Top-box score

In the survey "Rate the service from 1 to 5" - 60 people chose 5, 25 chose 4, and 15 chose 3 or lower. Top-box score is the share of respondents who chose the highest (most favorable) scale option

In the example: 60 out of 100 = 60%. The metric highlights the "top performers" rather than those "generally satisfied". Research shows that customers giving the top rating are more likely to be loyal and to recommend than those who answered "okay". In SurveyNinja, when you calculate the share of respondents who picked a particular option - that is the top box. Link to CSAT and NPS.

Top box is not the average. It is "how many gave the maximum".

Definition

Top-box score is the share of respondents who chose the highest (most favorable) answer option on a scale, out of the total number who answered. Formula: (number who chose the top option / total number of responses) × 100%. It is used in satisfaction surveys, CSAT, and service evaluation. Top box puts the emphasis on "excellent" rather than "good or better". Link to the Likert scale and rating scales - the top box is always tied to a specific question and its answer options.

In short: "what percentage gave the maximum" - 5 out of 5, 10 out of 10, "very satisfied".

Formula and examples

Top box = (number who chose the highest option / total number of responses) × 100%.

Scale 1-5. "Rate the quality of service" - options from "very poor" to "excellent". The top box is the share who chose "excellent" (5).

NPS 0-10. The top box for NPS is the share of promoters (9-10). It is often this figure that gets reported: "40% gave a 9 or 10".

Scale 1-7. The top box is only 7. Alternatively, the top-2-box is 6 and 7 together.

Emojis (CSAT). "Very satisfied" is the top box. Visual scales like emojis and stars in CSAT make these answers easy for respondents to read at a glance.

Top box vs the average and top-2-box

The average blends all ratings together. A distribution of 50% fives and 50% threes gives an average of 4. The top box is 50%. The interpretation differs: the average says "generally fine", while the top box says "half are delighted, half are not".

Top-2-box is the share who chose the two highest options. On a 1-5 scale: 4 and 5 together. On NPS: 9 and 10 (promoters). The top-2-box is softer than the top box: it includes both "good" and "excellent". For long scales (7, 10, or 11 points) the top-2-box is often more informative - a single top point can be rare.

Why use the top box

It predicts behavior better. Customers giving the maximum rating return and recommend more often than those who are merely "satisfied" (4 out of 5). The top box reveals the true enthusiasts.

A focus on excellence. The goal is not "generally happy" but "delighted". The top box sets a high bar.

Comparing locations. Two stores both average 4.2. But one has a top box of 70% and the other 40%. Where is the experience really better?

Prioritization. A low top box with an acceptable average is a signal: "good, but not excellent". There is room to grow.

Common mistakes

Confusing it with NPS. NPS is the difference between promoters and detractors. The top box is simply the share of the top option. For an NPS question the top box equals the share of 9-10, but NPS itself uses a different formula.

Different scales mean different top boxes. A top box on a 1-5 scale and on a 1-10 scale are not directly comparable. A 60% top box on a 5-point scale and a 60% top box on a 10-point scale mean different things.

A small sample. With 20 responses the top box jumps from one measurement to the next. You need a sufficient volume for a stable figure.

Central tendency. If respondents cluster toward the middle (central tendency bias), the top box will be low. Keep this in mind when interpreting the results.

In SurveyNinja: how to calculate the top box

Create a survey with a scale question (a rating of 1-5, 1-10, emojis, or stars). SurveyNinja reports show the distribution across options. The top box is the share who chose the top option. Use filtering and cross-tabulation: the top box by department, point of sale, or period. Hidden variables pass along the segment - analyze the top box by segment. Export to Excel - for your own calculations and dashboards.

Link to CSAT and CSI

CSAT (Customer Satisfaction) is often measured on a 1-5 or 1-7 scale. The top box is the share who answered "very satisfied" / "excellent". A classic CSAT can be calculated as an average, but the top box adds another angle. The CSI (Customer Satisfaction Index) is a composite indicator that may include the share of highest ratings. More on this in CSAT and satisfaction.

Benchmarks

Benchmarks depend on the industry and the question. For post-purchase CSAT: a top box of 50-70% is a good level, and above 70% is excellent. For support evaluation it is a little lower - 40-60% is often realistic. For NPS top box (9-10): 40-50% is a healthy level of promoters. Compare against your own trend and industry data, not against abstract "norms".

Case: top box by point of sale

A chain of 12 stores. A monthly survey: "Rate your visit 1-5". The chain average is 4.3. Management is pleased. They broke it down by store: in three of them the top box (a rating of 5) was below 45%, while in the rest it was 55-75%. The averages in the "weak" stores were 4.1-4.2 - not critical. But the top box showed there were fewer delighted customers there. A detailed review covered queues, assortment, and courtesy. After the changes, the top box in the problem locations rose by 10-15%. The average barely moved - but the share of "excellent" grew.

Top box and bottom box

The opposite of the top box is the bottom box: the share of the worst ratings (1 out of 5, or 0-6 on NPS). The top box is "how many are delighted", the bottom box is "how many are disappointed". Together the two metrics give a full picture of the distribution. A high top box with a low bottom box is the goal. A low top box with a high bottom box is a serious problem. Watch both.

When to use the top box

It fits when an "outstanding" rating matters, not just an "acceptable" one. Evaluating premium service, loyalty, and recommendations - the top box is relevant. For routine processes (delivery on time, yes/no) it can be excessive. Choose the metric to suit the goal: the average for the overall picture, the top box to focus on top performers.

Track the top box over time: a rise is a sign of improving experience, a drop is a signal to investigate the causes. Combine it with qualitative analysis: why did the ratings rise or fall.

The trend matters more than a one-off measurement: a top box of 55% this month against 48% last month is progress.

Top-box score is the share of respondents who chose the highest scale option. It highlights the "top performers" and predicts loyalty better than the average. In SurveyNinja - through the distribution across options and filters by segment. Link to CSAT, NPS, and bottom box.