Telus International AI Rater Exam Guide 2026: Pass Guidelines Part 1, 2 & 3 ($14–$20/hr)
Disclosure: Some links in this article are affiliate links. If you click and make a purchase, we may earn a commission at no extra cost to you. This does not influence our editorial recommendations - we only recommend products and services we genuinely believe in. Read our full affiliate disclosure.
The Search Quality Rater (now officially designated as AI Community Rater or Personalized Search Engine Evaluator) position at Telus International represents one of the most stable, flexible non-phone remote positions in the online economy. Raters work directly on evaluating the relevance, safety, and factual accuracy of search engine and generative AI algorithms.
However, the barrier to entry remains high: the 3-part qualification exam tests your command of Google's 170+ page Search Quality Evaluator Guidelines. Over 65% of first-time applicants fail during Part 2 or Part 3.
This operational study guide breaks down each stage of the 2026 Telus Rater examination, detailing the exact scoring rubrics, Page Quality (PQ) signals, and Needs Met (NM) logic required to pass on your first attempt.
1. Exam Overview and Timeline Structure
Once your resume passes automated screening and you complete identity verification, Telus issues an exam link with a strict 7-day window.
You must pass each part sequentially. The system grades your answers automatically; if you pass Part 1, you instantly unlock Part 2. If you fail any section, the exam concludes immediately.
2. Part 1: Theory and General Guidelines Rules
Part 1 consists of 32 open-book multiple-choice questions. It tests whether you understand definitions, terminology, and standard evaluation principles.
Key Concepts Tested in Part 1
- MC vs SC vs Ads: Main Content (MC): Any part of the page that directly helps the page achieve its purpose (text, images, interactive tools). Supplementary Content (SC): Helpful navigation links, related article suggestions, or site architecture that enhances user experience. * Advertisements (Ads): Monetization units. Ads do not automatically make a page low quality unless they disguise themselves as MC or impede content readability.
- YMYL (Your Money or Your Life): Topics that could impact health, financial stability, civil rights, or physical safety require the highest standard of formal E-E-A-T.
- E-E-A-T Dimensions: Experience (first-hand practice), Expertise (formal knowledge), Authoritativeness (industry recognition), and Trustworthiness (the overarching anchor).
3. Part 2: Page Quality (PQ) Evaluation
Part 2 requires you to assess 7 live websites on a 5-tier scale: Lowest, Low, Medium, High, and Highest. You must provide written justification explaining your reasoning.
Passing Strategies for Part 2
- Verify Creator Identity: Check the "About Us" page, author bio, and privacy policy. If a website offers medical dosage advice but provides no author credentials or publisher background, it defaults to Low or Lowest.
- Evaluate Purpose: What is the page trying to accomplish? A satirical website or entertainment blog can achieve a High rating if it fulfills its entertainment purpose without claiming factual accuracy.
- Beware of Negative Reputation: Conduct a reputation search on Google:
"[Website Name]" -site:[WebsiteDomain.com] reviews. Significant BBB complaints or news reports regarding fraudulent billing justify a Low/Lowest rating.
4. Part 3: Needs Met (NM) Rating System
Part 3 is where most applicants fail. You are given 7 complex search queries with multiple result blocks. You must assign two ratings to each block: Page Quality (PQ) and Needs Met (NM).
The 5 Needs Met Ratings
| Needs Met Tier | Criteria | Query Example |
|---|---|---|
| Fully Meets (FM) | Completely resolves the query with zero extra search. Reserved almost exclusively for unambiguous, fact-based Know Simple queries. | Query: how tall is the eiffel tower -> Answer block: 330 meters |
| Highly Meets (HM) | Thorough, authoritative, highly satisfying result with minimal limitations. | Query: symptoms of type 2 diabetes -> Mayo Clinic comprehensive guide |
| Moderately Meets (MM) | Adequate and relevant, but lacks depth, has minor outdated info, or appeals to a secondary user intent. | Query: best laptops 2026 -> Forum thread with 2 brief suggestions |
| Slightly Meets (SM) | Distantly related, lacks specific details, or addresses an obscure interpretation of the query. | Query: buy winter tires -> Blog explaining how tires are manufactured |
| Fails to Meet (FailsM) | Completely off-topic, factually wrong, broken link, or non-functional page. | Query: weather in chicago -> Page displaying hotels in Miami |
Critical Trap: Never give a Fully Meets rating to broad or ambiguous queries like "running shoes" or "healthy dinner". Users have different preferences, budgets, and dietary restrictions; no single result can completely resolve such queries.
5. Pay Rates, Work Hours, and Practical Tips
Compensation Structure
- United States: $14.00 – $17.50/hr (often categorized as W-2 with state-specific paid sick leave).
- Canada & UK: $13.00 – $16.50/hr (independent contractor or localized payroll).
- Europe & Latin America: $8.50 – $14.00/hr (standardized contractor rails).
Scheduling and Flexibility
- Minimum Hours: Typically 10 to 15 hours per week to remain in good standing.
- Maximum Hours: 20 hours per week during normal volume; frequently uncapped up to 35 hours per week during holiday search surges.
- AET Tracking: Telus enforces Average Estimated Time (AET) on every task. Submitting tasks significantly faster than AET results in quality review suspensions.
Frequently Asked Questions
In the United States, Telus AI Raters earn between $14.00 and $17.50 per hour as W-2 or 1099 workers. In international locales (UK, Canada, Europe, South America, and Asia), rates range between $9.00 and $16.00 per hour, calibrated to local economic indices.
The exam consists of three distinct parts completed over a 7-day testing window: Part 1 is an open-book multiple choice test on the General Guidelines (32 questions); Part 2 evaluates Page Quality (PQ) across 7 webpage tasks; Part 3 evaluates Needs Met (NM) across 7 multi-result search queries.
Yes, Telus frequently offers candidates a second attempt if they pass Part 1 and fail Part 2 or Part 3 by a narrow margin. However, receiving a retake invitation depends entirely on workforce demand in your specific geographic locale.
Applicants fail Part 3 because they fail to properly identify User Intent, incorrectly assigning 'Fully Meets' ratings to broad, ambiguous queries, or misunderstanding how to rate mobile device action queries (e.g., Device Action / Call intent vs informational search).
Alex Morgan is the founder and lead editor of RemoGrid. With over six years of hands-on experience in remote operations, cross-border freelance workflows, and AI tool benchmarking, Alex independently tests and audits software platforms to help modern digital workers build sustainable online income streams. He regularly reviews international payment systems (Wise, Stripe, Payoneer, local mobile wallets) and conducts real-world usability benchmarks across AI productivity tools.
