Customer Retention Statistics Updated November 2026

Human Written & Fact Checked · Updated

195 Customer Retention Statistics for 2026

The best-known customer retention statistic is also the least settled. The 1990 Harvard Business Review article that started it said companies could lift profits by almost 100% by keeping 5% more of their customers. Bain's 2001 brief put the gain at more than 25% for financial services, and the 25% to 95% range comes from a 2014 article. This page gives each version with its own source.

Measured churn is easier to pin down, because companies publish it. Verizon reported consumer postpaid phone churn of 0.92% for 2025, Antenna put average monthly churn across US premium streaming services at 4.6%, and Costco's renewal rate in the US and Canada stood at 92.3% at the end of fiscal 2026. Each of those is a separate measure with its own definition, so the page keeps them apart.

The rest of the page covers net revenue retention at listed software firms, loyalty programmes, why customers switch, satisfaction indexes in the UK and North America, complaints to regulators and ombudsmen, customer service jobs and waiting times, and AI in customer service. Every figure links to the primary document and carries the date of the data.

Headline figures

Does keeping 5% more customers raise profits by 25% to 95%?

There is no single agreed number. The claim began in a 1990 Harvard Business Review article, was restated for financial services in 2001 and took its popular range from a 2014 article, and the three versions do not say the same thing.

The three statements are given as their authors made them. The 1990 article concerned service businesses, and the 2001 brief gave its figure for financial services only. None of them is a measured average across all industries. No primary study behind the "five to 25 times" acquisition cost figure was found for this page, so it appears only as the 2014 article states it.

What is Verizon's postpaid phone churn rate?

Below 1% a month for consumer phone lines on Verizon's own figures, and the 2025 rate was above the 2020 one. The answer also depends on which lines are counted.

0.67% 2020 0.71% 2021 0.81% 2022 0.83% 2023 0.83% 2024 0.92% 2025
View data table
Verizon Consumer wireless retail postpaid phone churn, 2020 to 2025, as reported in the company's annual filings. Each year comes from the release for that year; the link is to the latest one.
PeriodValue
20200.67%
20210.71%
20220.81%
20230.83%
20240.83%
20250.92%
Verizon Consumer wireless retail postpaid phone churn, 2020 to 2025, as reported in the company's annual filings. Each year comes from the release for that year; the link is to the latest one. (Verizon (Form 10-K filings), 2026)
  • Across all Verizon Consumer wireless retail connections, prepaid included, churn was 1.61% in 2025. Across retail postpaid connections of every device type it was 1.15%. (Verizon (2025 Form 10-K), 2026)
  • Verizon Business reported wireless retail postpaid phone churn of 1.23% in 2025, against 1.10% in 2024 and 1.13% in 2023. (Verizon (2025 Form 10-K), 2026)
  • Across all Verizon Business wireless retail postpaid connections, churn was 1.58% in 2025, 1.47% in 2024 and 1.48% in 2023. (Verizon (2025 Form 10-K), 2026)

Verizon defines wireless churn as the rate at which service to retail, retail postpaid or retail postpaid phone connections is terminated on average in the period. These are Verizon's own figures for its own segments. AT&T and T-Mobile each define churn in their own way, so the three carriers are shown in separate sections and are not ranked.

What is AT&T's postpaid phone churn rate?

AT&T's postpaid phone churn stayed under 1% a month in every year from 2018 to 2025 on its own reporting. It rose in 2025 after falling in 2024.

0.90% 2018 0.95% 2019 0.79% 2020 0.76% 2021 0.81% 2022 0.81% 2023 0.76% 2024 0.90% 2025
View data table
AT&T Mobility postpaid phone churn, average monthly rate, 2018 to 2025, as reported in the company's annual reports. Each year comes from the release for that year; the link is to the latest one.
PeriodValue
20180.90%
20190.95%
20200.79%
20210.76%
20220.81%
20230.81%
20240.76%
20250.90%
AT&T Mobility postpaid phone churn, average monthly rate, 2018 to 2025, as reported in the company's annual reports. Each year comes from the release for that year; the link is to the latest one. (AT&T (annual reports), 2026)
  • AT&T Mobility postpaid phone churn was 0.90% in 2025, against 0.76% in 2024 and 0.81% in 2023. (AT&T (2025 annual report), 2026)
  • Across all AT&T postpaid devices, churn was 1.05% in 2025, 0.92% in 2024 and 0.98% in 2023. (AT&T (2025 annual report), 2026)
  • AT&T had 74.2 million postpaid phone subscribers at the end of 2025. (AT&T (2025 annual report), 2026)
  • In the third quarter of 2026 AT&T reported postpaid phone churn of 0.92% and 379,000 postpaid phone net additions. (AT&T (Q3 2026 results), Oct 2026)

AT&T divides the subscribers who cancelled in a month by the subscribers at the start of that month, and averages the monthly rates over the period. Its 2022 to 2024 figures exclude disconnections from the shutdown of its 3G network. Postpaid churn across all devices includes tablets and wearables as well as phones. These figures are not comparable with other carriers.

