A Likert scale is a survey question that measures an attitude or opinion by asking respondents how much they agree or disagree with a statement, using an ordered set of answer points (most often 5 points from "strongly disagree" to "strongly agree"). It turns a subjective feeling into a number you can average, compare, and track over time.

What a Likert scale is

A Likert scale (named after psychologist Rensis Likert) presents a single statement and a symmetric, ordered list of response options. Each option carries a numeric value, so "strongly disagree" might equal 1 and "strongly agree" might equal 5. Because the points are ordered and evenly spaced in meaning, you can calculate an average score for a question or combine several related statements into one overall index.

The defining features of a true Likert scale are:

  • A statement, not a question. Respondents react to "The checkout process was easy," not "How easy was checkout?"
  • A balanced range. The same number of positive and negative options sit on either side of a neutral midpoint.
  • Ordered, labeled points. Every point has a word label, not just a number, so the meaning is unambiguous.

Examples

A classic 5-point agreement scale:

  • The statement: "I would recommend this product to a friend."
  • The points: Strongly disagree, Disagree, Neutral, Agree, Strongly agree.

Likert-style scales also work for dimensions other than agreement, as long as the labels stay balanced and ordered:

  • Frequency: Never, Rarely, Sometimes, Often, Always.
  • Satisfaction: Very dissatisfied, Dissatisfied, Neutral, Satisfied, Very satisfied.
  • Quality: Very poor, Poor, Fair, Good, Excellent.
  • Likelihood: Very unlikely, Unlikely, Neutral, Likely, Very likely.
  • Importance: Not important, Slightly important, Moderately important, Important, Very important.

5-point vs 7-point scales

Both are valid; the choice is a trade-off between simplicity and sensitivity.

  • 5-point scales are fast to answer, easy to label clearly, and work well on mobile. They are the safe default for most studies.
  • 7-point scales add two more points (for example, "somewhat agree" and "somewhat disagree") and capture finer differences in strength of opinion. They suit research where you need to detect small shifts, such as tracking studies or academic work.

Keep these rules whichever length you pick: use an odd number of points if you want to allow a genuine neutral answer, keep the labels balanced so positive and negative sides mirror each other, and label every point rather than only the ends.

When to use a Likert scale

Use a Likert scale when you want to measure the degree of an attitude, not just its presence:

  • Gauging agreement with statements about a brand, product, or experience.
  • Measuring satisfaction, ease, trust, or perceived quality.
  • Building a multi-statement index (for example, several agreement items that together measure "perceived value").
  • Running tracking studies where you compare average scores across waves or segments.

When not to use one

Reach for a different question type when:

  • You need an exact count or number. Use a number or rating question instead.
  • You are measuring likelihood to recommend at the company level. Use a dedicated NPS question, which has its own 0 to 10 scale.
  • You want reasons or rich detail. Pair the scale with an open-text follow-up, or use a open-ended question.
  • The concept is not really a spectrum (for example, "Which feature did you use?"). Use a single or multiple choice question.

How to add this in SegmentOS

SegmentOS includes a dedicated Likert Scale question type (available on Premium and Pro plans).

  1. In the survey builder, add a question and set its type to Likert Scale.
  2. In the Likert editor, pick a Quick Template to fill the points instantly. Built-in templates include Agreement, Frequency, Satisfaction, Quality, Likelihood, and Importance.
  3. Or set your own number of scale points and type a label for each one, so every point on the scale is clearly named.
  4. Optionally assign a score to each point so the responses roll up into an average you can analyze and segment.

For a full list of available formats, see Question types. To make sure your statements are neutral and not nudging respondents, read How to avoid leading and biased survey questions.

Frequently asked questions

How many points should a Likert scale have?

Five points is the most common and works for nearly all studies. Use seven points when you need to detect finer differences in opinion, such as in tracking or academic research, and keep the number odd if you want to allow a true neutral answer.

Should a Likert scale include a neutral middle option?

Include a neutral midpoint when "no strong opinion" is a genuine, meaningful answer, which is usually the case. Omit it (using an even number of points) only when you deliberately want to force respondents to lean one way.

What is the difference between a Likert scale and a rating scale?

A Likert scale asks for agreement with a statement using fully labeled, balanced points, while a general rating scale asks respondents to rate something on a numeric range that often labels only the endpoints. In SegmentOS the Likert Scale type labels every point, whereas the rating scale type sets a min and max value with optional end labels.

Can I turn Likert answers into a number?

Yes. Each point on a Likert scale maps to a numeric value, so you can average responses, compare segments, and track scores over time. In SegmentOS you can assign a score to each scale point for exactly this purpose.