Lesson 5 of 9 · 5 min
Two short interviews and a paid test
Two short rounds beat one long one. The gap between them lets you think and lets the candidate prepare properly, and a candidate who clears round one and stalls in round two has just told you their experience was overstated. Trust what you saw, not what they claimed.
Round one: thirty minutes, filtering
- Their most recent Shopify role: the store, the volume, what they owned.
- One operational problem they solved, and how. Specifics are the signal; “I helped manage operations” is the absence of one.
- Three scenarios: a difficult refund, a near-stockout, a customer angry at the brand.
- Their questions for you. A strong candidate has at least three.
Round two: forty-five minutes, working session
- Share a duplicate test store, never your live admin.
- Ask them to find a specific report, then draft a reply to a real ticket.
- Talk through their approach as they go. You are watching how they think, not whether they get it right.
- Pay, hours, start date, last questions.
The paid test task
Give your two or three finalists the task from lesson two and pay for it; $25 to $75 is normal for one to two hours. Score each on three things: fidelity to the brand voice document, quality of the decisions, attention to detail. The work product tells you more than any answer in either round.
Then do not sort by price
The most common mistake at this point is sorting the shortlist by hourly rate, ascending. Sort by the test score, then negotiate within the band you can afford. A strong candidate at the top of your band beats a weak one at half the price every time; the timesheet saving is visible and the downstream cost is not.
Checked From the complete guide and the job description guide, unchanged in substance.
Both guides give the same two-round structure and the same test-task range. Nothing here needed cutting; it is procedure, and each step can be judged on its own. The only editorial change is ordering: the guides put the test task before the interviews in one place and after in another, and here it follows round two because you cannot score voice fidelity before the brand document has been discussed.