What is T-Mobile's churn rate?

T-Mobile reported postpaid phone churn below 1% a month from 2020 to 2025. From 2026 it reports postpaid account churn, a different measure that does not continue the earlier series.

0.90% 2020 0.98% 2021 0.88% 2022 0.87% 2023 0.86% 2024 0.93% 2025
View data table
T-Mobile US postpaid phone churn, average monthly rate for the full year, 2020 to 2025, from the company's year-end results. Each year comes from the release for that year; the link is to the latest one.
PeriodValue
20200.90%
20210.98%
20220.88%
20230.87%
20240.86%
20250.93%
T-Mobile US postpaid phone churn, average monthly rate for the full year, 2020 to 2025, from the company's year-end results. Each year comes from the release for that year; the link is to the latest one. (T-Mobile US (year-end results), Feb 2026)
  • T-Mobile US postpaid phone churn was 0.93% in 2025, against 0.86% in 2024. (T-Mobile US (Q4 2025 results), Feb 2026)
  • T-Mobile US prepaid churn was 2.72% in 2025, against 2.73% in 2024, and 2.76% in the fourth quarter of 2025. (T-Mobile US (Q4 2025 results), Feb 2026)
  • T-Mobile US postpaid account churn was 1.09% in the third quarter of 2026, against 0.97% in the third quarter of 2025. (T-Mobile US (Q3 2026 results), Oct 2026)
  • Over the nine months to September 2026, T-Mobile US postpaid account churn was 1.04%, against 0.94% in the same period of 2025. (T-Mobile US (Q3 2026 results), Oct 2026)

These are T-Mobile's own figures. The 2021 figure includes the integration of Sprint customers, and the 2025 figure includes UScellular customers acquired in August 2025. The third-quarter 2026 release gives account churn and no phone churn, so the 2026 figures are not added to the chart. Prepaid and postpaid churn are separate measures.

What is the churn rate for streaming services in the US?

Streaming subscribers come and go quickly. On Antenna's panel data, cancellations outran sign-ups in the third quarter of 2026, and a group of repeat cancellers accounts for a large part of the movement.

  1. 1 Non-serial churners 61%
  2. 2 Standard serial churners 54%3-4 cancellations in two years
  3. 3 Super-heavy serial churners 38%7+ cancellations
Share of US premium streaming subscribers still subscribed three months after signing up, by cancellation history, for subscribers from the third quarter of 2023 on Antenna's panel data. (Antenna (serial churn analysis 2024), Dec 2024)
  • Churn across US premium streaming services was flat or lower year on year in 7 of the 11 months from September 2024 to August 2025, against only 2 of the previous 21 months. (Antenna (Premium SVOD 2025 review), Feb 2026)
  • US premium streaming subscriptions grew 7% in 2025, down from 12% in 2024. Gross additions also grew 7% in 2025. (Antenna (Premium SVOD 2025 review), Feb 2026)
  • US premium streaming services recorded 36.8 million gross additions and 38.5 million cancellations in the third quarter of 2026, a net loss of 1.7 million subscriptions. (Antenna (Premium SVOD Q3 2026), Nov 2026)
  • 78% of US premium streaming subscribers have signed up for an ad-supported plan at some point in the past four years. (Antenna (Adds and Ads report), Jun 2026)
  • There are more than 110 million ad-supported streaming plans in the US, excluding Amazon Prime Video. (Antenna (Adds and Ads report), Jun 2026)

Antenna divides the cancellations in a month by the subscribers at the end of the previous month, using its own panel of US consumer transaction records. It is a monthly rate for one category and is not comparable with carrier churn or with survey-based churn. The serial churner figures date from 2024.

What is a serial churner?

  • Antenna counted about 29.5 million US serial churners at the end of the third quarter of 2024: people with three or more premium streaming cancellations in two years. They made up 23% of subscribers, up from 11.8 million people at the start of 2022. (Antenna (serial churn analysis 2024), Dec 2024)
  • Serial churners accounted for 41% of gross additions and 42% of cancellations in US premium streaming in 2024 to the third quarter. (Antenna (serial churn analysis 2024), Dec 2024)

Where do cord cutters go after cancelling pay TV?

  • Within a month of cancelling pay TV, 31% of US cord cutters sign up for a new streaming service, five times the normal sign-up rate. (Antenna (cord cutters analysis), Aug 2026)
  • 72% of traditional cord cutters in the US already had at least one premium streaming subscription before they cancelled pay TV. (Antenna (cord cutters analysis), Aug 2026)

How many people cancel a streaming subscription?

