Behaviorally Anchored Rating Scale (BARS) Guide and Examples
A Behaviorally Anchored Rating Scale (BARS) is a performance rating scale where each score is tied to a concrete description of on-the-job behavior, so a "4" means the same observable thing to every manager. You build one from real examples of good and poor performance in a specific role, and it gives clearer, more defensible ratings than a bare 1 to 5 scale.
This guide is for HR leaders, people managers and HR analysts who want performance reviews employees actually trust.
Key takeaways
- BARS replaces vague labels like "meets expectations" with specific, observable behaviors at each rating level.
- Anchors come from real critical incidents collected from managers and strong performers, not from invented ideals.
- The biggest practical benefit is clearer feedback and better rater alignment; research on accuracy gains is mixed.
- BARS is role-specific and takes real effort, so start with the few competencies that matter most.
- Rater training and periodic anchor updates are still required.
Why do traditional rating scales fall short?
Most employees are skeptical of their reviews, and often with reason. A 1 to 5 score on "communication" or "professionalism" leaves a lot of room for interpretation. One manager's 3 is another manager's 4, so ratings end up describing the manager as much as the employee. When pay and promotions depend on those scores, the inconsistency becomes a fairness problem. Our piece on common performance appraisal problems covers the usual failure modes, from leniency to recency bias.
BARS, first described by Patricia Smith and Lorne Kendall in the early 1960s, tackles the interpretation problem directly by defining what each rating looks like in practice.
What does a BARS look like?
Instead of rating "customer communication" from 1 to 5 with no context, a BARS describes the behavior that corresponds to each level for that role. Here's an example for a customer service representative:
| Rating | Behavioral anchor: customer communication |
|---|---|
| 5 | Gives complete resolution information without being asked, anticipates follow-up questions, and confirms the customer is satisfied before closing. |
| 4 | Resolves the stated issue fully and usually explains next steps, with occasional prompting. |
| 3 | Addresses the stated issue accurately but rarely adds context or confirms resolution. |
| 2 | Gives partially correct answers; customers often need a second contact to finish. |
| 1 | Gives incomplete or inaccurate responses that require repeated follow-up from the customer. |
A second example, for a first-line supervisor rated on "coaching":
- 5: Holds regular one-to-ones, gives specific feedback within a day of observing an issue, and agrees a written development action with the employee.
- 3: Gives feedback when asked or during formal reviews; it's accurate but often general.
- 1: Rarely gives feedback; problems surface first in the annual review.
And for a sales representative rated on "pipeline management": a 5 keeps every opportunity updated with a dated next step and flags stalled deals to the manager; a 1 leaves opportunities without next steps and forecasts from memory. Behavioral definitions like these are also how teams make high-performance sales team characteristics concrete.
Each anchor should describe something a manager could actually observe. "Has a great attitude" doesn't work. "Volunteers to cover open shifts during peak weeks" does.
How do you build a BARS step by step?
Building a valid BARS is structured but not technically hard. It needs the right people in the room and a few facilitated sessions. Here's the usual process:
- Pick the role and competencies. Choose three to six dimensions that really separate strong from weak performance in that job.
- Collect critical incidents. Ask managers and high performers for specific examples of effective and ineffective behavior they've seen. Aim for many short, concrete stories.
- Sort incidents into dimensions. A separate group assigns each incident to a competency. Drop incidents that people can't agree on. This is the "retranslation" step.
- Scale the incidents. Experts rate how effective each incident is. Keep the ones with strong agreement and use them as anchors at each level.
- Draft and test the scale. Have several managers rate the same sample employees or recorded scenarios, then compare scores and reword anchors that caused disagreement.
- Check against outcomes. Over time, see whether ratings line up with results such as quality scores, customer satisfaction or sales.
Whatever format you use, the standards should meet the test the U.S. Office of Personnel Management sets for performance elements and standards: measurable, understandable, verifiable, equitable and achievable. BARS anchors are one practical way to get there.
Is BARS more accurate than other rating methods?
BARS is widely used because its anchors give employees clear feedback (they can see exactly which behaviors to change) and give managers a shared reference point. Studies comparing it to other formats have found mixed results on rating accuracy and bias, though, so don't expect the scale alone to fix inconsistent ratings. The gains come from combining it with rater training and calibration.
That still matters a lot when ratings drive pay. If one manager rates generously and another harshly, pay decisions inherit that bias and erode trust. Clearer anchors and calibration sessions reduce the gap. Analytics can help spot it too; see AI in compensation and benefits for how teams check rating and pay patterns for fairness.
What are the limitations of BARS?
- Time to build: creating anchors for many roles and competencies is a significant investment.
- Role-specific: scales don't transfer across job families without rework.
- Anchors age: what excellent looked like two years ago may not fit today's tools or priorities.
- Bias isn't removed: managers still need training to separate observation from interpretation.
- Anchor matching: raters sometimes struggle when an employee shows behaviors from two different levels.
If full development is too heavy, use a hybrid. Apply behavioral anchors only to the two or three most critical dimensions, or adapt an established competency framework rather than starting from scratch. You'll capture much of the benefit with far less effort.
How do you put BARS into practice?
Most performance platforms, including Workday, SAP SuccessFactors and Lattice, let you configure custom rating scales with descriptions for each level, so you rarely need a separate tool. Our guide to using your HRMS for performance appraisals covers setup basics. Before rollout, answer four questions: how many roles need their own scale, who owns maintenance, how managers will be trained, and when anchors will be reviewed.
Pilot in one department or job family first. A common pattern is that the pilot surfaces two or three anchors managers read differently, and fixing those before company-wide rollout saves a lot of friction. Teams also often find the workshops themselves valuable: getting managers and employees to agree on what excellent work looks like is useful whether or not anyone looks at the final scale. Once ratings are flowing, connect them to development plans and track them over time, as described in our HR analytics learning roadmap.
FAQ
What is the difference between BARS and a graphic rating scale?
A graphic rating scale asks raters to score a trait, such as communication, on a numeric or labeled scale with little or no description. BARS attaches a specific behavioral example to each point on the scale for a particular role. That makes BARS slower to build, but it gives raters and employees a clearer shared picture of what each rating means.
How many anchors should each dimension have?
Most BARS use five or seven levels, with a behavioral anchor at each level or at least at the low, middle and high points. More levels add precision but make it harder for raters to choose. Five levels with clear anchors at every point is a practical default for most organizations, especially when managers are new to behavioral rating.
Can small companies use BARS?
Yes. A small company doesn't need a formal research process. Pick one or two key roles, gather examples from the people who know the work best, and draft anchors for three or four competencies. Test them in one review cycle and refine. Even a lightweight version gives clearer feedback than an unanchored 1 to 5 scale.
What to do next
- Choose one high-headcount role where ratings cause the most disagreement.
- Run a one-hour critical incident session with three managers and two strong performers.
- Draft anchors for three competencies and test them by having managers rate the same sample cases.
- Configure the scale in your performance system and run a calibration meeting after the first cycle.
Sources
- U.S. Office of Personnel Management: Rating
- U.S. Office of Personnel Management: Performance management
- Smith, P. C., and Kendall, L. M. (1963). Retranslation of expectations: An approach to the construction of unambiguous anchors for rating scales. Journal of Applied Psychology, 47(2), 149-155.
Last reviewed: October 2026.

Comments
Post a Comment