Craft transfers instantly. Context doesn't.

You've seen the post. Somebody needs a designer for a redesign, a launch, and a pile of usability debt. Five months of work, tops. So why hire full-time? Rent a senior designer by the month instead. It ends with a question about how your team handles short-term design needs, which is LinkedIn for “DM me.”

I should love this post. I've been the rented designer more than once, and my own site offers embedded engagements. I'm the product it's selling.

I half love it. The staffing logic is right. The pricing is wrong, and the thing it prices wrong is the thing that decides whether the whole model works.

The part that's right

Design workload is phase-shaped. Redesigns end, launches end, a usability backlog shrinks if you actually work it. Hire a full-timer for a five-month bump and one of two things happens: they get absorbed into work that didn't need a new person, or they leave within a year. Both cost more than the engagement would have.

And the hiring process makes it worse. A senior search drags through sourcing, interviews, and notice periods, so your permanent hire often shows up right as the phase that justified them is wrapping. Someone who starts in two weeks and stops when the work stops isn't a corner cut. That's the right shape for the problem.

If the post stopped there, I'd have nothing. It doesn't stop there. It assumes the designer's first productive day is day one.

What it prices, and what it forgets

Look at the bullet list in any version of this post. Improve user flows. Design high-converting screens. Work with developers daily. Ship without hand-holding.

All of that is craft. Craft travels with the person. It's in my hands the morning I arrive, full strength, whatever the product is.

But design output is craft applied to context, and context doesn't travel at all. What the product is for. Who operates it, how often, under what pressure. Which complaints matter and which are noise. Why the workflow is shaped the way it is, what got tried before I showed up, what the company already decided it won't do. That half lives in the company. I have to acquire it on site, every single time, and no invoice line item mentions it.

I can't point you at a study measuring designer ramp-up, so this is mechanism, not measurement. But the mechanism isn't subtle. A design decision is a bet on what users need. The bet is capped by what I know about those users. Drop me into your product with zero context and I'll produce competent, generic work: the patterns that are right on average across every product. For a checkout funnel, average is exactly what you want. For an operational tool, average is a slow-motion disaster.

Seniority shrinks the ramp. It doesn't delete it.

The pitch uses seniority to wave the ramp away: this person is experienced, so they deliver from week one. That's backwards. Experience doesn't exempt you from learning the domain. Experience is being good at learning the domain. Sharper questions in week one. A nose for which parts are load-bearing. Knowing what you don't know yet, which is the rarest skill on anyone's CV. A ramp that eats a junior's quarter takes me maybe three weeks.

Three weeks isn't zero. On a five-month contract that's fifteen percent of the clock, gone before my first defensible decision.

Every domain I've worked in had this floor. Procurement: I had to learn how approval authority actually moves through an organization before an approval screen was worth drawing. Localization: what a deadline means when broadcast schedules are involved. Loyalty: what a mispriced reward does to a real budget. None of that arrived with me. And each one was the difference between designing the workflow and decorating it.

“High-converting screens” is the tell

Notice which product the post is imagining. High-converting screens. Funnels. Signup, landing, checkout. Those conventions have been documented for a decade, the metrics are unambiguous, and one product's checkout is mostly another's. Context is cheap there, so a stranger's craft lands in days. That's where the monthly model earned its reputation, and it earned it honestly.

Now try an operational product. An approval queue for procurement and an approval queue for media localization look identical in a portfolio and behave nothing alike, because the difference isn't on the screen. It's in escalation rules, exception volumes, compliance, and who gets shouted at when the queue jams. There, context is most of the cost of the work. A contract priced as if craft were the whole job is mispriced by exactly the context.

Same designer. Same seniority. Different product, different math. Anyone selling you the model without asking which product you have is selling it to both of you.

When it works, when it quietly doesn't

It works when the phase is real: an end you can name in one sentence, not a standing function wearing a project name for budget reasons. It works when the domain is legible, meaning decisions are recorded and somebody can hand me the why, not just the what. A company with written-down decisions cuts weeks off any ramp. And it works when someone inside owns design after I leave. Owns, not maintains. Can say why the screens are shaped this way and is allowed to revisit it.

The failures show up later, which is why the model's reputation runs ahead of its results.

Failure one: a strategy problem wearing a project deadline. Sometimes “redesign” is five months of work. Sometimes it's the company not knowing what the product should become, hoping a confident senior hand settles the question as a side effect. It won't. I can propose direction. Direction only holds if someone with tenure carries it, and the contract ends whether or not anyone picked it up.

Failure two: the unowned decision. I ship good work and leave. Six months later something needs to change and nobody can tell which parts were decisions and which were accidents. So nobody touches it, the next person routes around it, and the product grows a region marked “here be a consultant.” I've written about this disease in design systems, where it's called undocumented deviation. Same disease, different vector.

Failure three is just arithmetic. Solve every phase with a fresh engagement and you pay for the ramp every time. Three engagements, three ramps, the same domain learned three times by three strangers. “No long-term commitment” converts the one asset that compounds into a recurring cost. Past some frequency, the full-time hire this post argued against gets cheaper, ramp included, because the ramp happens once.

The exit is part of the deal

The ramp is the hidden cost at the front. The exit is the hidden cost at the back, and it's the one you actually control.

An engagement that ends with shipped screens and nothing else delivered less than it looks. The durable part is written: decisions recorded with reasons a stranger can evaluate, deviations documented as deviations, and the cut-or-postponed work parked with conditions for bringing it back. That last one matters more here than anywhere, because the person who knows why it was deferred is about to stop answering your messages.

Judge the engagement at exit plus six months, not at the goodbye demo. The demo measures craft. Six months measures whether the context I acquired on your payroll stayed, or left in my head.

Better questions than the post asks

• ·Does the phase have an end you can name in one sentence, or is it a strategy question wearing a deadline?
• ·What does the designer read in week one, and whose job was it that this exists before day one instead of being assembled during the ramp you're paying for?
• ·Which parts of your product are pattern-shaped, where senior craft lands immediately, and which are domain-shaped, where nothing useful happens until the context is in?
• ·Who owns design decisions the day after the engagement ends?
• ·What does the exit owe in writing: recorded decisions, documented deviations, parked work with re-entry conditions?
• ·Is this the first engagement or the third, and how many times have you paid for the same ramp?

The post is right that hiring isn't always the answer. But full-time versus monthly was never the real choice. The real choice is between a company that holds its own context and one that rents context acquisition over and over and books it as flexibility. Craft you can rent by the month. Context you can only build. The only open question is who you're building it in.