A large minority of US consumers cancel at least one service every six months in Deloitte's surveys, and many of them come back. This is a survey measure of people, not a company churn rate.

  1. 1 Cancelled a service, 2026 43%
  2. 2 Cancelled a service, a year earlier 39%
  3. 3 Cancelled and returned, 2026 27%
  4. 4 Cancelled and returned, a year earlier 23%
Share of US consumers who cancelled at least one paid streaming video service in the previous six months, and who cancelled and then resubscribed to the same service, in Deloitte's May 2026 survey and a year earlier. (Deloitte (Digital Media Trends, fall 2026), Oct 2026)

Deloitte surveyed 3,518 US consumers online in May 2026 and 3,575 in October-November 2025, weighted to the census. Its churn figure is the share of people who cancelled any service in six months. It is not a monthly churn rate and is not set against Antenna's figures in the section above.

What churn do other subscription companies report?

It depends on the business and on how each company counts. Some report a monthly churn rate, others an annual retention or attrition rate, and the two kinds of figure cannot be set side by side.

  • SiriusXM average self-pay monthly churn was 1.5% in 2025, against 1.6% in 2024. (SiriusXM (Q4 2025 results), Feb 2026)
  • SiriusXM had 31.3 million self-pay subscribers at 31 December 2025, down 301,000 on a year earlier. (SiriusXM (Q4 2025 results), Feb 2026)
  • In the third quarter of 2026 SiriusXM average self-pay monthly churn was 1.5%, against 1.6% a year earlier, and it had 31.26 million self-pay subscribers at 30 September 2026. (SiriusXM (Q3 2026 results), Oct 2026)
  • SiriusXM added a net 5,000 self-pay subscribers in the third quarter of 2026, against a net loss of 40,000 a year earlier. (SiriusXM (Q3 2026 results), Oct 2026)
  • Peloton's average net monthly paid Connected Fitness subscription churn was 2.2% in the quarter to 30 June 2026, against 1.8% a year earlier and 1.2% in the previous quarter. (Peloton (Q4 fiscal 2026 results), Aug 2026)
  • In the quarter to 30 September 2026 the same Peloton measure was 1.7%, against 1.6% a year earlier. (Peloton (Q1 fiscal 2027 results), Nov 2026)
  • Zoom's online customers, who buy on a self-serve basis, had average monthly churn of 2.9% in the quarter to 31 July 2026. (Zoom (Q2 fiscal 2027 results), Aug 2026)

Every figure here is the company's own measure. SiriusXM counts subscribers who pay for themselves. Peloton nets reactivations against cancellations and pauses. Zoom measures lost monthly recurring revenue. The companies are listed, not ranked.

Which companies report annual retention or attrition?

  • ADT, the US home security company, reported trailing twelve-month gross customer revenue attrition of 13.1% at the second quarter of 2026, against 12.8% a year earlier. (ADT (Q2 2026 results), Jul 2026)
  • Lemonade, the insurer, reported annual dollar retention of 85% at 30 June 2026, up 1 percentage point on the second quarter of 2025. (Lemonade (Q2 2026 shareholder letter), Jul 2026)
  • Frontdoor, the US home warranty company, reported a customer retention rate of 79.7% at 30 September 2026, against 79.4% a year earlier. (Frontdoor (Q3 2026 results), Nov 2026)
  • American Express Global Business Travel, which manages corporate travel, reported a customer retention rate of 95% in the second quarter of 2026, including CWT. (American Express Global Business Travel (Q2 2026 results), Aug 2026)

ADT measures revenue lost over twelve months. Lemonade measures the share of in-force premium retained over twelve months. American Express Global Business Travel sells to businesses and bases its rate on transaction value. Each is that company's own definition.

How many broadband customers did Comcast lose?

  • Comcast reported net losses of 144,000 domestic residential broadband customers in the third quarter of 2026, against net losses of 91,000 in the third quarter of 2025. (Comcast (Q3 2026 results), Oct 2026)
  • In the same quarter Comcast reported 368,000 domestic wireless line net additions and 47.8 million residential connectivity customer relationships. (Comcast (Q3 2026 results), Oct 2026)

Comcast publishes net customer losses, which are disconnections minus additions. It does not publish a churn rate in the release.

What net revenue retention do software companies report?

There is no shared formula, so there is no league table. Each company defines the measure itself: a rate above 100% means the same customers spent more than a year earlier, and a rate below 100% means they spent less.

