The statistics library

612 sourced numbers on customer support

The figures cited across reports, landscape guides, benchmarks, and dispatches — each traceable to its linked source and context. Repeated figures stay separate when their publication context differs. Copy a citation or share a number straight to LinkedIn.

427 numbersClear search and filters

63%

consumers willing to switch to a competitor after just one bad experience — up 9% year on year

Zendesk CX Trends 2025 Report Insight: Your CSAT response rate is lying to you

78%

industry-average call-center CSAT (good = 75-84%, world-class = 85%+)

SQM Group, Call Center Industry Standards / KPI Benchmarks (2024) Insight: Your CSAT response rate is lying to you

satisfaction yes; behaviour no

What the largest meta-analysis of recovery studies found the paradox moves: a significant lift to satisfaction, but no significant effect on repurchase intention, word-of-mouth, or corporate image

de Matos, Henrique & Rossi, 'Service Recovery Paradox: A Meta-Analysis,' Journal of Service Research 2007 Insight: The service-recovery paradox, tested

5.4%

Share of real service encounters in which the paradox actually occurred, across 11,000+ retail-banking interviews — and only when initial service was merely 'satisfactory,' never 'very satisfying'

Michel & Meuter, 'The Service Recovery Paradox: True but Overrated?,' Journal of Service Management 2008 Insight: The service-recovery paradox, tested

4 conditions

The paradox appears only when the failure is minor, the customer's first with the firm, caused by an unstable one-off, and outside the firm's control — change any one and it collapses

Magnini, Ford, Markowski & Honeycutt, 'The Service Recovery Paradox: Justifiable Theory or Smoldering Myth?,' Journal of Services Marketing 2007 Insight: The service-recovery paradox, tested

~80% of price

How large a make-good had to be — roughly 80% of the original price — before post-recovery satisfaction beat an error-free experience, in a 2022 study of the compensation threshold

Edstrom et al., 'Where service recovery meets its paradox,' Journal of Service Theory and Practice 2022 Insight: The service-recovery paradox, tested

1.5M riders; cash > apologies

In a field experiment across ~1.5 million ride-hailing customers, monetary compensation beat verbal apologies and repeated apologies sometimes did worse than none — a failed recovery is worse than the failure

'Toward an Understanding of the Economics of Apologies,' NBER Working Paper 25676 Insight: The service-recovery paradox, tested

1.6x

How much more likely a customer is to buy again when you move them from a 1-2 star experience up to a merely-fine 3-star one — the realistic recovery target, not delight

Qualtrics XM Institute 2024 (reported by CX Dive) Insight: The service-recovery paradox, tested

70%

Average first-contact resolution across call-center industries (good = 70–79%, world-class = 80%+)

SQM Group, Call Center Industry Standards / KPI Benchmarks Insight: What good looks like: support benchmarks by team size and industry

34.4 seconds

Benchmark average speed of answer for call centers, against the 80/20 service-level rule

SQM Group, Call Center Industry Standards / KPI Benchmarks Insight: What good looks like: support benchmarks by team size and industry

Average NPS for B2B software and SaaS — the lowest-scoring B2B vertical in the benchmark

Retently, What is a Good Net Promoter Score? (2026 NPS Benchmarks) Insight: What good looks like: support benchmarks by team size and industry

14%

Share of service issues customers fully resolve in self-service, with no human (only 36% even for 'very simple' issues)

Gartner survey, Aug 2024 (reported by destinationCRM) Insight: What good looks like: support benchmarks by team size and industry

88%

2023 industry benchmark internal quality score (IQS) from QA conversation reviews

Zendesk QA (Klaus), Customer Service Quality Benchmark Report 2023 (cited on Zendesk blog) Insight: What good looks like: support benchmarks by team size and industry

76.9 / 100

US national cross-industry CSAT — essentially flat and not materially higher since 2017

American Customer Satisfaction Index (ACSI), National Q4 2025 Press Release Insight: What good looks like: support benchmarks by team size and industry

78%

Industry-average call-center (phone) CSAT — higher than most async channels

SQM Group, Call Center Industry Standards / KPI Benchmarks Insight: What good looks like: support benchmarks by team size and industry

Average NPS for financial services, the top-scoring B2B industry in the benchmark

Retently, What is a Good Net Promoter Score? (2026 NPS Benchmarks) Insight: What good looks like: support benchmarks by team size and industry

63%

Consumers willing to switch to a competitor after just one bad experience (up 9% year-on-year)

Zendesk CX Trends 2025 Report Insight: What good looks like: support benchmarks by team size and industry

$8.01 vs $0.10

What Gartner found a live-channel contact (phone, email, chat) costs on average, against about $0.10 for a self-service contact — roughly an 80x gap.

Gartner, 2019 Customer Service & Support Leader poll (reported by destinationCRM) Insight: Blended cost-per-contact hides the customers who actually cost you

$1.84 vs $13.50

Median cost per contact for self-service versus human-assisted channels — a single 'blended' average sits between them and describes neither.

Unthread, Support Cost per Resolution Statistics (Gartner data) Insight: Blended cost-per-contact hides the customers who actually cost you

$17-25

Cost per contact for phone/voice, the most expensive assisted channel — several times the cost of a chat or email contact.

Unthread, Support Cost per Resolution Statistics (Which-50 data) Insight: Blended cost-per-contact hides the customers who actually cost you

Share of customers who report fully resolving their issue through self-service — the cheap channel frequently fails and reprices the contact as an expensive one.

Gartner, press release, Sept 2019 (Customer Service & Support Leader poll) Insight: Blended cost-per-contact hides the customers who actually cost you

150-300%

Share of total profit generated by a firm's most profitable 20% of customers; the least profitable 10-20% erase 50-200% of it (the 'whale curve'). Cost-to-serve, not revenue, drives the split.

