Hire product designers, with judging criteria agreed up front
Send the brief and we typically reply with a shortlist within a few days. The pilot runs two to three weeks against agreed criteria.
Tell us what you need
We usually reply within 24 hours. Everything you share stays private.
Why design hires go wrong, in order
1
Nobody wrote down what good looks like.
The work starts, and the only tool anyone has for judging it is taste, so arguments only get louder and the schedule absorbs them.
- Why 2 waits for 1Every other problem here is recoverable once you can tell whether the work is right. Without that, fixing the others changes nothing.
2
The screens cannot be built as drawn.
Handover happens, your software engineers open the file, and half the states are missing: empty, error, loading, long strings, small screens. What comes back is a negotiation, not an implementation.
- Why 3 waits for 2This one is expensive and at least it surfaces. It shows up in the sprint, where somebody can see it.
3
Research left with the person who did it.
Interviews happened, findings existed, and none of it is written anywhere your team can reach. The next designer starts again and you pay for the same learning twice.
- Why 4 waits for 3You feel this on the second engagement rather than the first, which is why it is almost never priced in.
4
Accessibility gets treated as something to fix later.
Contrast, labels and alternative text are settled at design time whether or not anyone means to settle them. Deciding after the build means rebuilding it.
If you would put these in a different order, say so on the call. The order is the argument, and we would rather have that argument before the work starts than after it.
The failures a designer decides, not a developer
WebAIM scans the top million home pages every February. In 2026 it found detectable WCAG 2 failures on 95.9% of them, at an average of 56.1 errors per page. What matters for a design hire is which failures they are.
Show data table
| Item | share of home pages with this failure |
|---|---|
| Low contrast text | 83.9% |
| Missing alternative text | 53.1% |
| Missing form input labels | 51% |
| Empty links | 46.3% |
| Empty buttons | 30.6% |
| Missing document language | 13.5% |
Read that list again and five of the six are decisions somebody takes in the design file. Contrast is a palette decision. Alternative text, form labels and the accessible names on links and buttons are all decided in the file, before anyone writes code. A designer who leaves them open hands your software engineers a defect list and calls it a handover. Atyantik builds to WCAG 2.2 by default, which is why these are part of the pilot criteria rather than a later phase.
Written down before anyone starts
How the design work gets judged
Before the start
Criteria agreed in scoping, not discovered in review.
We turn your brief into three to five things the pilot must produce, each judged without appeal to taste. You sign it before anyone starts.
Pilot criteria sheet the shape of the form, not a client sheet1. Click-through prototype of onboarding, every state drawn, including empty and error. 2. Tested with five people who match your customers. 3. A written list of what stopped them.
A pilot criteria sheet, three to five lines, agreed in scoping.
During
Checked against the sheet, not against opinion.
At the end of the pilot we walk the sheet line by line, met or not met. A wrong line is rewritten for the next cycle, not quietly softened.
The same sheet, end of pilot met or not met, no partial credit1. Met. 2. Met. 3. Not met. Three of five sessions run, notes incomplete.
A line-by-line walk of the same sheet, at the end of the pilot.
After
Everything the designer made is yours.
Source files, component libraries, interview notes and findings live in your accounts from Day 1. You keep all of it if we stop working together.
What sits in your accounts from Day 1, not at offboardingsource files component library interview notes findings
Files, findings and libraries, in your accounts from the first day.
Your first month with a product designer
Days 1 to 3
The brief
Light
You describe the work, the stack it has to fit and what must be true at the end.
How it feels
Faster than expected, and slightly exposing, because writing down what good looks like is genuinely harder than it sounds at first.
What we do about it
Draft the criteria within days of your brief, send them back for you to change, and lock them only when you say they are right.
Within a few days
The shortlist
Your call
We present designers who match the stack and the domain, and you can interview them or skip it.
How it feels
Quick, and slightly uncomfortable if you walked in expecting to be sold to rather than simply shown who is available.
What we do about it
Tell you plainly when our bench does not match your brief, and recruit for the role instead of sending a near-miss.
Two to three weeks
The pilot
Real work
The designer works the actual problem against the signed sheet, with your software engineers seeing the files early.
How it feels
This is the hardest part. You are paying attention to work you cannot yet judge, and the criteria are the only thing holding the conversation steady.