178% Jan 2022 158% Jan 2023 131% Jan 2024 126% Jan 2025 125% Jan 2026
View data table
Snowflake net revenue retention rate at each fiscal year end, 31 January 2022 to 31 January 2026, on the company's own definition. Each year comes from the release for that year; the link is to the latest one.
PeriodValue
Jan 2022178%
Jan 2023158%
Jan 2024131%
Jan 2025126%
Jan 2026125%
Snowflake net revenue retention rate at each fiscal year end, 31 January 2022 to 31 January 2026, on the company's own definition. Each year comes from the release for that year; the link is to the latest one. (Snowflake (fiscal year-end results), Feb 2026)
  • Snowflake reported a net revenue retention rate of 126% for the quarter ended 31 July 2026. (Snowflake (Q2 fiscal 2027 results), Sep 2026)
  • Amplitude reported pro forma net revenue retention of 105% for the quarter ended 30 September 2026. (Amplitude (Q3 2026 results), Nov 2026)
  • Asana reported a dollar-based net retention rate of 97% across all customers for the quarter to 31 July 2026. (Asana (Q2 fiscal 2027 results), Sep 2026)
  • Asana's rate was 98% for its Core customers and 98% for customers spending $100,000 or more a year. (Asana (Q2 fiscal 2027 results), Sep 2026)
  • BlackLine reported a dollar-based net revenue retention rate of 103% at 30 September 2026. (BlackLine (Q3 2026 results), Nov 2026)
  • Braze reported dollar-based net retention of 110% for all customers for the twelve months to 30 April 2026, against 109% a year earlier. (Braze (Q1 fiscal 2027 results), May 2026)
  • For Braze customers with annual recurring revenue of $500,000 or more, the rate was 111%. (Braze (Q1 fiscal 2027 results), May 2026)
  • GitLab reported a dollar-based net retention rate of 117% for the quarter ended 30 April 2026. (GitLab (Q1 fiscal 2027 results), Jun 2026)
  • Klaviyo reported a net revenue retention rate of 110% at 31 March 2026, up two percentage points year on year. (Klaviyo (Q1 2026 results), May 2026)
  • PagerDuty reported a dollar-based net retention rate of 97% at 30 April 2026, against 104% a year earlier. (PagerDuty (Q1 fiscal 2027 results), May 2026)
  • Procore Technologies reported a gross revenue retention rate of 95% and a net revenue retention rate of 106% for 2025. (Procore (Q4 2025 results), Feb 2026)
  • Sprout Social reported a dollar-based net retention rate of 100% in 2025, against 104% in 2024. (Sprout Social (Q4 2025 results), Feb 2026)
  • Excluding small and medium-sized business customers, the Sprout Social rate was 102% in 2025, against 108% in 2024. (Sprout Social (Q4 2025 results), Feb 2026)
  • Twilio reported a dollar-based net expansion rate of 108% for the full year 2025, against 104% in 2024. (Twilio (Q4 2025 results), Feb 2026)
  • In the fourth quarter of 2025 the Twilio rate was 109%, against 106% in the fourth quarter of 2024. (Twilio (Q4 2025 results), Feb 2026)
  • UiPath reported a dollar-based net retention rate of 109% for the quarter to 31 July 2026. (UiPath (Q2 fiscal 2027 results), Sep 2026)
  • Weave Communications reported a dollar-based gross retention rate of 89% and a net retention rate of 93% at 31 December 2025. (Weave (Q4 2025 results), Feb 2026)
  • Zoom reported a trailing twelve-month net dollar expansion rate of 99% for its Enterprise customers in the quarter to 31 July 2026, against 98% a year earlier. (Zoom (Q2 fiscal 2027 results), Aug 2026)

Net revenue retention is defined by each company in its own filing, and the time windows, customer groups and revenue measures differ. Gross retention leaves out expansion, so it cannot exceed 100%. The companies are listed alphabetically after Snowflake and are not ranked or averaged.

How many customers belong to loyalty and membership programmes?

Tens of millions at single retailers, on the companies' own counts, and members account for most of the sales at some of them. Each programme counts members in its own way.