Kaplan & Narayanan (2001), 'Measuring and Managing Customer Profitability' — via Wikipedia: Customer Profitability Analysis Insight: Blended cost-per-contact hides the customers who actually cost you

~$0.50

Cost per contact for an AI agent or chatbot — the new bottom of the range, and why blended cost-per-contact is now a moving target.

Unthread, Support Cost per Resolution Statistics (Fullview data) Insight: Blended cost-per-contact hides the customers who actually cost you

40%

of productive time can be lost to shifting between tasks — the tax every extra chat quietly imposes

American Psychological Association — Multitasking: Switching costs Insight: Concurrency: the hidden cap on chat quality

35 seconds

median first-response time customers now get on live chat — the speed promise concurrency has to keep

LiveChat — Customer Service Report (2024) Insight: Concurrency: the hidden cap on chat quality

27.4%

of customers abandon the chat queue before an agent ever replies

LiveChat — Customer Service Report (2024) Insight: Concurrency: the hidden cap on chat quality

64.2%

average live-chat CSAT across rated conversations — well below the channel's happy reputation

LiveChat — Customer Service Report (2024) Insight: Concurrency: the hidden cap on chat quality

4.1 / 5

overall live-chat satisfaction across 220M+ interactions in 18 industries

Comm100 — 2026 Live Chat Benchmark Report Insight: Concurrency: the hidden cap on chat quality

2–3

concurrent chats — the working ceiling experienced practitioners converge on; new agents cap at one

Call Centre Helper — How to Handle Multiple Chat Conversations Effectively Insight: Concurrency: the hidden cap on chat quality

9 in 10

CX leaders who say human-handled contacts are getting more complex as automation absorbs routine volume

Cresta contact-center leader survey, reported by CX Current Insight: The judgment calls no macro can make

83.3%

Average maximum occupancy target entered by contact-center professionals across more than 160,000 calculations

Call Centre Helper, How to Calculate Occupancy in the Contact Centre Insight: The weekly support operating review

~2%

Share of customer conversations commonly reviewed under traditional manual QA

Zendesk QA, QA automation overview Insight: Calibration clinic: ten disputed support interactions

20–35%

Typical paid time consumed by shrinkage such as training, meetings, breaks, and absence

Society of Workforce Planning Professionals, Calculating Call Center Staff Insight: From forecast to intraday: the complete WFM planning loop

25–50%

Near-term improvement in resolution times reported after adopting KCS practices

Consortium for Service Innovation, KCS v6 Adoption Guide Insight: Knowledge ownership without a knowledge team

95%

Consumers who expect an explanation for decisions made by AI in customer service

Zendesk CX Trends 2026 Insight: Escalation authority: who decides what when risk rises

31%

Contact-center agents who reported being likely to leave their role within six months

Verint State of Agent Experience 2026, reported by No Jitter Insight: How to build a support portfolio that proves scope

+34%

Productivity gain observed for the least-experienced agents using a generative-AI assistant

Brynjolfsson, Li & Raymond, NBER Working Paper 31161 Insight: The human-control map for support automation

7 minutes

Industry benchmark average handle time for call centers — the middle term in every staffing calculation

SQM Group, Call Center Industry Standards / KPI Benchmarks Insight: AHT is a budgeting metric, not a coaching metric

34.4 seconds

Average speed of answer benchmark, measured against the 80/20 service-level rule that AHT helps you staff for

SQM Group, Call Center Industry Standards / KPI Benchmarks Insight: AHT is a budgeting metric, not a coaching metric

70%

Average first-call resolution across industries — 'good' is 70-79%, world-class is 80%+ — the metric rushing quietly erodes

SQM Group, Call Center Industry Standards / KPI Benchmarks Insight: AHT is a budgeting metric, not a coaching metric

88%

2023 industry benchmark internal quality score from QA conversation reviews — what a stopwatch can never see

Zendesk QA (Klaus), Customer Service Quality Benchmark Report 2023 (cited on Zendesk blog) Insight: AHT is a budgeting metric, not a coaching metric

63%

Share of consumers willing to switch to a competitor after just one bad experience, up 9% year on year

Zendesk CX Trends 2025 Report Insight: AHT is a budgeting metric, not a coaching metric

46%

Share of their time service reps actually spend with customers — the rest goes to administrative and internal tasks, the bucket after-call work lives in

Salesforce, State of Service (customer service statistics) Insight: After-call work: the invisible queue

30-90 sec

Typical after-call work per call across the industry, and one of the three components of average handle time alongside talk time and hold time

Balto, 'What Does After-Call Work (ACW) Mean in a Call Center?' Insight: After-call work: the invisible queue

3 minutes

Average time agents spend searching for answers in the 45% of calls that require it — hidden work that spills into handle time and wrap-up

Verint, State of Agent Experience 2026 (survey of 1,000 agents) Insight: After-call work: the invisible queue

31%

Contact-center agents who say they are likely to leave their role within the next six months as repetitive busywork erodes the job

Verint, State of Agent Experience 2026 Insight: After-call work: the invisible queue

85%

Occupancy ceiling contact centres are advised not to exceed — and after-call work sits in the handling-time numerator that drives it

Call Centre Helper, 'How to Calculate Occupancy in the Contact Centre' Insight: After-call work: the invisible queue

40-60 sec

Per-call after-call work that DailyPay agents saved once AI auto-summaries replaced typing notes from scratch

Observe.AI, 'Hidden Benefits of Contact Center AI Summarizer Tools' Insight: After-call work: the invisible queue

The Support Brief

The most useful new numbers, sourced in one monthly issue.

Monthly. No ads. Unsubscribe anytime. Privacy