What Is a Good CSAT Score? (2026 Benchmarks by Industry, Channel, and Scale)

A good CSAT score is between 75% and 85% in most industries, with top performers consistently above 90%. The US national baseline across all sectors is 76.9 out of 100, per the American Customer Satisfaction Index (ACSI, Q4 2025). What "good" actually means for your contact center depends on three things most CSAT guides skip: your industry's ACSI tier, the channel you measure on (voice, chat, email), and whether you report CSAT as a percentage or as an average on a 1 to 5 scale. Balto , the AI Workforce for the contact center, tracks CSAT across every interaction and shows exactly where your score sits against the right benchmark, so you can stop arguing about the number and start moving it.

What Is a Good CSAT Score? The Short Answer

For most contact centers, a CSAT score between 75% and 85% is good. Anything above 90% puts you in top-performer territory. Below 60% is a structural problem that usually points to process, policy, or routing, not individual agent performance.

If you measure CSAT as a percentage (satisfied customers divided by total responses times 100), aim for your industry's ACSI tier plus 2 to 3 points of headroom. If you measure on a 1 to 5 scale, a 4.0 to 4.2 average usually maps to the "good" band when scored correctly as the percentage who rated a 4 or 5 (the standard CSAT-5 method). For the exact formula and 6-step walkthrough, see our guide to how to calculate CSAT in a call center .

The real question is not "what is a good CSAT score in general" but "what is a good CSAT score for my industry, my channel, and my scale." That is what the rest of this guide answers.

Why There Is No Single "Good" CSAT Score

Every page that gives you a single CSAT number as "the benchmark" is oversimplifying. CSAT is useful precisely because it is comparative, and four variables change what good looks like for you.

The ACSI Baseline vs. Your Industry Average

The American Customer Satisfaction Index is the only cross-industry, methodologically consistent CSAT baseline in the US. Its Q4 2025 national aggregate is 76.9 out of 100. That is the single most useful anchor to start from.

But ACSI reports by sector, and sector scores vary widely. Banks score 80, specialty retail scores 80, supermarkets score 78, and most telecom scores in the high 60s to mid 70s. Comparing your telco contact center to the national 76.9 average tells you much less than comparing it to your sector's own ACSI benchmark.

Channel Effects (Voice, Chat, Email)

Channel changes everything. In-app and chat CSAT routinely score 10 to 15 points higher than email CSAT in vendor-reported data, and voice sits between them. This is not because voice agents are worse than chat agents. It is a function of survey design, response rate, and self-selection bias per channel.

Comparing a voice-only team to a chat-only competitor's headline number is apples to oranges. Benchmark voice against voice.

Scale Effects (1-5 vs. 1-10)

Teams running CSAT-3 (a 1 to 3 scale) see the highest response rates. CSAT-5 is the industry default. 10-point scales get the lowest completion. The same customers answering the same question on different scales will produce different raw averages, and the "satisfied" threshold shifts (top 2 responses on each scale). If your scale changed in the last year, your trend line is partly real and partly a scale artifact. Our guide to CSAT vs NPS vs CES covers how different scales interact with each metric.

Calculation Method (% Satisfied vs. Average Rating)

The number that gets called "CSAT" is calculated two ways in the wild. The correct method is the percentage of respondents who rated 4 or 5 on a 1 to 5 scale. A 4.2 average rating is NOT the same as a 4.2% CSAT, and treating them as the same is the single most common reporting mistake. We cover this fully in the "Out of 5" section below.

CSAT Score Tiers: The Framework Most Teams Use

The most widely used tier framework for CSAT percentages, popularized by SurveyMonkey and echoed across the industry, groups scores into four bands.

0 to 50% (Needs Improvement). Structural problem. Usually tied to process, policy, or routing. Agent-level coaching alone will not move this number.

50 to 70% (Fair). Below most industry benchmarks. Common pattern: inconsistent service across agents, with the bottom quartile dragging the mean.

70 to 85% (Good). The band most contact centers target and where most industries' ACSI numbers land. Operationally sustainable, but has clear headroom for a well-run team.

85 to 100% (Excellent). Top quartile. Scores above 90% are typically described as world-class and require closed-loop coaching tied to QA standards to sustain.

Use these tiers as a sanity check, not a scoreboard. A 72% CSAT at a national bank is below benchmark. A 72% CSAT at a telecom provider is often above benchmark. Industry context is everything.

What Is a Good CSAT Score by Industry (ACSI 2026 Benchmarks)

ACSI publishes sector-specific CSAT studies throughout the year. The table below pulls 2026 numbers from its most recent industry studies. Treat these as the authoritative benchmarks for your industry. Vendor-reported numbers from CX surveys disagree widely because their underlying samples differ, so stick with ACSI where it is available.