76.2m 2024 81.0m 2025 84.1m 2026
View data table
Costco paid members worldwide at fiscal year end, in millions, from the membership table in the company's fiscal 2026 annual filing.
PeriodValue
202476.2m
202581.0m
202684.1m
Costco paid members worldwide at fiscal year end, in millions, from the membership table in the company's fiscal 2026 annual filing. (Costco (fiscal 2026 Form 10-K), Oct 2026)
  • Costco had 84.1 million paid members and 150.4 million cardholders worldwide at the end of fiscal 2026. (Costco (fiscal 2026 Form 10-K), Oct 2026)
  • Executive members generated approximately 75.2% of Costco's worldwide net sales in fiscal 2026. (Costco (fiscal 2026 Form 10-K), Oct 2026)
  • At the end of fiscal 2024 Costco's member renewal rate was 92.9% in the US and Canada and 90.5% worldwide. (Costco (fiscal 2024 Form 10-K), Oct 2024)
  • At the end of fiscal 2020 Costco's member renewal rate was 91% in the US and Canada and 88% worldwide. (Costco (fiscal 2020 Form 10-K), Oct 2020)
  • Ulta Beauty Rewards had more than 46 million members at the end of fiscal 2025, and approximately 95% of Ulta's total sales came from members. (Ulta Beauty (fiscal 2025 Form 10-K), Mar 2026)
  • McDonald's systemwide sales to loyalty members were more than $11 billion in the third quarter of 2026 alone. (McDonald's (Q3 2026 results), Nov 2026)
  • Chewy's Autoship customer sales, from its repeat-delivery subscription, were $10.5 billion in fiscal 2025, or 83.3% of net sales. (Chewy (fiscal 2025 results), Mar 2026)
  • Chewy reported 21.3 million active customers and net sales per active customer of $591 for fiscal 2025. (Chewy (fiscal 2025 results), Mar 2026)
  • Delta Air Lines' total loyalty revenue grew 18% in the September 2026 quarter, and its remuneration from American Express grew 15%. (Delta Air Lines (September quarter 2026 results), Oct 2026)
  • Delta expects its American Express remuneration to exceed $9 billion for the full year 2026, a company expectation stated in October 2026. (Delta Air Lines (September quarter 2026 results), Oct 2026)
  • Enrolments in American Airlines' AAdvantage programme rose 27.4% year on year in the third quarter of 2026, and spending on its co-branded cards with Citi grew nearly 10%. (American Airlines (Q3 2026 results), Oct 2026)

Member counts and sales shares are each company's own figures. Costco's renewal rate is a trailing calculation covering renewals seven to eighteen months before the reporting date. McDonald's counts members active in the last 90 days of the quarter, and Ulta reports total members. The programmes are not ranked against each other.

How much do US credit card rewards pay out?

The CFPB figures come from its analysis of data supplied by large US card issuers.

Why do customers switch brands?

A lower price is the reason most often given in Capgemini's consumer survey, but service failures push people out too, and many leave without saying why. The surveys below ask different questions of different samples, so each is reported separately.

  1. 1 Competitor offers lower regular price 74%
  2. 2 Brand reduced pack size or quality without clear notice 71%
  3. 3 Competitor offers better-value loyalty programme 65%
  4. 4 Competitor offers a better promotion 62%
  5. 5 Bundle value (product plus service or warranty) 61%
  6. 6 Competitor offers better experience 61%
  7. 7 Consumer reviews 61%
Share of consumers who find each reason compelling for switching brands or retail stores, in a Capgemini survey of 12,000 consumers in 12 countries in October 2025. (Capgemini Research Institute (consumer trends 2026), Jan 2026)
  • A competitor offering a lower regular price is a compelling reason to switch brands or stores for 74% of consumers in a 12-country survey, and a brand reducing pack size or quality without clear notice for 71% (October 2025). (Capgemini Research Institute (consumer trends 2026), Jan 2026)
  • 65% of consumers in the same survey say a competitor offering a better-value loyalty programme is a compelling reason to switch. (Capgemini Research Institute (consumer trends 2026), Jan 2026)
  • For 21% of consumers in the Genesys survey, one bad experience is enough to switch, a 24% increase on the prior year. (Genesys (State of Customer Experience, 5th edition), Jul 2026)
  • 85% of consumers in the Genesys survey say poor service has caused them to spend less or stop doing business with a brand altogether. (Genesys (State of Customer Experience, 5th edition), Jul 2026)
  • 86% of consumers in a 22-country survey say responsiveness and accurate resolution highly influence their purchase decisions (June 2025). (Zendesk (CX Trends 2026), Nov 2025)
  • 85% of customer experience leaders in the same research say customers will drop brands that cannot resolve issues on first contact. This is a survey of business respondents, not consumers. (Zendesk (CX Trends 2026), Nov 2025)
  • 74% of consumers are frustrated when they have to repeat information to different agents, and 81% want agents to continue the conversation without backtracking. (Zendesk (CX Trends 2026), Nov 2025)
  • 73% of customers in an 18-market survey say companies treat them as an individual and not as a number, up from 39% in 2023. (Salesforce (State of the AI Connected Customer), Oct 2024)

Capgemini surveyed 12,000 consumers in 12 countries. Genesys surveyed 5,811 consumers in more than 20 countries, Zendesk 6,182 consumers and 5,115 business respondents in 22 countries, and Salesforce 16,585 consumers and business buyers in 18 markets. These are the companies' own surveys with different questions and samples, so the percentages are not comparable with each other.

How much do customers cut spending after a bad experience?

The sales-at-risk figures are a modelled estimate by Qualtrics: survey percentages multiplied by household consumption. They are not measured losses.

Do customers tell companies before they leave?

How does service affect Net Promoter Score?

