Customer Effort Score (CES): The Complete Guide for 2026

Customer Effort Score (CES) is a single-question survey metric that measures how much work a customer had to do to accomplish something with your product or team. The standard question asks customers to rate, usually on a 1 to 7 agreement scale, how strongly they agree that the company made it easy to handle their issue. You then average every response. The higher the average, the lower the effort. CES matters because effort predicts loyalty far better than delight does. Customers rarely churn because you failed to amaze them. They churn because something was annoying.

Picture a customer who loves your product but spent 40 minutes on a chat thread to cancel a duplicate charge. They got the refund. They were polite throughout. Then they quietly downgraded the next month. CSAT would have flagged that interaction as a 4 out of 5. NPS would have shown nothing. CES would have caught it on the spot, because the customer had to repeat their account ID three times and escalate to a supervisor. That is the gap CES is built to close.

Key Takeaways

CES measures perceived ease, not satisfaction or loyalty. It is the strongest single predictor of repurchase and renewal in transactional moments.
The formula is a simple average of all responses on a 1 to 7 (or 1 to 5) scale. Some teams report it as a percentage of "agree" responses (5, 6, or 7).
Send CES right after a specific interaction: support resolution, onboarding completion, checkout, feature adoption. Never send it as an annual relationship survey.
Use NPS for relationship health, CSAT for moment-to-moment happiness, and CES for friction. They are complements, not substitutes.
Wording matters more than the scale. Double-barreled questions and vague verbs destroy your data.
You can build a working CES survey in PollPe in under a minute and collect unlimited responses on the free tier.

What is Customer Effort Score?

Customer Effort Score is a transactional survey metric introduced by the Corporate Executive Board (CEB, now Gartner) in a 2010 Harvard Business Review article titled "Stop Trying to Delight Your Customers." The authors analyzed more than 75,000 customer interactions and found something the industry was not ready to hear: exceeding expectations did almost nothing for loyalty. Reducing effort did almost everything.

The original CES used a 1 to 5 scale and a clunky question about "how much effort did you personally have to put forth." CEB revised it in 2013 to what most teams use today, called CES 2.0:

"The company made it easy for me to handle my issue."

Strongly Disagree (1) ... Strongly Agree (7)

The shift to an agreement statement on a 7-point scale fixed two problems. It read more naturally in the customer's voice. And it gave you more granularity to detect drift over time.

CES sits in the same family as NPS and CSAT but answers a narrower question. NPS asks "would you recommend us." CSAT asks "were you satisfied." CES asks "was that easy." Of the three, CES correlates most tightly with future behavior in transactional moments. The CEB study found that 94 percent of customers who reported low effort said they would repurchase. Only 4 percent of high-effort customers said the same.

How to Calculate CES (with examples)

The formula is straightforward:

CES = (Sum of all individual scores) / (Total number of responses)

Say you sent a post-support CES survey and got 200 responses on a 1 to 7 scale. The total of all scores is 1,140. Your CES is 1,140 / 200 = 5.7. That is a healthy number for SaaS support, where most teams land between 5.2 and 6.0.

Some teams prefer a percentage version, similar to NPS top-box scoring:

CES % = (Number of responses scoring 5, 6, or 7) / (Total responses) × 100

If 156 of those 200 respondents scored 5 or higher, your CES % is 78 percent. Both methods are valid. Pick one and stay consistent. Mixing them across quarters makes trends meaningless.

A small note on scale choice. The 7-point agreement scale is the industry default and what most benchmark data uses. A 1 to 5 scale is acceptable if you are matching an existing CSAT scale and want visual consistency in your dashboards. Avoid 1 to 10. It dilutes the signal and creates an awkward middle.

CES vs NPS vs CSAT: Which to Use When

These three metrics get blurred together constantly, especially in early-stage CX programs. Here is the cleanest way to think about them.