What we do about it
Check in weekly against the sheet, surfacing a miss the week it happens. We put the designer in front of your software engineers early.
End of the pilot
The decision
Binary
We walk the sheet together, and on a miss we replace the designer and re-run the pilot.
How it feels
Anticlimactic, which is the point, because nobody is arguing about taste once the sheet already says what counted as done.
What we do about it
Say it first if we think the fit is wrong, at the review or before it, rather than waiting for you to raise it.
Most engagements run three months or longer, because that is when a new person's productivity curve flattens, though shorter is fine for clearly scoped work. The length follows the work, not the contract.
How long a design search actually takes
If you are weighing a search against bringing someone in on a pilot, the honest thing to say about the search is that nobody agrees how long it takes.
Show data table
| Item | reported days to hire |
|---|---|
| Recruiting from Scratch, product designers, own placements | 29 days |
| Ashby, business roles, median | 30 days |
| Ashby, technical roles, median | 40 days |
| SHRM, US nonexecutive roles, median time to fill, 2025 | 44 days |
These are industry figures about hiring an employee, and none of them is Atyantik's. What they show is that the spread between the lowest and the highest figure is fifteen days, about as long as the whole pilot. That is the argument for testing the fit on real work rather than forecasting it.
What US employers advertise for one product designer
United States, all locations
$157k to $213k · Median $180k
Recruiting from Scratch, June 2026, measured this over 1,000 recent product designer postings scraped from company career pages. It is base salary only, before employer taxes, benefits, equipment, software licences and whatever the search itself costs.
Staffing the designer, and what widens it
One designer, embedded
Two to three weeks, then ongoing
What it includes
- a product designer in your standups and your tools
- source files and research in your accounts from Day 1
- the written pilot criteria
- replacement if the pilot misses
Widens when you need customer research beyond your team, the surface spans more than one product, or WCAG 2.2 AA conformance is contractually required.
Designer plus software engineer
Two to three weeks, then ongoing
What it includes
- everything above
- a front-end software engineer building alongside the designer so buildability is tested in the same week it is drawn
Widens when the work touches a design system that has drifted from the code, or the component library needs to match what ships.
A small product team
Three months and up
Length follows the work, not the contract.
What it includes
- designer, software engineers, and QA
- the criteria sheet re-cut per milestone rather than once
- the shape described on hiring a dedicated team
Widens when you are building a new surface rather than improving one, or there is a fixed external date such as a launch or audit.
Send the brief with your role, your timeline and the stack the work has to fit. You get a shortlist and what it costs, so you can put our number next to the one above and do the arithmetic with both sides filled in.
Incredible partners who delivered a critical project on an abridged timeline. The team blew everyone away with what we accomplished.
Said about a delivery engagement rather than a design one. We are showing it because it is real and the person put his name to it, not because it proves anything about a design pilot.
Self-reported by Atyantik; not independently audited.
Atyantik engagement terms, verified 2026-05-14.
Atyantik build commitment, not an audit result.
Before you send the brief
Can we interview them ourselves?
What if design is not really the problem?
Will they work in our hours?
Who owns the design files and the research?
What if the designer is not right?
Is there a minimum commitment?
When a design pilot is the right call, and when it is not
This works when all four of these are true. If any one of them is not, we are the wrong call and we would rather say it now than three weeks in.
- You can name three to five things the first three weeks must produce, or will work them out with us first. Without that sheet there is nothing to judge the work against and we are back to arguing about taste.
- Someone on your side can give the designer real access: to your software engineers, to the product decisions, and ideally to your actual customers. A designer working from a brief alone will produce something plausible and untested.
- The product has a decided purpose. If what the thing is for is still open, design cannot close it, and we will tell you that instead of taking the work.
- You want the design and the software engineering to argue with each other early. If you would rather receive finished screens and hand them on, a fixed-scope design studio will serve you better.
If the last one is you, look at how we scope a fixed design project instead, or at our work on design systems. Neither of those needs a pilot.
Send us the brief
Tell us the role, the timeline and the stack the work has to fit. We come back within one business day, and typically with a shortlist within a few days of that. Before anyone starts we agree in writing the three to five things the first three weeks have to produce. If those are missed, we replace the designer and re-run the pilot with the next candidate.
You will speak to someone who has led real builds, not to an account manager.
- We usually reply within 24 hours
- Everything you share stays private
- You keep the files and the research from Day 1