  • 67% of digital transactions are completed without issues and 33% have notable issues, in a December 2024 survey of US customers. (Bain & Company (digital experience survey), Jan 2026)
  • US customers with issue-free digital transactions gave vendors a Net Promoter Score of 57. For transactions with notable issues the score dropped 88 points to -31. (Bain & Company (digital experience survey), Jan 2026)
  • Among US industries in Q3 2025, big box retail has the highest average Net Promoter Score at 37 and utilities the lowest at 13, on the Qualtrics ranking. TV and internet service providers score 16. (Qualtrics XM Institute (NPS by industry 2025), Apr 2026)

Bain and Qualtrics each calculate Net Promoter Score from their own surveys of US consumers, and the two sets of scores are not comparable.

What does poor service cost UK customers?

Which? surveyed 4,101 UK adults in May 2024 and 3,372 UK adults who had contacted customer service in June 2023. The £298 million and 27.3 million hours are its own estimates. The Salesforce figures are the UK part of an 18-market survey.

How satisfied are UK customers?

More satisfied than two years ago. The UK Customer Satisfaction Index has climbed back close to its July 2022 highpoint, and banks and building societies score above utilities and public services.

  1. 1 Banks and building societies 82.0
  2. 2 Tourism 80.9
  3. 3 Retail (food) 80.6
  4. 4 Leisure and hospitality 80.6
  5. 5 Telecommunications and media 76.6
  6. 6 Services 76.6
  7. 7 Transport 75.7
  8. 8 Public services (national) 74.9
  9. 9 Public services (local) 74.0
  10. 10 Utilities 73.7
UKCSI scores out of 100 by sector, July 2026, for the 10 sectors listed on the Institute of Customer Service page. (Institute of Customer Service (UKCSI July 2026), Jul 2026)

The UKCSI is published twice a year by the Institute of Customer Service from an online survey of UK consumers; the January 2026 edition covered more than 15,000 consumers and 272 organisations. It is a UK index on its own scale and is not comparable with the Forrester index in the next section. The £7.3 billion is the Institute's estimate, not measured spending.

Is customer experience getting better or worse in North America?

Better in 2026, after four years of decline on Forrester's index. More US and Canadian brands improved than declined in 2026, though US federal agencies moved the other way.

  • Improved 26%
  • Unchanged 68%
  • Declined 7%
How CX Index scores changed for US and Canadian brands in 2026: the share of brands with a statistically significant gain, a decline, or no change. Shares are rounded. (Forrester (CX Index 2026), Jun 2026)
  • 68% of North American brands in Forrester's 2026 CX Index were unchanged, and three US industries declined: airlines, credit card issuers and the US federal government. (Forrester (CX Index 2026), Jun 2026)
  • In 2025, US and Canadian consumer perceptions of customer experience quality dropped for a fourth consecutive year, to what Forrester called a new all-time low. (Forrester (CX Index 2025), Jun 2025)
  • In the US, 25% of the brands Forrester evaluated in both 2024 and 2025 had statistically significant losses in 2025, and 7% improved. (Forrester (CX Index 2025), Jun 2025)
  • Of 375 brands evaluated in both 2025 and 2026 across Asia Pacific, Europe and North America, 41% improved their Total Experience Score and 3% declined. (Forrester (Total Experience Score 2026), Jun 2026)
  • Globally, 20% of CX Index scores improved and 8% declined among brands evaluated in both 2025 and 2026. (Forrester (Total Experience Score 2026), Jun 2026)
  • Forrester's US Federal CX Index average fell from 61.9 in 2025 to 59.4 in 2026, which Forrester describes as the lowest US federal average since 2018. (Forrester (US Federal CX Index 2026), Jul 2026)
  • Seven US federal agencies had statistically significant declines in Forrester's 2026 index and none improved. (Forrester (US Federal CX Index 2026), Jul 2026)

Forrester's indexes come from its own surveys, with more than 224,000 customers and 462 brands in 2026. They measure brand-level change on Forrester's scale and are not comparable with the UKCSI.

How many customer complaints do regulators and ombudsmen receive?

Millions a year in US consumer finance alone, most of them about credit reporting. These are counts of complaints sent to one body, not a measure of service quality across a whole market.

c. 352,400 2019 994,000 2021 1,657,600 2023 3,187,900 2024 6,635,400 2025
View data table
Consumer complaints received by the US Consumer Financial Protection Bureau in the years its 2025 annual report gives: 2019, 2021, 2023, 2024 and 2025. The 2019 figure is approximate, and the years are not evenly spaced.
PeriodValue
2019c. 352,400
2021994,000
20231,657,600
20243,187,900
20256,635,400
Consumer complaints received by the US Consumer Financial Protection Bureau in the years its 2025 annual report gives: 2019, 2021, 2023, 2024 and 2025. The 2019 figure is approximate, and the years are not evenly spaced. (CFPB (Consumer Response annual report 2025), Mar 2026)

Each count covers complaints received by one body. The CFPB handles complaints about consumer financial products in the US, and a complaint count is not a count of consumers. The UK ombudsmen count cases escalated to them. None of the counts measures service quality across a whole market, and they are not compared with each other.

