Most architects come to outsourcing after a stretch of overload: three deadlines land in the same fortnight, hiring takes months, and the work still has to go out. The temptation is to pick whoever answers first and quotes lowest. That decision is usually the one people regret.
A drafting partner sits inside your production process. Choosing badly costs more than the fee, because you end up redrawing sets you already paid for.
Start with a paid test, not a proposal
Proposals tell you how well a firm writes. A small paid task tells you how well they draw. Send one sheet or a small package, something real rather than a demo, and judge what comes back on the things that actually matter: layer discipline, annotation quality, whether the dimensions close, and whether they asked sensible questions before starting.
The questions are the tell. A team that comes back with three sharp queries about your title block and datum conventions is reading the drawing. A team that returns silence and then a set full of assumptions is not.
Ask who actually does the work
Find out who is on your project, how many years they have drafted, and whether the same people stay on it. High turnover inside a supplier shows up as inconsistency in your sheets. Ask what happens when the person who knows your standards is on leave.
Check how they handle your standards
Any competent partner should work to your office CAD standards rather than impose their own. Ask what they do when you hand over a template, and what they do when you have no documented standard at all. The right answer to the second is that they help you set one up and then follow it consistently, not that they quietly use theirs.
Understand the review step
Ask what happens between a drafter finishing and the file reaching you. If the answer is nothing, you have just outsourced drafting and insourced checking, which is not much of a saving. There should be a named senior reviewing dimensions, annotations, layers and sheet setup before anything leaves.
Get the commercial terms in writing early
You want clarity on three things before you start: what a revision includes, what counts as new scope, and who owns the files. Ambiguity here is where relationships sour. Ownership should never be a discussion, the drawings you pay for are yours.
Judge communication before you judge price
You will spend more time explaining intent than you expect. A single point of contact who understands architecture, reachable during some overlap with your working day, is worth more than a small hourly saving. Response time on a query matters more than response time on an invoice.
A short checklist
- Did they ask good questions before drawing?
- Can they work to your standards, not just their own?
- Who reviews the set before it reaches you?
- What is the revision policy in writing?
- Who owns the drawings and the models?
- Is there a named contact with overlap hours?
Run one small project through that list before you commit volume. The cost of testing properly is a fraction of the cost of choosing wrong.