How to Compare VA Platforms' Skills Assessments Before You Trust the Scores

A VA platform tells you a candidate scored 87%. But 87% on what, exactly? Tested by whom, under what conditions, and against which benchmark? If you cannot answer those questions, that score is little more than a marketing number dressed up as data.

This matters more than most hiring guides admit. According toTestGorilla's State of Skills-Based Hiring 2025 report, 76% of employers now use pre-hire skills tests, and 71% believe skills-based hiring is more predictive of on-the-job success than CVs. The demand for assessed talent is real. The problem is that not every assessment is built the same way, and a high score on a poorly designed test predicts almost nothing.

The question is not whether a platform tests candidates. The question is whether their tests are worth trusting.

Before you shortlist a single candidate based on a skills badge or a percentage score, here is exactly what to evaluate.

What Makes a Skills Assessment Actually Valid?

Before comparing platforms, you need a baseline for what "good" looks like. Assessment science gives us clear guidance here.

Research published byCogn-IQ summarising decades of meta-analyses shows that work sample tests, which simulate the actual tasks a hire will perform, have a validity coefficient of up to r = .54 for predicting job performance. In plain terms: the closer a test is to the real job, the more reliably it predicts whether someone will succeed in it.

By contrast, years of self-reported experience predicts performance at only r = .18. That CV a candidate spent hours polishing? Statistically, it tells you very little.

The three things a valid assessment must do

• Test the actual skill, not a proxy. A typing speed test tells you how fast someone types. It does not tell you whether they can draft a professional email, manage an inbox, or handle a CRM. Role-specific task simulations are far more predictive than generic aptitude tests.
• Be proctored or otherwise tamper-resistant.TestGorilla's 2025 data found that 7 in 10 candidates who admitted cheating on skills tests used AI to do it. An unproctored test with no verification layer is an open invitation to game the system.
• Produce a score you can benchmark. A score without context is meaningless. You need to know what the average applicant scores, what the top performers score, and where your candidate sits relative to that distribution.

If a platform cannot clearly explain how their tests are structured, proctored, and benchmarked, treat their scores with scepticism.

Five Questions to Ask Any VA Platform About Their Assessments

Use this as your due diligence checklist when evaluating any VA marketplace that advertises tested or verified candidates.

1. What exactly does the test measure?

Push past the label. "English proficiency" could mean a multiple-choice grammar quiz or a live voice sample reviewed by a human assessor. Those are not equivalent. Ask the platform to describe the format, duration, and output of each test they run.

2. Who administers the test, and can candidates retake it?

If a candidate can retake the same test until they pass, the score reflects persistence, not ability. Platforms that allow unlimited retakes without question-set rotation are effectively handing out gold stars.

3. Can you see the candidate's actual test output?

A score is a summary. The raw output, the written response, the voice recording, the completed task, tells you far more. If a platform only shows you a number and refuses to share the underlying evidence, that is a red flag.

4. Is the test role-specific or generic?

A VA hired to manage your calendar needs different skills from one hired to handle customer support or run lead generation. A single generic "admin skills" test applied to every candidate on a platform is not meaningful differentiation. Look for platforms that match assessments to the specific role.

5. What is the benchmark, and how many candidates have taken it?

A score of 78% means nothing without knowing whether the average candidate scores 40% or 75%. Ask for the score distribution, sample size, and how often the test content is updated to prevent coaching.

Key takeaway: A platform that cannot answer these five questions confidently is not running assessments. It is running a vetting theatre designed to look credible at first glance.

The Multi-Measure Difference: Why One Test Is Never Enough

The strongest predictor of a good hire is not a single high score. It is a combination of verified signals across different dimensions.

TestGorilla's 2025 research found that 91% of employers using multi-measure testing, combining skills tests with behavioural and cognitive assessments, report making quality hires. That figure drops significantly when employers rely on a single test type. The research also found that employers who validate skills before screening CVs are more likely to make quality hires (96%) compared to those who test after CV review (87%).

The practical implication for VA hiring is straightforward:

A platform that only offers one or two of these data points is giving you a partial picture. You are then making a $500 to $1,500+ per month hiring decision on incomplete evidence.

According to aSHRM 2024 study cited in multiple skills-based hiring reports, 62% of HR professionals say skill validation remains a genuine challenge, because candidates routinely claim competencies they cannot actually demonstrate. Multi-measure testing is the most reliable way to close that gap before someone is sitting in your team. When youhire virtual assistants, reviewing verified data points ensures you understand the true caliber of the applicant pool.

How DonutJobs Approaches Candidate Verification

DonutJobs was built around a specific belief: that a score without evidence is not a vetting standard, it is a convenience feature.

