The email arrives on a Thursday afternoon: 'Great chatting with you — next step is a take-home project. Should take about 6-8 hours.' No compensation mentioned. No deadline flexibility. Just a GitHub repo link and a due date three days out.
You've been here before. Maybe you did the last one and got ghosted anyway. Maybe you're mid-contract, billing 40 hours a week, and '6-8 hours' is a polite fiction for 12. The question isn't whether take-homes are annoying. It's whether this specific one is worth your time — and if so, on what terms.
Here's the framework. Use it before you open the assignment, not after you've already sunk three hours into it.
The four questions that decide everything
Before you touch the keyboard, answer these:
- How long is it, really? Not the stated estimate — the actual scope. A 'quick data pipeline exercise' that requires setting up infrastructure, writing tests, and documenting design decisions is not two hours. Read the requirements literally and estimate honestly.
- Does it resemble billable client work? A scoped feature build, a working API, a dashboard — that's deliverable-shaped work product, not an assessment. Compare that to a bounded algorithm problem or a bug-fix exercise, which tests judgment without asking you to build something they could plausibly reuse.
- Has a human spoken to you yet? A take-home dropped before any live conversation is a filtering tool applied at scale, often to hundreds of candidates. A take-home requested after a real conversation with a hiring manager or technical lead is a much stronger signal of genuine interest.
- What stage of the loop is this? Early-stage take-homes replace a phone screen. Late-stage take-homes replace a final round. The later it sits in the process, the more it should look like a targeted exercise tied to the actual role — not a generic template.
Read the room: what the answers tell you
Run your answers through this table before you commit.
| Signal | Green light — proceed | Red flag — negotiate or decline |
|---|---|---|
| Time estimate | 2-3 hours, clearly scoped | 6+ hours, vague scope, 'take your time' |
| Resemblance to real work | Bounded problem, synthetic dataset | Full feature build using their actual stack or proprietary-adjacent logic |
| Human contact so far | Hiring manager or lead already spoke with you | Recruiter-only contact, auto-generated invite |
| Loop stage | Mid-loop, tied to a specific role you understand | First step for every applicant, no context given |
Two green lights is workable. Three or four red flags is a strong signal to negotiate the terms or pass entirely.
Should I do a take-home test at all?
This is the question everyone actually wants answered, and the honest answer is: it depends on your leverage and your calendar, not on some universal rule that take-homes are always exploitative or always fine.
If you're between contracts with bandwidth to spare and the role is genuinely appealing, a well-scoped 2-3 hour exercise is a reasonable ask — it's roughly the same time cost as a technical phone screen, just asynchronous.
If you're currently billing 40 hours a week, have three other processes in motion, and the assignment smells like unpaid labor dressed up as a hiring step, your calculus changes. Your time has a market rate. Companies that forget that during interviewing tend to forget it during rate negotiations too — that pattern is worth noticing early.
Script one: timebox the work and say so upfront
You don't have to silently comply or silently refuse. Say what you'll actually do, in writing, before you start:
'Thanks for sending this over. I've got about 2.5 hours I can dedicate to it this weekend given my current schedule. I'll timebox my work to that window and submit whatever's complete at that point, along with notes on what I'd do next with more time. Let me know if that works on your end.'
This does three things. It sets an honest expectation instead of you silently rushing or silently blowing past their estimate. It signals you respect your own time without sounding combative. And it gives the interviewer a chance to push back if 2.5 hours genuinely won't be enough to evaluate you fairly — which is useful information either way.
Script two: propose a live exercise or a paid pilot instead
For senior consultants, especially those with a portfolio or a strong reference chain already, it's fair to counter-propose entirely:
'I'd rather not commit unpaid hours to a take-home at this stage, but I'm glad to do a live pairing session or whiteboard walkthrough of a similar problem — that tends to give both sides a clearer read in less time. Alternatively, if there's a small paid pilot or trial engagement that makes sense for this role, I'm open to discussing that.'
This works especially well later in a loop, when you already have a real relationship with the hiring manager. It rarely works as your opening line with a recruiter you've never spoken to — save it for after a human has vouched for the conversation.
When to just decline
Sometimes the right move is a polite no. Decline when the assignment is a full working feature using their actual production stack, when no human has spoken to you at any point, when the estimate is dishonest on its face, or when the company has a public reputation for using 'candidate projects' as free labor. None of that requires an angry email. A short, professional pass — 'this isn't the right fit for my current bandwidth' — closes the loop cleanly and keeps the door open if the role or process changes later.
If you want a second opinion on a specific assignment before you commit hours to it, the team at Josh Pros LLC has seen enough of these across different clients to help you read the signals fast. Email contact@joshpros.com or visit https://joshpros.com and we'll talk it through.
#TakeHomeAssignment #TechInterviews #ContractConsulting #JobSearchTips #InterviewNegotiation #TechContractors #ITConsulting #CareerAdvice #HiringProcess #JoshProsLLC #TechnicalScreening #ContractWork
Talk to a real recruiter, not a bot.
We'll tell you the rate, the client, and the terms before you interview. And if we're not the right fit, we'll say so.