Hiring A Technical Artist

What You're Actually Buying

Here's what you're really paying for. A good TA turns "the art is done but it doesn't work in the game" from a monthly crisis into a non-event. In practice that looks like: time back (a tool that saves 8 artists 20 minutes a day buys back 13 hours a week, basically a part-time hire who never calls in sick), calmer deadlines (asset validation catches problems the day they're made, not the night of the milestone build), frame rate that holds (because someone actually owns the budgets), and fewer arguments (art-versus-engineering negotiations that end in a spec instead of a grudge).

The failure mode of not having one: your most senior artists quietly become bad part-time TAs - losing production time to write fragile scripts - while your engineers get interrupted for "art problems" they're not close enough to solve well. You're already paying for tech art; you're just paying two departments to do it badly.

When You Need One

• ~8-10 artists is the classic first-TA threshold - enough repetition that tooling compounds.
• Any custom pipeline (your own engine, unusual DCC mix, heavy outsourcing intake) moves the threshold earlier.
• Symptoms checklist: artists keep a wiki of manual export steps; every milestone has a "get the art in" crisis; performance is discovered rather than designed; one engineer spends Fridays on "art requests." Two or more of these - you needed a TA last quarter.

Writing A Posting That Doesn't Scare Off The Good Ones

The classic broken TA posting asks for a senior shader author, a pipeline engineer, and a character rigger in one body at one salary. Those are three different specializations; the posting reads as "we don't know what we need," and experienced candidates - the ones with options - skip it.

The fix is one honest sentence before you write requirements: "What will this person do in their first three months?" Fix exports and build validation -> you want a pipeline TA. Own the material library and framerate -> shader/tech-art generalist. Character deformation quality -> rigging/tech-anim. Write the posting around that answer:

• Must-haves: five or fewer. Python + one relevant DCC + your engine covers most TA roles. Every extra "required" bullet costs you real applicants - especially the excellent self-taught ones who take requirement lists literally.
• Name the first project. "Your first quarter: rebuild the Blender->Unreal export path" attracts exactly the people who like that problem - self-selection does your screening.
• Degree requirements filter out the wrong people. A large share of working TAs are self-taught artists who learned to code. Portfolio over pedigree, always.
• Post the salary band. TA salaries are opaque (we wrote a whole page on it); ranges get you candidates who'd otherwise not risk the process.

Evaluating Candidates When You're Not Technical

• Look for breakdowns, not beauty. A TA portfolio should explain how and why, with numbers ("export time: 15 min -> 30 sec"). If it's only pretty renders, you may be looking at an artist who applied to the wrong posting.
• The one interview question that always works: "Tell me about a tool you built that people actually used - and what they complained about." Real tool-builders light up and get specific about the complaints. That specificity is the credential.
• Put one artist and one engineer in the loop. The artist tests whether they listen; the engineer tests whether the technical depth is real. A candidate both groups like is the entire job description verified.
• Time-box any test to a respectful size (a few hours, scoped) and consider paying for it. The best candidates decline week-long free labor - filtering for desperation is a strange hiring strategy.

Keeping Them

TAs leave for predictable reasons, all cheap to prevent. Give the work a lane: a TA who reports into art but takes tickets from everyone becomes the studio's ticket queue; give them a backlog, a prioritizer, and the authority to say "next sprint." Fund the multiplier: tool maintenance and pipeline work looks like "not making the game" to naive roadmaps - it's why the game ships; budget for it explicitly. Level them fairly: TA sits between art and engineering ladders and often falls into the gap; decide which ladder (engineering-adjacent usually retains better, see salary) and be consistent. And send them to GDC - the tech-art community is tight, generous, and your TA will return with a year's worth of stolen good ideas. Cheapest R&D you'll ever buy.