Every candidate on DonutJobs goes through a structured, multi-stage verification process before their profile is visible to employers. That process covers four core areas:

• Hardware and connectivity check - confirming the candidate has a reliable setup for remote work
• Written English assessment - testing grammar, vocabulary, and written communication quality
• Typing speed and accuracy test - establishing a baseline for admin and support productivity
• Voice sample recording - giving employers an unfiltered listen before any interview is scheduled

The result is thatover 92% of active DonutJobs candidates have completed all four verification checks, across a pool of more than 50,000 active candidates. You are not browsing self-reported profiles and hoping the claims hold up. You are reviewing verified data before you spend a single minute of your time on a call.

Employers can alsosee exactly how DonutJobs verifies candidates and what the process looks like end to end, which is something most platforms are reluctant to publish because their process does not hold up to scrutiny.

The practical difference: DonutJobs delivers shortlisted, verified candidates in as little as 48 hours, with the underlying test evidence available, not just a badge on a profile. For businesses comparingVA hiring platforms by candidate quality and verified data, this level of transparency is the standard to hold every other platform to.

If the job is one of DonutJobs' managed services, the refund terms are also clearly defined. The policy states that managed virtual assistant services and the Executive Assistant retainer include a 14-day, first-month refund window if the service is not a fit, plus a free replacement if the assistant is the wrong person for the role. For subscription plans, the first purchase has a 48-hour refund window, and used platform access can void that eligibility. That means refund language should stay specific to the product, not treated as one blanket promise.

What to Do Before You Commit to a Platform

Evaluating platforms does not require a lengthy procurement process. A few targeted steps will tell you most of what you need to know.

1. Request a sample candidate profile. A trustworthy platform will show you exactly what data you receive before you subscribe. If they hesitate or send a generic brochure, that is your answer.
2. Ask for the score distribution on their English or typing test. How many candidates take it? What is the average score? What percentage pass? These numbers reveal whether the bar is meaningful or just decorative.
3. Check whether you can access the raw test evidence. Can you listen to a voice sample before booking an interview? Can you read a written English response? Platforms that hide the evidence are protecting their pass rates, not your hiring decision.
4. Understand the replacement policy. Even excellent assessments produce occasional mismatches. A platform confident in their vetting will offer a free replacement guarantee. For managed assistant services, DonutJobs provides a free replacement at any point while the service is running if the assistant is the wrong fit for the role.

For a broader comparison of what different platforms offer, theDonutJobs guide to comparing VA platforms breaks down the key differences across screening depth, pricing, and hiring speed.

FAQ

What is the most reliable type of skills assessment for VA hiring?

Work sample tests, where candidates complete tasks that mirror the actual job, are consistently the most predictive assessment type according to meta-analyses reviewed byCogn-IQ. For VA roles, this means asking candidates to draft an email, respond to a support query, or complete a data entry task, not just answer multiple-choice questions about what they would theoretically do.

Can candidates cheat on VA platform skills tests?

Yes, and it is more common than most platforms acknowledge.TestGorilla's 2025 research found that 7 in 10 candidates who admitted cheating used AI to do it. Platforms that use unproctored, text-only tests with no identity verification or question rotation are particularly vulnerable. When evaluating a platform, ask specifically how they prevent AI-assisted or coached responses.

How many assessments should a VA candidate complete before I trust their profile?

Research consistently shows that multi-measure testing produces better hiring outcomes than any single assessment.91% of employers using multi-measure testing report making quality hires, compared to lower rates among those using a single test. For VA hiring specifically, look for platforms that verify at least three to four distinct dimensions: written English, spoken English, role-relevant skills, and technical setup.

What does DonutJobs' verification process actually include?

DonutJobs verifies candidates across four areas: hardware and internet connectivity, written English proficiency, typing speed and accuracy, and a voice sample recording. Over 92% of active candidates on the platform have completed all four checks, across a pool of more than 50,000 verified profiles. Employers receive access to this evidence before committing to an interview or a hire.

How do I know if a platform's "verified" badge is meaningful?

Ask the platform three questions: What percentage of candidates on the platform have completed the full verification? What is the average score on their English test? And can you see the raw test output, not just the score? If they cannot answer all three clearly, the badge is a design element, not a hiring signal.

What should a VA platform's replacement guarantee tell me about their assessments?

A lot. Platforms that offer short or no replacement windows are implicitly telling you they are not confident in their own vetting. DonutJobs offers a 90-day free replacement, which reflects the confidence built into a structured, multi-stage verification process. A generous replacement policy is one of the clearest signals that a platform's assessments are genuinely predictive.

A skills score is only as valuable as the process behind it. The platforms that make it hardest to look behind the curtain are usually the ones with the least to show you.

Before you hire your next VA based on a badge or a percentage, run through the five questions in this guide. Demand to see the raw evidence. Understand what was tested, how it was administered, and what the benchmark actually means.

DonutJobs was built for employers who want answers to those questions before they commit. With over 50,000 verified candidates, four-stage verification, and a 90-day replacement guarantee, the evidence is there from the start.

Get started with DonutJobs today and access pre-tested candidates with verified scores you can actually trust, in as little as 48 hours.