> ## Documentation Index
> Fetch the complete documentation index at: https://help.equip.co/llms.txt
> Use this file to discover all available pages before exploring further.

# What Is a Good Assessment Score?

> Benchmarks for interpreting candidate scores, and how to set sensible cutoffs.

<Tip>
  There is no universal "good" score. As a rule of thumb, read a candidate's percentage as a **percentile**: a 70% score suggests the candidate performed better than roughly 70% of practitioners with those skills at that experience level. Equip works best when you sort candidates by score and shortlist the top performers, rather than judging any single score in isolation.
</Tip>

## Score Benchmarks

When Equip's content team designs questions, they follow a **percentage as percentile** rule of thumb. If you create an assessment for a Digital Marketer with 0 to 2 years of experience and a candidate scores 70%, you can interpret that as: this candidate is better than about 70% of practicing digital marketers with 0 to 2 years of experience.

The same logic applies at any point on the scale:

| If a candidate scores | Read it as                                                                          |
| --------------------- | ----------------------------------------------------------------------------------- |
| 95%                   | Better than roughly 95% of professionals with those skills at that experience level |
| 70%                   | Better than roughly 70% of such professionals                                       |
| 50%                   | Around the middle of the field for those skills and that experience level           |

<Note>
  If your assessment mixes questions at different difficulty levels, for example 4 Medium skills and 2 Hard skills, the percentile reading becomes a **weighted average** across those experience levels rather than a single clean benchmark. See [Question Difficulty Levels](/question-difficulty-levels) for how difficulty maps to experience.
</Note>

## Context Matters

Asking "what is a good assessment score?" is a bit like asking "what does an apartment cost?". The honest answer is: it depends.

When you create an assessment, Equip selects questions based only on the **skills** and **difficulty levels** you choose. It knows nothing about your company, your hiring bar, or how selective the role needs to be.

Before you anchor on a number, consider three questions:

| Question                   | Why it changes the bar                                                                                                                                        |
| -------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| How critical is the hire?  | A first hire for a role usually needs to be exceptional; the 100th hire on a large team may not                                                               |
| How selective is the role? | Hiring the best backend developer in the market demands a higher bar than filling an average-difficulty position                                              |
| How are you using Equip?   | If the assessment is your primary filter, set a stricter bar; if it is one signal among many and you will interview everyone anyway, a moderate score is fine |

The same 65% can be a strong signal in one context and a weak one in another. Decide what the score is for before deciding what counts as good.

## Using Cutoffs

Recruitment is often a **relative game**. You are rarely looking for an objectively perfect candidate; you are looking for the best people in the pool you actually have. This is where Equip is most effective.

Instead of fixing a cutoff in advance, let the pool set it:

<Steps>
  <Step title="Let the results come in">
    Wait until most candidates have completed the assessment, so the score distribution reflects your real pool.
  </Step>

  <Step title="Sort by Overall Score">
    Open the assessment results and sort candidates by their **Overall Score**, which combines each test's percentage with its weightage. See [Test Scores and Overall Score](/test-scores-and-overall-score) for the exact formula.
  </Step>

  <Step title="Shortlist the top N">
    If you need 5 candidates from a pool of 100, take the top 5. You can also apply score filters and update candidate statuses, such as **Shortlisted** or **Rejected**, in bulk.
  </Step>
</Steps>

<Warning>
  Avoid reusing a fixed cutoff, such as "always shortlist above 60%", across different assessments. Each assessment has its own skills, difficulty mix, and test weightages, so the same number does not mean the same thing everywhere.
</Warning>

Note that some tests do not produce a simple percentage at all. English Communication results use **CEFR levels**, psychometric tests are unscored personality profiles, and both psychometric and typing tests carry no weightage toward the Overall Score.

## Related Resources

* [Test Scores and Overall Score](/test-scores-and-overall-score) - How per-test scores and weightages combine into the Overall Score
* [Question Difficulty Levels](/question-difficulty-levels) - How Medium, Hard, and Expert map to years of experience
* [CEFR Scores](/cefr-scores) - Interpreting English Communication results on the A1 to C2 scale
* [Psychometric Scores](/psychometric-scores) - Why personality tests carry no points or weightage