What the work actually involves
You will be handed artefacts that look like the documents a retail or clienteling function produces: client outreach sequences, CRM note summaries, appointment briefs, styling recommendations, seasonal sell-through commentary, loss-prevention or returns write-ups, store performance decks. Some are model-generated; some are prompts you are asked to answer yourself as a reference standard. Your job is to judge them the way a seasoned professional would — is the clienteling advice something you would let a sales associate send to a top-tier client, does the Excel logic behind the sell-through figure hold, is the tone right for a luxury house versus a mass-market chain — and then write feedback structured enough that an engineer who has never set foot in a boutique can act on it.
Most tasks are graded not on your verdict alone but on the reasoning trail. A rating with no cited line, no stated rubric criterion, and no explanation of what a correct version would look like is close to worthless to the project.
What the screen looks for
- Verifiable specifics from your own history. Which brands or categories, what price point, whether you carried a personal client book and how large, whether you owned a target and hit it.
- Depth under follow-up. The screener will push: how you handled a lapsed VIP, what your actual outreach cadence was, how clienteling behaviour differs between a $400 and a $40,000 transaction.
- Evaluation judgment. Can you distinguish fluent-but-hollow copy from substantively correct work, and can you apply a rubric you disagree with rather than substituting your own taste?
- Tooling. Excel and PowerPoint come up because much of the corpus is spreadsheets and decks; you should be comfortable checking a formula and reading a chart critically.
Logistics
Fully remote and asynchronous, drawn from a task queue rather than scheduled shifts. Contributors are based in the US, Europe, UK, or Canada. Volume fluctuates with project demand — expect bursts rather than a steady weekly load, and treat the pay band as observed rates on comparable Mercor projects rather than a guarantee of hours or rate.