What the work actually involves

A task designer writes a scenario — say, an analyst asked to compute repeat purchase rate from a store's order data, or to reconcile discount attribution across a Plus store's markets. Your job is to read it against reality and say where it breaks. Does the scenario assume a field exists on the Order object that doesn't? Does it treat `total_price` as net of refunds? Does it call something "conversion rate" while describing sessions-to-checkout? You return written feedback naming the divergence, why an operator would never frame it that way, and what the corrected version should say. Some engagements add short verbal review calls with the customer's task team.

Expect roughly 5–10 hours a week, remote and largely asynchronous, with occasional scheduled calls. Volume tends to arrive in batches tied to the customer's task-authoring cycles rather than evenly across the week.

What the screen is looking for

micro1 runs an AI interview first, then a recruiter and team conversation. The AI stage will push on specifics: which data objects you've worked with directly, how a metric is defined and where definitions diverge between Shopify's native reports and a warehouse query, what breaks at Plus scale. Vague ecosystem familiarity reads badly here — the follow-up questions go one layer deeper than most candidates expect, and answers that stay at the level of "I've managed Shopify stores for years" tend to stall.

  • Depth over breadth: hands-on years across store ops, development, or analytics
  • Fluency in the data model — orders, products, customers, reporting — including its quirks
  • Analytics grounding: cohort, retention, conversion, revenue measurement
  • Ability to write feedback someone else can act on without a follow-up thread

Pay and logistics

Observed rates on comparable micro1 expert engagements fall in the $60–100/hr range; the actual offer depends on the customer project and your depth. No AI or ML background is required or expected. Multi-store, agency, or consulting exposure helps, as does SQL or a BI tool like Looker, but neither is a gate.