Banking and Credit Unions

Banking consistently tops the ACSI consumer industries, and 2026 is no exception. The February 2026 ACSI Finance Study puts banks overall at 80, with regional and community banks at 83 (the highest sector score of the study) and super-regional banks at 77. Credit unions sit at 78. Financial advisors score 82.

For a contact center supporting retail banking, 80 is the benchmark to beat. For community banks, 83 is the credible target. These are higher-than-average benchmarks because banking customers reward reliability and clarity, both of which agent-driven service controls.

Retail (General, Specialty, Online, Supermarkets)

The January 2026 ACSI Retail Study clustered retail between 78 and 80. Specialty retail leads at 80, followed by general merchandise and online retail at 79, with supermarkets at 78. Discount and department stores tend to score slightly lower, in the mid-to-high 70s.

Retail contact centers typically handle a mix of order status, returns, and loyalty questions, and the small score differences between sub-sectors usually come from fulfillment and returns experience rather than contact-center performance.

SaaS and Software

ACSI does not publish a dedicated SaaS CSAT score comparable to its consumer studies. Vendor-reported numbers for SaaS cluster around 78%, per third-party CSAT benchmark reports (Koji 2026, Lorikeet). B2B SaaS CSAT varies more than consumer CSAT: enterprise accounts tend to score higher (80 to 85) while self-serve and PLG accounts can drop into the 60s. Segment before you benchmark.

Healthcare and Insurance

Healthcare and insurance CSAT is not covered in a single ACSI study aggregated the way banking and retail are, so most public numbers come from vendor data. Commonly cited ranges put health insurance CSAT in the low-to-mid 70s and healthcare provider CSAT slightly higher at 76 to 80 (Fullview, Lorikeet 2026). Treat these as directional, not ACSI-grade.

Compliance volume and sensitivity of the conversations make this an industry where agent assist and 100% QA coverage pay off more than almost anywhere else.

Telecom

Telecom has historically been one of the lowest-scoring sectors across every CSAT index. The May 2026 ACSI Telecom Study continues that pattern, with most major carriers scoring in the high 60s to mid 70s. A telecom contact center hitting 75% CSAT is often above benchmark, not below. Compare against the sector, not the national average.

What Is a Good CSAT Score by Channel (Voice, Chat, Email)

Channel-specific CSAT numbers are not published consistently by ACSI, so the figures below come from vendor-reported CX research (Zendesk CX Trends 2026, GetMacha 2026). Treat them as directional benchmarks, not ACSI-grade. The clearest pattern across every public dataset: email trails every other channel, often by 10 to 15 points.

Voice (Phone and IVR)

Voice CSAT scores typically land between 82 and 83% in vendor data, with Zendesk's 2026 trends report putting phone at 82. Voice is the channel where real-time agent assist matters most, because it is the only channel where the customer and agent are both speaking live and the agent has seconds, not minutes, to resolve the issue. Scores on voice skew to the extremes (very happy or very frustrated) because response bias is highest post-call.

Chat (Live Chat and In-App)

Chat CSAT scores higher than voice in most reports. Zendesk 2026 puts live chat with a human at 88% and in-app help at 91%. Chat gets the highest completion rate too (15 to 25%) because the survey arrives in the same window where the customer just finished the conversation. Compare chat scores to other chat scores, not to voice.

Email

Email CSAT trails every other channel. Zendesk 2026 puts email at 76%, GetMacha has it at 74. Response rates are also lowest (3 to 7%), which means sample-size effects dominate for smaller queues. Report email CSAT with weekly rolling averages, not daily numbers, and expect the band to be meaningfully lower than voice and chat without reading that as a service failure.

Omnichannel Interpretation

If you report a single company-wide CSAT across channels, use a weighted average, not a simple mean. The weight is the number of valid responses per channel. This prevents a high-response, low-score channel from being masked by a low-response, high-score channel, and it keeps the number honest. For deeper multichannel interpretation, see our guide to analyzing Voice of the Customer .

What Is a Good CSAT Score Out of 5? (And Why Average ≠ CSAT)

This is the single most common CSAT reporting mistake, and it matters enough to break out separately.

The Common Mistake: Treating a 4.2/5 Average as 4.2% CSAT

When CSAT runs on a 1 to 5 scale, teams sometimes report "our CSAT is 4.2" and others read that as "our CSAT is 4.2%." Those are two completely different numbers. A 4.2 average on a 1 to 5 scale is a strong score. A 4.2% CSAT is a crisis.

The confusion happens because the word "CSAT" is used two ways in the wild: as the raw average rating and as the industry-standard percentage of satisfied responses. Pick one, document it, and use it consistently.

The Correct Conversion: 1-5 Scale to CSAT Percentage

The industry-standard CSAT calculation on a 1 to 5 scale is the percentage of respondents who rated 4 or 5, divided by total responses, times 100. If 82 of 100 respondents rated a 4 or 5, your CSAT is 82%, not 4.2 and not 4.2%.