Metric	Question Type	Best For	Scale	Timing
CES	"Was that easy?"	Transactional friction	1 to 7 agreement	Right after an interaction
NPS	"Would you recommend us?"	Relationship and brand health	0 to 10	Quarterly or post-renewal
CSAT	"Were you satisfied?"	Specific touchpoints	1 to 5 satisfaction	Right after an interaction

CES and CSAT often look interchangeable. They are not. CSAT measures emotional response. CES measures structural friction. A customer can be highly satisfied with a support agent who was kind and apologetic, while still rating the experience as high effort because the system made them repeat themselves four times. CSAT will hide that. CES will surface it.

Use NPS for the big picture. We have a full breakdown in our NPS survey best practices guide if you are setting up a relationship program.

The honest answer for most teams is that you should run all three, just at different moments. CES on transactions, CSAT on moments, NPS on relationships.

CES Question Wording That Actually Works

The single biggest mistake teams make with CES is rewriting the question. Resist the urge. The CEB 2.0 wording was tested across thousands of interactions and translated into dozens of languages. It works.

Recommended wording:

"The company made it easy for me to handle my issue."

Replace "the company" with your brand name. Replace "handle my issue" with the specific action if it is more natural. For onboarding, "complete setup." For checkout, "place my order." For a feature trigger, "create my first report."

Bad examples to avoid:

"How easy or difficult was it to use our product and resolve your issue today?" (Double-barreled. Use is one thing, resolution is another.)
"Rate your effort." (Vague. Effort doing what?)
"Was our agent helpful and quick to resolve your problem?" (This is a CSAT question wearing CES clothing.)

Always pair the rating with one open-text follow-up. The single most valuable field on a CES survey is "What could we have made easier?" Anyone who scores 4 or below should see this. Anyone who scores 6 or 7 can skip it. That logic takes ten seconds to set up in PollPe with conditional branching.

For more on writing questions that produce clean data, see how to write survey questions.

When to Send a CES Survey (Trigger Points)

CES is a transactional metric. It only works when the interaction is fresh in the customer's memory. The window is roughly 24 hours, ideally less. Beyond that, recall blurs and you start measuring mood instead of effort.

The four canonical trigger points:

Post-support resolution. The classic CES use case. Fire the survey within an hour of ticket closure. Email or in-app, depending on your support channel.
Post-onboarding milestone. Right after a new user completes setup, imports data, or invites their first teammate. This catches activation friction your funnel metrics will miss.
Post-purchase or post-checkout. For ecommerce or self-serve SaaS, fire CES after the first successful transaction. You will learn whether your billing and provisioning flow has any sharp edges.
Post-feature adoption. When a user hits a key feature for the first time, ask if it was easy. This is gold for product teams shipping new flows.

Avoid sending CES on a fixed calendar cadence. Avoid bundling it with NPS in the same email. Avoid sending it to customers who did not actually complete the action you are measuring.

CES Benchmarks by Industry

Benchmark data should always be treated as directional. Your industry, customer segment, and survey wording all shift the numbers. With that caveat, here are the ranges most CX research firms publish for 7-point CES:

SaaS support: 5.2 to 6.0 average. Top-quartile teams hit 6.2.
B2B onboarding: 4.8 to 5.6. This is usually the lowest CES surface in SaaS, because setup is genuinely effortful.
Ecommerce checkout: 5.5 to 6.3. High because the flow is short.
Telecom and utilities support: 4.0 to 4.8. Notoriously low across the industry.
Consumer banking app interactions: 5.0 to 5.8.

A more useful frame than the absolute number is the trend. A CES that drops 0.3 points quarter over quarter is a flashing red light, even if your number is still "good." Track movement, not vanity scores.