How many complaints reach the UK Financial Ombudsman?

The ombudsman counts complaints escalated to it after a firm's own complaints process, not every complaint made to a firm. Quarterly figures are rounded to the nearest 100.

How many energy complaints reach the UK ombudsman?

These are cases escalated to the ombudsman after the energy supplier's own process.

How much do US cardholders dispute?

  • US cardholders disputed $9.8 billion in credit card charges in 2024, resulting in $5.9 billion in chargebacks. (CFPB (credit card market report 2025), Dec 2025)
  • On US general purpose cards in 2024, a cancelled recurring transaction such as a subscription or membership fee was the most common reason for a dispute, at 40% of disputes, followed by services not received at 21% and cancelled or credit not issued at 18%. (CFPB (credit card market report 2025), Dec 2025)

How many people work in customer service in the US?

Millions, and the official projection is for fewer by 2035. Most of the openings the Bureau of Labor Statistics expects come from replacing workers who leave.

  1. 1 Retail trade 15%
  2. 2 Insurance carriers and related activities 12%
  3. 3 Business support services 8%
  4. 4 Professional, scientific and technical services 7%
  5. 5 Healthcare and social assistance 6%
Share of US customer service representative jobs in 2025 in the five industries the Bureau of Labor Statistics lists as the largest employers. (BLS (Occupational Outlook Handbook), Oct 2026)
  • The median pay of US customer service representatives was $44,770 a year in 2025, or $21.53 an hour. (BLS (Occupational Outlook Handbook), Oct 2026)
  • US employment of customer service representatives is projected to decline 5% from 2025 to 2035, a fall of 141,800 jobs, on the BLS projection. (BLS (Occupational Outlook Handbook), Oct 2026)
  • About 289,500 openings for US customer service representatives are projected each year, on average, over the decade to 2035. (BLS (Occupational Outlook Handbook), Oct 2026)
  • In the BLS table of occupations with the largest projected job declines for 2025-2035, customer service representatives are projected at 2,524,100 jobs in 2035, a fall of 5.3%. Cashiers (down 200,600) and general office clerks (down 156,200) are listed above them. (BLS (Employment Projections, Table 1.6), Aug 2026)
  • Median hourly pay for US customer service representatives in May 2025 was $22.47 at insurance carriers, $21.77 in professional, scientific and technical services, $21.49 in healthcare and social assistance, $17.96 in retail trade and $17.68 in business support services. (BLS (Occupational Outlook Handbook), Oct 2026)
  • The lowest-paid 10% of US customer service representatives earned less than $15.27 an hour in May 2025, and the highest-paid 10% earned more than $30.57. (BLS (Occupational Outlook Handbook), Oct 2026)
  • 12% of service employees left their company in the past year, in a 2025 survey of customer service professionals in 40 countries. (Salesforce (State of Service, 7th edition), 2025)
  • Service representatives in the same survey spend 46% of their time with customers, and 82% of service professionals say customer expectations are higher than they used to be. (Salesforce (State of Service, 7th edition), 2025)

The BLS figures cover all customer service representatives in the US, including those who work outside call centres, and the 2035 figures are projections. Business support services includes telephone call centres. The Salesforce figures come from its own survey of 6,500 service professionals.

How long do callers wait for HMRC in the UK?

These are HMRC's own operational figures for the UK tax authority's helplines. Its satisfaction measure comes from its own customer survey and is not comparable with the UKCSI.

How many calls does the IRS answer in the US?

The figures come from the National Taxpayer Advocate, an independent office within the IRS, and count calls answered by telephone assistors as a share of all calls received.

How is AI changing customer service?

Companies are investing faster than customers are warming to it. Business surveys report rising AI budgets and a growing share of cases handled by AI, while consumer surveys report worries about data, jobs and reaching a human.

  1. 1 Important to know when talking to an AI agent 72%
  2. 2 Believe companies are reckless with customer data 64%
  3. 3 Say AI makes it more important for companies to be trustworthy 61%
  4. 4 Comfortable with an AI agent making financial decisions 17%
What customers in 18 markets say about AI and trust, in a Salesforce survey of consumers and business buyers in July-August 2024. (Salesforce (State of the AI Connected Customer), Oct 2024)

Consumer surveys and business surveys ask different people different questions and are listed separately. Qualtrics surveyed more than 20,000 consumers in 14 countries, Zendesk 6,182 consumers in 22 countries, Salesforce 16,585 consumers and business buyers in 18 markets and Genesys 5,811 consumers in more than 20 countries.

