong candidates and a quality-of-life signal that hurts brand. Tight scoping is the design discipline that matters more than the platform choice. A 2-hour take-home with crisp acceptance criteria and a reviewer rubric outperforms a 6-hour open-ended take-home on signal quality, completion rate, reviewer time, and candidate experience.

Does using a platform like HackerRank or CodeSignal change the math?

Platforms add a per-seat or per-test fee ($8 to $40 amortised per use depending on tier) but reduce reviewer time materially because auto-scoring handles the first-pass filter. The net effect is usually cost-neutral to slightly cost-positive at scale, with the real benefit being consistency of scoring rather than dollar savings. The cost case for a platform peaks above 200 assessments per year.

What a take-home assessment costs in 2026.