I used to think a 4 out of 5 on a customer satisfaction survey was good news. Then I started paying attention to top-box scores, and I realized I'd been lying to myself for years.
Here's the problem: when you average Likert scale responses, you smush together the people who love you and the people who merely tolerate you into one flattering number. A top-box score cuts through that fiction. It isolates the people who gave you the highest possible rating and asks the uncomfortable question: what percentage of your customers are genuinely enthusiastic, not just satisfied?
That distinction is the difference between customers who recommend you and customers who switch the moment something slightly better comes along.
What Is a Top-Box Score?
A top-box score is the percentage of survey respondents who selected the single most positive response option on a rating scale. On a 5-point Likert scale (Strongly Disagree to Strongly Agree), the top box is "Strongly Agree." On a 10-point satisfaction scale, it's a 10. On a purchase intent scale, it's "Definitely Would Buy."
The related metric, the Top-2-Box score (T2B), combines the two most positive responses ("Strongly Agree" + "Agree," or 9-10 on a 10-point scale). According to SurveyMonkey, the Top-2-Box score is the most widely used summary metric in market research because it provides a cleaner signal than the raw mean.
Both metrics serve the same purpose: they compress a distribution of responses into a single percentage that tells you how many people feel strongly positive about whatever you're measuring.
Why Top-Box Scores Matter More Than Averages
Let me show you why this matters with an example. Suppose you survey 100 customers on a 5-point satisfaction scale and get these results:
Rating | Response A (Count) | Response B (Count) |
5 (Strongly Agree) | 30 | 10 |
4 (Agree) | 40 | 60 |
3 (Neutral) | 20 | 20 |
2 (Disagree) | 5 | 5 |
1 (Strongly Disagree) | 5 | 5 |
Mean | 3.85 | 3.70 |
Top-Box % | 30% | 10% |
Top-2-Box % | 70% | 70% |
Look at what happens. The means are similar (3.85 vs. 3.70). The Top-2-Box scores are identical (70%). But the top-box scores are dramatically different: 30% vs. 10%. Scenario A has three times as many truly enthusiastic customers as Scenario B. If you only looked at averages or T2B, you'd think these were nearly identical results. They're not.
This is why MeasuringU argues that top-box scores are often better predictors of actual behavior than mean scores. The people in the top box are the ones who repurchase, who refer friends, who write positive reviews, who forgive your occasional mistakes. The people in the second box? They're satisfied, sure, but they're also open to alternatives.
Where Top-Box Scores Are Used
Top-box analysis shows up across virtually every area of market research and marketing analytics:
Customer Satisfaction (CSAT)
The most common application. Companies ask "How satisfied are you with [product/service]?" on a 5- or 7-point scale and report the top-box or T2B as their headline CSAT number. Qualtrics includes top-box analysis as a standard metric in its experience management platform.
Purchase Intent
Concept testing surveys frequently ask "If this product were available today, how likely would you be to buy it?" on a 5-point scale (Definitely Would Buy to Definitely Would Not Buy). The top-box score ("Definitely Would Buy") is the standard metric reported to stakeholders, because research consistently shows that only "Definitely" responders convert at meaningful rates.
Net Promoter Score Complement
NPS uses a 0-10 scale and classifies 9-10 as "Promoters." That's essentially a top-2-box approach on a wider scale. But some organizations track the single top-box (10 only) to identify their most extreme advocates.
Advertising Testing
When testing ad concepts, researchers measure agreement with statements like "This ad makes me want to learn more about the product." The top-box score separates ads that genuinely move people from ads that are merely inoffensive. This connects directly to advertising awareness measurement and conversion rate prediction.
Brand Health Tracking
Brand trackers measure agreement with attributes like "This is a brand I trust" or "This brand is innovative." Top-box scores on these attributes over time reveal whether your brand equity is strengthening or weakening among your most committed customers.
Top-Box vs. Top-2-Box: When to Use Which
Scenario | Recommended Metric | Why |
Identifying true advocates | Top-Box (single) | Isolates genuine enthusiasm |
Reporting overall satisfaction | Top-2-Box | More stable, larger sample in the box |
Purchase intent testing | Top-Box ("Definitely") | Only "Definitely" converts reliably |
Comparing across segments | Top-2-Box | Larger percentages reduce noise |
Tracking over time | Both | Top-Box shows intensity shifts; T2B shows breadth |
Small sample sizes (n < 100) | Top-2-Box | Single top-box may be too volatile |
I generally recommend tracking both. The top-2-box gives you breadth (how many people feel positive), while the single top-box gives you intensity (how many people feel strongly positive). The gap between the two is itself an interesting diagnostic: a wide gap means you have lots of people who like you but few who love you.
Best Practices for Top-Box Analysis
Choose the right scale length. For top-box analysis, 5-point and 7-point scales are most common. SurveyMonkey's research shows that 5-point scales are easier for respondents and yield comparable insights to 7-point scales for most applications. Longer scales (10-point) can provide more granularity but risk confusing respondents about the difference between, say, a 7 and an 8.
