The VantaFive People Evaluation System connects agency hiring to ongoing performance reviews. Define the role, assess evidence against a scorecard, verify it through interviews and references, then use the same standard to develop the person.
Replace impressions with a defined standard
Most people decisions begin with impressions: a confident interview, a likable personality, or a sense that someone could be excellent. Those impressions are hard to compare and even harder to revisit when performance does not match expectations.
VantaFive’s system begins before the interview. It defines what success in a role means and gathers evidence against that definition. The framework draws on EOS, developed by Gino Wickman, and Who: The A Method for Hiring by Geoff Smart and Randy Street, adapted through experience in agency businesses.
It is an evaluation system. It starts when a candidate is worth assessing and continues after they join. It does not replace the separate work of finding candidates.
Define the role before evaluating the person
Start with a one- or two-sentence mission: why does the role exist, and what does success look like? Define the important outcomes before discussing candidate strengths.
Build the scorecard in three groups: values, hard skills, and soft skills. Keep each group focused on the capabilities that matter for this role. Give every criterion a behavioral description that two reviewers could apply to the same evidence.
- Write the role mission and expected outcomes.
- Choose five to ten meaningful criteria per group; remove filler.
- Describe what consistent performance looks like in observable terms.
- Agree on the evaluation standard before interviewing.
Use consistent scoring
VantaFive scores demonstrated values from 1 to 3 and skills from 1 to 5. The values standard is consistent demonstration. Skills can have a development plan when a gap is compatible with the role’s requirements.
The numerical standard helps reviewers communicate, but a number without evidence is not an assessment. Record the example, where it came from, and what remains uncertain.
| Scale | Meaning |
|---|---|
| Values: 1 | The behavior has not been demonstrated. |
| Values: 2 | The behavior is demonstrated inconsistently. |
| Values: 3 | The behavior is demonstrated consistently. |
| Skills: 1–5 | Assess capability against the defined role standard. |
| V5 skill benchmarks | B: 4.0+ average; A: 4.5+; A+: 4.7+. These are internal evaluation conventions. |
Screen, then investigate
Schedule a short screening call to establish whether a full interview is worthwhile. The process starts with a 15-minute screening call while leaving room for the conversation to run longer when useful. Ask each candidate a consistent set of relevant questions.
In the full interview, work through the person’s relevant career history. Ask what they were hired to do, what went well, what was difficult, who they worked with, and why they left. Understand how success was measured.
For important accomplishments, check the context against the previous year, the plan, and peers. Capture evidence that supports or challenges each scorecard criterion.
Cross-check the picture
Use references to verify specifics from the interview. Seek the perspective of people who directly observed relevant work. Do not treat a candidate’s self-assessment or a generic compliment as the complete picture.
Where judgment is central to the role, consider a small work sample. Give the candidate a bounded, realistic problem and ask them to explain their decisions. The explanation should show assumptions, tradeoffs, and the reasoning behind the result.
The public framework recommends an assignment that can reveal useful thinking in roughly one or two hours. Long exercises often measure available free time more than capability.
Use AI as an evidence assistant
AI can help organize an interview transcript by scorecard criterion and locate supporting passages. Verify those passages against the transcript. Do not let a fluent summary conceal missing evidence.
The hiring team makes the decision. AI-generated ratings are suggestions to examine, and unsupported ratings should be disregarded. Handling recordings and candidate information requires the organization’s normal consent and data practices.
Keep the scorecard alive
Share expectations with the person who joins the team. Use the same scorecard in quarterly reviews, with specific examples of progress and gaps. Turn skill gaps into a concrete development conversation. That continuity matters as AI changes the work through which junior employees develop judgment.
The outcome is continuity: the standard used to hire someone is also the standard used to help them improve.
- Before the interview: a clear mission and agreed scorecard.
- After the interview: evidence, scores, and unresolved questions.
- Before a decision: references and any relevant work sample.
- After hiring: visible expectations and recurring feedback.
A worked criterion
Illustrative example for an account leadership role: “Client communication” means identifying a material issue, explaining its impact clearly, proposing the next action, and keeping the client informed until it is resolved.
A useful evidence note records a specific situation, the candidate’s actions, the result, and what a reference can verify. “Great communicator” alone does not show any of those things. The Green Flag Process shows the kinds of client concerns an account leader needs to handle. This example is illustrative; it is not a real candidate evaluation.
Three things to remember.
- Define observable criteria before interviewing.
- Use the interview and references to gather evidence against those criteria.
- Reuse the scorecard for development after hiring.
About this article
Adapted and expanded from “The VantaFive People Evaluation System” on LinkedIn ↗, originally shared by John Ghiorso on . The Journal edition was first published on September 8, 2026.
The practical steps also draw on the Hiring A Players modules in the VantaFive Operating System.
Framework influences: Who: The A Method for Hiring by Geoff Smart and Randy Street, and EOS, developed by Gino Wickman.
More about John Ghiorso →