BookbagBookbag
Benchmarks

CSAT vs CES vs NPS: Which to Track and Benchmarks

Three metrics, three questions: was this interaction satisfying? How hard was it? And would you recommend us? Here's which ones matter most for ecommerce — and the benchmarks to aim for.

The Bookbag Team·June 2026· 10 min read

The three metrics defined

CSAT, CES, and NPS are all customer feedback metrics, but they measure fundamentally different things and are appropriate at different moments in the customer journey.

CSAT — Customer Satisfaction Score

CSAT measures how satisfied a customer was with a specific interaction. It's asked immediately after a support ticket closes: 'How satisfied were you with your support experience?' — rated on a 1–5 scale or thumbs up/down. CSAT is interaction-level: it reflects the quality of that specific conversation, not the customer's overall relationship with the brand.

CSAT is the most commonly used support quality metric for ecommerce because it's directly tied to the interaction being measured and gives immediate, actionable feedback.

CES — Customer Effort Score

CES measures how easy it was for the customer to resolve their issue. It's asked post-interaction: 'How easy was it to get help?' — typically rated on a 1–7 scale. CES is based on the insight that reducing customer effort is more predictive of loyalty than maximizing delight. It's particularly relevant for self-service and support channel design.

CES is less commonly used in ecommerce support than CSAT, but it's useful for diagnosing channel and process friction — if customers find it hard to get help, that shows up in CES before it shows up in churn.

NPS — Net Promoter Score

NPS measures overall brand loyalty: 'How likely are you to recommend [Brand] to a friend or colleague?' — rated 0–10. NPS is relationship-level, not interaction-level: it reflects the customer's cumulative experience across all touchpoints, not just the support interaction. NPS is typically surveyed periodically (quarterly or annually) rather than post-interaction.

For support teams specifically, NPS is a lagging indicator — it reflects many factors beyond support quality. It's a useful executive metric but not ideal for measuring support performance specifically.

Benchmarks for each metric

NPS for ecommerce brands varies widely by category — luxury and specialty brands often run higher NPS than commodity or high-return-rate categories. The ranges above are for general ecommerce. CSAT ranges are for post-interaction surveys specifically. CES ranges assume the 'easy to get help' framing on the 1–7 Likert scale.

MetricScaleIndustry typical (ecommerce)StrongExcellent
CSAT% positive (thumbs up or 4–5 rating)75–85%88–92%93%+
CES (effort score)1–7 (lower = easier)4.5–5.5 avg3.5–4.5 avgUnder 3.5 avg
CES (easy + somewhat easy %)% rating 5–7 of 755–70%72–82%85%+
NPS-100 to +10020–4545–6565+
Important caveat

Survey response rates affect all three metrics. Post-ticket CSAT response rates typically run 10–30%; low response rates bias toward vocal extremes. NPS surveys see 5–20% response rates on average. CES surveys are similar to CSAT. Improving response rates (asking immediately, keeping it to one question, embedding in the ticket close email) gives you a more representative signal.

Which metric to prioritize for ecommerce support

For most ecommerce support teams, CSAT is the right primary metric. It's interaction-level (tied directly to the support experience), immediate (feedback while the interaction is fresh), and widely understood. It's the metric that most directly answers 'is our support good?'

CES is a valuable secondary metric, particularly if you suspect process friction — customers who find it hard to reach support or navigate the return process will show up in CES before CSAT. Add CES when CSAT is strong but you sense there's an effort problem you're not measuring.

NPS is a brand-level metric, not a support-level metric. If your company already runs NPS surveys, segment by customers who recently interacted with support vs. those who haven't — the difference tells you the net impact of support interactions on brand loyalty. But don't use NPS as your primary support quality metric — it's too far removed from the interaction to be actionable for support teams.

MetricBest forFrequencyUse case
CSATSupport quality trackingPer interactionPrimary support metric — track always
CESProcess friction diagnosisPer interaction or periodicAdd when you suspect channel/process friction
NPSBrand loyalty and retentionQuarterly / annuallyExecutive metric — segment by support interaction

How AI affects each metric

AI support affects CSAT, CES, and NPS in distinct ways. Understanding the mechanism helps you predict what will improve — and what requires separate investment.

CSAT is most directly affected by AI. For interactions that AI handles well (fast, accurate, complete), CSAT from AI-resolved contacts is comparable to or better than human-resolved contacts. Speed is the primary driver: customers who previously waited hours now get an instant answer, which shows up immediately in CSAT.

CES improves significantly with AI, because the primary measure of effort is 'how hard was it to get help?' — and AI reduces effort to near-zero for the customer. Instead of navigating to a help center, searching for an article, and reading it, the customer just types their question and gets the answer. That's a low-effort experience.

NPS is affected indirectly and over a longer time horizon. Customers who consistently get fast, accurate support are more likely to repurchase and recommend the brand. But NPS also reflects pricing, product quality, and many other factors — support is one input among many.

MetricAI impactTimelineKey condition
CSATPositive: instant, accurate responses raise scoresWithin weeks of deploymentAI must give accurate, complete answers
CESStrongly positive: near-zero effort for common questionsImmediateAI must be easy to access and use
NPSModerately positive: better support improves loyalty3–6 months lagEffect is real but diluted by other NPS factors

Running all three efficiently

The good news is that all three metrics can be collected with minimal survey overhead. The key is timing and brevity.

  1. 1CSAT: embed a one-click rating (thumbs up/down or 1–5 stars) in the ticket close email or the chat post-conversation summary. One question, immediate send, no link to a separate survey page.
  2. 2CES: add a second single question in the same ticket close email: 'How easy was it to get your issue resolved?' (1–7 scale). Placing it right after CSAT in the same email achieves similar response rates with minimal extra friction.
  3. 3NPS: run quarterly or twice a year via your email marketing tool. Segment responses by 'had a support interaction in last 90 days' vs. not — the delta is your support NPS impact.
  4. 4Review CSAT weekly at a minimum — watch for drops that signal a new problem (bad AI response, new ticket type, seasonal frustration).
  5. 5Review CES monthly and segment by channel. If chat CES is low (high effort), that points to a different problem than if email CES is low.
  6. 6Use NPS annually to evaluate whether support improvements are translating to brand-level loyalty — and to make the case for support investment to leadership.

Key takeaways

  • CSAT measures interaction quality, CES measures effort, NPS measures brand loyalty — all different things.
  • CSAT is the primary metric for ecommerce support teams: interaction-level, immediate, and actionable.
  • Ecommerce CSAT benchmarks: 75–85% typical, 88–92% strong, 93%+ excellent.
  • AI improves CSAT through speed and accuracy; it improves CES by reducing customer effort to near-zero.
  • NPS is a brand metric, not a support metric — useful for executives but not actionable for daily support management.

Frequently Asked Questions

Turn support into your competitive edge

Join the ecommerce teams resolving more tickets, answering 24/7, and turning support into a revenue channel with Bookbag.