Common Mistakes That Tank CES Data

After reviewing hundreds of CES programs, the same mistakes show up again and again. Watch for these:

Sending too late. A survey that arrives three days after a support ticket is a satisfaction survey. The effort signal is gone.
Surveying everyone, every time. Survey fatigue is real. Sample 30 to 50 percent of interactions, rotate intelligently, and exclude customers who got a survey in the last 14 days. Our guide to improving survey response rates covers the sampling math.
Double-barreled questions. "Was our service fast and helpful?" measures two things at once. Split it or pick one.
No open-text follow-up. A score without a reason is a number you cannot act on. Always ask the low scorers what to fix.
Mixing CES and CSAT in the same field. Pick one metric per question. If you want both, ask them as separate questions.
Ignoring the middle. Teams obsess over detractors and promoters but ignore the 4s and 5s. That is where the largest population of "quietly frustrated" customers lives.
Reporting CES without segmenting. Aggregate CES is almost meaningless. Segment by plan, by region, by support channel, by tenure. The interesting story is always in the segments.

Building a CES Survey in PollPe (60 seconds)

Setting up a working CES survey in PollPe takes about a minute. Here is the flow.

Open app.pollpe.com and click "Create with Aria." Aria is our AI survey assistant. The Standard mode is free on every plan, including the free tier, so you do not need to upgrade to try it. Type something like: "Build a post-support CES survey using the standard 7-point agreement scale, with a follow-up open question for anyone who scores 4 or below." Aria drafts the full survey, including the conditional branching, in a few seconds. You review, tweak, and publish.

If you want deeper analysis of incoming responses, sentiment clustering, theme extraction, and trend detection, that runs through Aria Deep Analysis, which is included on Business and higher plans.

A few things worth flagging if you are comparing tools. PollPe gives you unlimited responses on the free tier. Typeform caps you at 10 responses per month on its free plan, and SurveyMonkey limits both responses and exports. If you are running a real CES program with a few thousand monthly tickets, those caps make the "free" plans unusable in week one. PollPe was built specifically so that response volume is never the constraint that pushes you onto a paid plan.

The builder UX is intentionally close to what your team probably already knows from Typeform, with the same one-question-at-a-time conversational feel and clean conditional logic. The pricing sits at a fraction of what you would pay there. Starter is ₹400 per month for small teams. Business is ₹2,500 per month and includes multi-language surveys across 15 languages, including Hindi, Telugu, and Tamil, which matters if you are running CES across India or any multilingual market. If pricing is the reason you have been holding off on better tooling, our Typeform alternatives breakdown walks through the math.

Start building your CES survey on the free tier in under a minute.

FAQs

Is CES better than NPS? Neither is "better." They measure different things. CES is the stronger predictor in transactional moments because effort is what makes customers churn quietly. NPS is the stronger signal for brand health and word-of-mouth. Run both.

What scale should I use, 1 to 5 or 1 to 7? Use 1 to 7 unless you have a specific reason not to. It is the industry standard, gives you more granularity, and matches the CEB 2.0 wording. Use 1 to 5 only if you need visual consistency with an existing CSAT scale.

How often should I send CES surveys? CES is event-triggered, not calendar-triggered. Send it right after the interaction you want to measure. For a single customer, cap exposure at one CES survey per 14 days to avoid fatigue.

Can CES predict churn? Yes, in transactional contexts. The CEB research found that low-effort experiences correlate with a 94 percent repurchase rate. High-effort experiences drop that to 4 percent. CES will not predict churn from pricing or competitive pressure, but it is excellent at predicting churn from operational friction.

How large a sample do I need? For directional segment-level analysis, aim for at least 100 responses per segment per period. For statistical significance on trend detection, you want 200 or more. If you are running a low-volume B2B program, focus on qualitative follow-up text rather than chasing statistical thresholds.

Should I show CES results to customers? Generally no. Unlike NPS, which some brands publish, CES is an internal operational metric. Use it to find and fix friction, not as a marketing badge.

What CES score should I aim for? On the 7-point scale, 5.5 is a reasonable internal floor for most SaaS teams. Anything below 5.0 indicates a structural problem in the touchpoint. Anything above 6.0 puts you in the top quartile.

Does CES work in non-English markets? Yes, with proper translation. The agreement-scale construct translates cleanly across most languages. If you are running surveys in Hindi, Telugu, Tamil, or another regional language, make sure the translation preserves the "ease" framing rather than substituting "satisfaction." For more on multilingual setup, see our notes on customer experience feedback questions.