Customer Effort Score

The customer effort score (CES) measures how much work a customer does to solve an issue, finish a purchase, or get a request handled. Low effort predicts loyalty better than delight. It’s a one-question survey, scored 1–7, fired right after the contact.

Matthew Dixon, Karen Freeman and Nicholas Toman popularised the metric in their 2010 Harvard Business Review essay, “Stop Trying to Delight Your Customers”. That historical paper argued reducing effort beats exceeding expectations.

CES sits beside customer satisfaction (CSAT) and net promoter score (NPS) in the voice-of-customer stack, but it does a different job. CSAT tells you how a customer felt; NPS tells you if they’d recommend you.

CES tells you why they might quit. That’s the reason contact centres treat it as an operating metric rather than a mood reading — a bad score names a broken step you can go and fix.

Key takeaways

• CES asks one question, usually “the company made it easy for me to handle my issue.”
• Research from advisory firm Gartner found effortless-service customers were 94% more likely to repurchase.
• The survey fires right after an interaction, so recall stays sharp and root causes stay visible.
• CES complements CSAT and NPS rather than replacing either of them.
• Contact centres and BPOs run CES as an operational KPI, not just a satisfaction check.

How it works

CES asks a single agree-or-disagree statement right after a support ticket, chat, purchase or self-service session. Respondents pick a number on a 1–7 or 1–5 scale where higher means easier, and the average across all responses is the score.

Two versions dominate the field. CES 1.0, launched in 2010, asked “how much effort did you personally have to put forth to handle your request?” on a 1–5 scale where lower scores were better.

CES 2.0, which the Corporate Executive Board — now part of Gartner — introduced in 2013, flipped the phrasing and the scale so higher numbers signal easier experiences, closer to how CSAT and NPS behave.

Where the survey fires matters more than which version you pick. A prompt inside the closed-ticket email catches the interaction; one buried in a quarterly newsletter catches nothing but mood.

The point isn’t the number — it’s step 4. A CES programme that never loops back into coaching and product design is telemetry without a steering wheel.

Sample size matters as much as phrasing. A CES average built on fewer than 100 responses per channel per month swings wildly, so most CX teams report a rolling 30-day figure beside the raw count.

Segment before you act. A blended 5.8 can hide a 4.1 on billing disputes, and the fix lives inside that segment rather than in the headline number.

Self-service journeys are the usual CES culprit. When a help-centre article half-answers a question, the customer starts again in chat, and the score records both attempts as one hard experience.

Examples

Named programmes make the metric concrete. Four teams, across three sectors and two continents, treat CES as an operating lever rather than a vanity number, and each one publishes enough detail to copy.

Zendesk, the San Francisco customer-service platform, ships CES as a native survey type inside its Support product. Its 2024 CX Trends report found 71% of consumers switch brands after a hard-to-resolve issue.

HubSpot, the Cambridge, Massachusetts marketing and service platform, added a native CES tool to Service Hub in 2019 and now scores every closed conversation. Support leads filter low-CES tickets by agent, a fast path from data to coaching.

Vanguard, the Pennsylvania asset manager, credits its CES-driven redesign of the account-transfer journey with cutting call volume by double digits, a case Gartner has cited since 2015.

The fix was pure friction removal: fewer forms, no re-authentication step, and one confirmation email instead of three.

The Australian Taxation Office (ATO) runs CES against its tax-agent portal each quarter and publishes the results in its annual report, a rare government signal that treats effort as a public-service KPI.

Offshore contact centres apply the same loop. Philippine BPOs running client CX desks pair CES with first call resolution, because a poor effort score beside strong FCR usually points at the process, not the agent.

Related terms

CES rarely travels alone. It sits inside a small family of customer metrics and CX disciplines, and reading it next to its neighbours is what turns a single number into a decision you can defend.

• Customer Satisfaction Score (CSAT): a happiness snapshot of how one interaction felt, where CES measures effort instead.
• Net Promoter Score (NPS): a long-horizon advocacy measure, while CES stays short-horizon and event-specific.
• First Call Resolution (FCR): the operational cousin that tracks whether the fix stuck on call one.
• Voice of the Customer (VoC): the umbrella feedback programme that CES scores feed into.
• Customer Experience (CX): the wider discipline that owns the metric and the journeys behind it.
• Self-Service: the channel whose design most directly raises or sinks CES numbers.
• Service Level Agreement (SLA): the contract clause where outsourced CES targets and penalties get written down.

FAQ

What’s a good customer effort score?

On the 1–7 CES 2.0 scale, most benchmarks put a healthy score at 5.5 or above, and anything under 5 flags systemic friction. SaaS and fintech firms typically score higher than telco and utilities.

How is CES calculated?

Add every response and divide by the number of responses; that mean is your CES. Some teams also report the share of respondents who answered 5 or above, which reads more like a top-box CSAT figure.

Is CES better than NPS?

Neither wins in isolation. CES is the sharper predictor of churn after a service event, while NPS captures brand-wide advocacy over time. Most mature CX teams run both, add CSAT, and triangulate.

When should you send a CES survey?

Fire it right after the touchpoint you want to measure: ticket close, chat end, order delivery or onboarding completion. Waiting more than 24 hours drops response rates sharply and blurs memory of the interaction.

Can outsourcing partners be measured on CES?

Yes, and increasingly they are. Contact-centre BPOs in the Philippines and India often write CES targets into their service level agreements, tying agent bonuses to score bands.

What’s the difference between CES 1.0 and CES 2.0?

CES 1.0 asked about effort directly on a 1–5 scale where lower was better, while CES 2.0 reversed it to a 1–7 agreement scale.

If you want a partner that ships CES dashboards from day one, talk to Outsource Accelerator’s advisory team about a shortlist of vetted CX providers.