What the work involves

You'll be evaluating AI-generated content and reasoning against the reality of running a listed business. Typical tasks include judging whether a model's suggested reply to a customer review sounds like something you'd actually post, checking if AI-drafted business descriptions, service categories, or hours updates match how Google Business Profile actually behaves, and flagging advice that would get a listing suspended or misrepresent a business. Some batches ask you to write reference answers from scratch — how you would handle a one-star review, a duplicate listing, a photo rejection — so the model has grounded human examples to learn from.

Sessions are typically short and self-contained: a batch of prompts with rating scales, plus written justification for your scores. The justification matters more than the score. Reviewers who write "tone is wrong for a hotel — this reads like a chain apology template, and it doesn't address the actual complaint about the room" are far more useful than reviewers who leave a 3 and move on.

What the platform screens for

  • Verifiable ownership or admin access to a Google Business Profile, ideally with a year or more of active history.
  • Evidence you actually use it — responding to reviews, updating hours, posting, handling Q&A, dealing with suspensions or verification issues.
  • Industry fit. Hospitality and lodging, food and beverage, retail, and consumer-facing professional services are the priority verticals, though there's no hard quota.
  • Clear written reasoning. The screen is AI-led and will follow up on vague answers; specifics about your own listing and your own decisions carry the interview.

No technical background is required beyond competent day-to-day management of your own listing. This is a domain-experience role, not an engineering one.

Logistics

Fully remote and asynchronous. Work arrives in batches with deadlines rather than fixed shifts, so most contributors fit it around running their business — evenings, slow hours, weekends. Volume is inconsistent by nature; treat this as supplemental rather than a replacement income. The $0–60/hr band reflects observed rates across task types on the platform, not a guarantee; simpler rating batches sit near the bottom of the range and expert-written reference tasks near the top.