What do customer service teams say about AI?

  • Service professionals expect AI to resolve 50% of service cases by 2027, up from 30% in 2025. This is an expectation given by respondents to a 2025 survey. (Salesforce (State of Service, 7th edition), 2025)
  • 39% of service professionals in the same Salesforce survey say they use agentic AI, and 79% of service leaders believe investing in AI agents is fundamental to meeting current business demands. (Salesforce (State of Service, 7th edition), 2025)
  • 80% of customer experience leaders say AI transparency will be required for any customer-facing AI within two years. (Zendesk (CX Trends 2026), Nov 2025)
  • 82% of senior customer service leaders say their teams invested in AI in the past year, and 87% plan to invest in 2026, in a survey of 2,470 support professionals. (Intercom (Customer Service Transformation Report 2026), Feb 2026)
  • 77% of customer service teams in the same Intercom survey say AI is meeting or exceeding their expectations, and 10% of organisations say they have reached a mature level of deployment. (Intercom (Customer Service Transformation Report 2026), Feb 2026)
  • 45% of customer service teams have updated job descriptions to include AI-related responsibilities, 40% say their human agents now focus more on training AI systems, and 27% say human agents mainly handle the most complex escalations. (Intercom (AI and support careers), Feb 2026)

These are surveys of people who work in customer service, run by companies that sell customer service software. They are kept apart from the consumer surveys above.

What have companies reported from their own AI assistants?

  • Klarna said its AI assistant handled 2.3 million conversations in its first month, two-thirds of its customer service chats, doing the equivalent work of 700 full-time agents. (Klarna (AI assistant release), Feb 2024)
  • Klarna said customers resolved their errands in under 2 minutes with the assistant, against 11 minutes before, and that repeat inquiries fell 25%. (Klarna (AI assistant release), Feb 2024)
  • Bank of America had 24.2 million active users of its Erica virtual assistant in the third quarter of 2026, up 18%. (Bank of America (Q3 2026 results), Oct 2026)
  • Bank of America had 50 million active digital banking users in the third quarter of 2026, and 71% of its total sales were digitally enabled. (Bank of America (Q3 2026 results), Oct 2026)

These are the companies' own claims and operating figures. The Klarna figures cover the first month of its assistant in early 2024 and were not independently audited.

Glossary

Churn rate
The share of customers or subscriptions lost in a period. Companies define it in their own way, and most carriers and subscription services report it as an average monthly rate.
Postpaid
A mobile plan billed after use, usually on a monthly contract. Prepaid plans are paid for in advance and are reported separately.
Premium SVOD
Subscription video on demand: the paid streaming video services tracked by Antenna in the US.
Serial churner
Antenna's term for a person with three or more premium streaming cancellations in the last two years.
NRR
Net revenue retention, also called dollar-based net retention or net expansion rate: revenue from a group of existing customers now, divided by revenue from the same group a year earlier. Each company sets its own formula.
GRR
Gross revenue retention: the same comparison with expansion left out, so it cannot exceed 100%.
ARR
Annual recurring revenue, the yearly value of subscription contracts.
Renewal rate
The share of members or contracts that renew. Costco calculates its rate on renewals seven to eighteen months before the reporting date.
NPS
Net Promoter Score: the percentage of customers who would recommend a company minus the percentage who would not, on a scale from -100 to 100.
UKCSI
UK Customer Satisfaction Index, published twice a year by the Institute of Customer Service and scored out of 100.
CX
Customer experience. Forrester's CX Index is its own survey-based measure of customer experience quality by brand.
CFPB
Consumer Financial Protection Bureau, the US regulator that receives complaints about consumer financial products.
BLS
Bureau of Labor Statistics, the US agency that publishes employment, pay and job projections by occupation.
HMRC
HM Revenue and Customs, the UK tax authority.

Methodology and sources

Every figure on this page comes from a primary document: a company filing or results release, a regulator or ombudsman report, a government statistics page, or the survey publisher's own report. Each figure links to that document and carries its date. Companies' operating figures use each company's own definition, so carrier churn, streaming churn, survey-based churn, renewal rates and net revenue retention are shown separately and are never ranked, averaged or charted together. Surveys run by software and consulting firms are labelled as their own surveys, with the sample and countries where the publisher gives them, and consumer surveys are kept apart from surveys of business respondents. Estimates and projections are labelled with who made them. Complaint counts are counts of complaints received by one body. The page does not state a single figure for the profit effect of retention or for the cost of acquiring a customer, because the sources do not agree on one. Figures the producer does not print, such as differences worked out between two years or two companies, are left out, as are figures that could not be opened on the producer's own page. The page is reviewed monthly and updated when a producer publishes a new release.