A 4.0 average usually maps to roughly 80% CSAT when scored correctly, but the mapping depends on your score distribution. A survey with mostly 4s and 5s and a few 1s will show a different average-to-percentage ratio than a survey with a flatter distribution. Always compute CSAT as the percentage and report it that way.

Rating Band Mapping (4.5+, 4.0-4.4, 3.5-3.9, etc.)

Some vendors do publish a rating-band interpretation of the 1 to 5 average for teams that want both views. One commonly cited scale: 4.5 and above sits in the top quartile, 4.0 to 4.4 is above average, 3.5 to 3.9 is average, 3.0 to 3.4 is below average, and anything under 3.0 needs immediate action. This is useful as a second sanity check on top of the CSAT percentage, not as a replacement for it. Report the CSAT percentage as the primary number and the average as context.

How to Interpret Your CSAT Score Against the Benchmark

A single CSAT number is almost never actionable on its own. Three moves make it one.

Step 1: Pick the Right Baseline

Choose ONE baseline before you look at your number. If you are a bank, that is the ACSI banking benchmark (80). If you are a retailer, that is the ACSI retail cluster (78 to 80). If your sector is not covered by a current ACSI study, use the national baseline of 76.9 and label it clearly. Writing down the baseline before you look at your number prevents you from retrofitting the benchmark to make your score look better.

Step 2: Weight for Sample Size and Response Rate

A per-agent CSAT of 100% based on 3 responses is not a coaching signal, it is noise. Set a minimum response threshold (usually 15 for agent-level, 30 for queue-level) and label anything below that as "insufficient sample" instead of reporting a misleading percentage. If your response rate drops below 5%, the number you are reporting is from a self-selected minority and should carry a caveat whenever it is reported externally.

Step 3: Compare Trend, Not Snapshot

A one-week CSAT is noise. A 4-week rolling average compared to the same 4 weeks last year is a trend. The useful question is not "is 78% a good score" but "is our 78% moving up or down, and against what context." Pair every weekly number with a 4-week rolling average and a year-over-year comparison before you act on it.

How to Move a Below-Benchmark CSAT Score

If you land below your industry benchmark, the number itself does not tell you how to fix it. Four plays, in order, do.

Segment Before Acting

A company-wide CSAT is a report, not a lever. The useful score is the one broken down by agent, queue, channel, shift, and interaction reason. One or two of these dimensions almost always carries most of the deficit. Start there before you launch any coaching program, change any policy, or buy any software.

Close the Agent Knowledge Gap

The single highest-ROI play in most contact centers is making sure every agent has the right answer to the most common questions in the moment they need it. That is almost never a "retrain the team" problem. It is a "put the answer in front of the agent while they are on the call" problem.

Automate QA to 100% Coverage

Manual QA scores 1 to 3% of calls. Below-benchmark CSAT with 1 to 3% QA coverage means you are looking at a sample so small that your QA is telling you almost nothing about what is actually happening in the other 97%. Automated QA scores every call against the same standards and flags the interactions most likely to drive low CSAT before the survey goes out. See our guide to automated QA software for the operational fundamentals.

Run Real-Time Agent Assist on the Same Standards

The closed loop is where the number actually moves. When agent assist runs the same playbook QA scores against, frontline agents fix in the moment what they would otherwise lose a CSAT point on. This is what lifts QA scores 10 to 20 points in the first months of deployment, which is a strong leading indicator of CSAT lift of similar magnitude. For the broader set of coaching plays, see our guide to proven ways to improve CSAT .

CSAT Benchmark Comparator

Enter your CSAT percentage, pick your industry and primary channel, and the tool below shows where you sit against the ACSI or vendor-cited benchmark plus a quick operational read.

How Balto Helps Teams Beat CSAT Benchmarks

Balto automates the parts of CSAT measurement and improvement that most teams try to do with spreadsheets and point tools. Insights scores CSAT across every interaction and segments the result by agent, queue, channel, and topic, which is the dimension where the benchmark gap usually lives. QA scores 100% of calls against the same standards that drive CSAT, so you are not sampling 1 to 3% of calls and hoping the sample is representative.

Real-time agent assist runs on those same standards during the live interaction. Agents are guided through the moments that drive CSAT, in the moments they happen, and the system learns from what works. That is what produces the 10 to 20 point QA lift that sits as a leading indicator for CSAT lift of similar magnitude. You stop benchmarking against the industry average and start setting it.

Benchmarking CSAT Is Only Useful If You Act On It

A benchmark tells you where you are. It does not tell you what to do next. The teams that actually move their CSAT number past the industry benchmark are the ones that treat the benchmark as the starting condition and build a closed loop of segmentation, agent coaching, 100% QA, and real-time guidance on top of it. For the broader CSAT definition and framework, see our full guide to CSAT in customer service .

FAQs