Label every scale point. Research from Alchemer indicates that fully labeled scales (where every point has a text label, not just the endpoints) produce more reliable data. Don't just label 1 as "Low" and 5 as "High." Label every point.
Be consistent across surveys. If you're tracking top-box scores over time, changing your scale length or wording invalidates comparisons. Lock in your methodology and stick with it.
Report alongside the full distribution. Top-box scores are a summary, not a replacement for the full picture. Always make the underlying distribution available. A top-box score of 25% means something very different depending on whether the remaining 75% are neutral or actively dissatisfied.
Consider weighted top-box for tighter comparisons. MeasuringU suggests that when comparing options with similar T2B scores, a weighted top-box (e.g., 2x for "Strongly Agree" + 1x for "Agree") can provide better discrimination.
The Relationship Between Top-Box Scores and Business Outcomes
High top-box scores correlate with several important business outcomes:
Customer retention. Customers who rate you at the top of the scale are significantly less likely to churn. This directly connects to your retention rate and churn rate metrics.
Word-of-mouth referral. Top-box responders are your most likely source of organic referrals, which lowers customer acquisition cost and builds brand power over time.
Revenue growth. Qualtrics research has shown that companies with higher top-box satisfaction scores grow revenue 2-3x faster than competitors with merely average satisfaction. The causal mechanism is straightforward: enthusiastic customers buy more, buy more often, and bring their friends.
Pricing power. Customers who rate you at the top of the scale are less price-sensitive. They've already decided you're worth it. This is the behavioral foundation of brand equity's financial value.
Limitations and Criticisms
Top-box scores aren't perfect, and I think it's important to be honest about their limitations:
Cultural response bias. Some cultures tend to use the extremes of scales more than others. Deloitte's global research has noted that top-box scores can vary by 10-15 percentage points across markets for identical products due to cultural differences in scale use.
Scale anchoring effects. The number of scale points changes the top-box percentage. A top-box on a 5-point scale (20% of the scale) isn't directly comparable to a top-box on a 10-point scale (10% of the scale).
Sample size sensitivity. With small samples, top-box percentages can swing wildly. You need adequate sample sizes (generally n > 100 per segment) for stable top-box metrics.
Doesn't capture the "why." A top-box score tells you how many people love you, not why. Always pair top-box questions with open-ended follow-ups. As QuestionPro recommends, ask "What's the main reason for your rating?" immediately after the scaled question.
Thought Leaders and Key Resources
Jeff Sauro at MeasuringU has published extensive research on top-box vs. mean comparisons and their predictive validity. Qualtrics and SurveyMonkey both provide excellent methodology guides for implementing top-box analysis. ESOMAR (the global association for market research) publishes standards for survey scale design that inform top-box best practices.
FAQs
What is a top-box score?
A top-box score is the percentage of survey respondents who selected the single most positive response on a rating scale. It isolates the most enthusiastic responders.
What is a top-2-box score?
A top-2-box (T2B) score combines the two most positive responses (e.g., "Strongly Agree" + "Agree") into a single percentage. It's the most widely used summary metric in market research.
Why are top-box scores better than averages?
Averages blend enthusiastic and merely tolerant responders into one number, masking important differences in customer intensity. Top-box scores isolate genuine enthusiasm, which better predicts behavior like repurchase and referral.
What scale should I use for top-box analysis?
Five-point and seven-point scales are most common. Five-point scales are simpler for respondents; seven-point scales offer slightly more granularity. Both work well for top-box analysis.
How do top-box scores relate to NPS?
NPS is essentially a top-2-box approach on a 0-10 scale, classifying 9-10 as "Promoters." NPS goes further by subtracting "Detractors" (0-6), which top-box scores don't do.
What's a good top-box score?
It depends heavily on the category and question type. For customer satisfaction, 30-40% top-box on a 5-point scale is generally strong. For purchase intent, 15-25% "Definitely Would Buy" is typically considered positive.
Can I compare top-box scores across different scale lengths?
Not directly. A top-box on a 5-point scale represents 20% of the scale; on a 10-point scale, it's 10%. Use the same scale consistently for valid comparisons.
How large does my sample need to be for reliable top-box scores?
Aim for at least 100 respondents per segment being compared. Smaller samples produce volatile top-box percentages.
Sources & References
- SurveyMonkey, "How to Use a Top 2 Box Score in Survey Analysis"
- MeasuringU, "Are Top Box Scores a Better Predictor of Behavior?"
- MeasuringU, "Top Box, Top-Two Box, Bottom Box, or Net Box?"
- SightX, "Top-Box Analysis"
- QuestionPro, "Top-Box Score: Deriving a New Measure"
- Qualtrics, "Top Box Metrics"
- Alchemer, "Top Box or Top 2 Box Reporting"
- Helio, "How to Use a Top 2 Box Score"
Written by Conan Pesci | April 5, 2026 | Markeview.com
Markeview is a subsidiary of Green Flag Digital LLC.