What is a good churn rate for a telecom operator?

On postpaid mobile, monthly voluntary churn at or below 1.1 percent is median and 0.7 percent is top decile. Prepaid runs far higher, with a 4.4 percent median and 2.9 percent top decile, because there is no contract to hold the base.

Why is prepaid churn so much higher than postpaid?

Prepaid has no contract, so churn is dormancy rather than a stated cancellation. The measured rate also depends heavily on the inactivity threshold: operators defining churn at 90 days of inactivity report materially lower numbers than those using 30 days.

Do bundles reduce telecom churn?

Converged bundles show roughly a third lower churn than mobile-only at the median, but part of that gap is selection: households that bundle were already more stable. Claims about bundle uplift need a holdout rather than a before-and-after comparison.

Why is reported churn lower than actual churn?

Three accounting choices flatter public figures: measuring disconnections rather than cancellation requests, averaging over a growing subscriber base, and reporting gross churn while heavy acquisition masks net losses.

The Markin ROI Report for Enterprise Growth TeamsRead now

MARKIN

Book a demo

Field notes

ResearchAug 31, 20268 min read

Telecom churn rate benchmarks, 2026

Median monthly churn is 1.1 percent postpaid, 4.4 percent prepaid and 1.6 percent on converged bundles. What separates the top decile from the rest.

Romà Llambés

#Churn
#Benchmarks
#Telecom

The median monthly telecom churn rate in 2026 is 1.1 percent for postpaid mobile, 4.4 percent for prepaid mobile and 1.6 percent for converged broadband bundles, based on the subscription bases we analyse. The gap between median and top decile is roughly half a point on postpaid, worth millions of lines per year on a national base.

These figures come from the same benchmark programme behind our 2026 cross-industry ARPU benchmarks: customer-level, monthly, consistently defined across operators. Where we cite operator-reported numbers we say so, because carrier-reported churn and observed churn are rarely the same thing.

The 2026 benchmark set

Monthly voluntary churn, active subscriber basis, trailing twelve months:

1Postpaid mobile: 1.1 percent median, 0.7 percent top decile. Contract structure keeps the floor low. The spread between median and best is driven less by network quality than by save-desk discipline and contract renewal timing, both of which are decisioning problems.
2Prepaid mobile: 4.4 percent median, 2.9 percent top decile. No contract means churn is a behavioural signal, not an event: most prepaid churn is silent dormancy rather than a stated exit. Operators that define churn as 90 days of inactivity report lower numbers than those using 30, so comparisons across operators need the definition stated.
3Converged bundles: 1.6 percent median, 1.0 percent top decile. Adding broadband or TV to a mobile line cuts churn by roughly a third at the median. The caveat is selection: households that bundle were already more stable, so bundle uplift claims need a holdout, not a before-and-after.
4MVNOs: 2.6 percent median. Price-led positioning attracts price-led customers. The MVNOs beating the median do it with onboarding quality and first-90-day engagement, not with deeper discounts.

Why operator-reported churn is lower

Three accounting choices flatter public numbers. Churn measured on disconnections rather than cancellations ignores customers who asked to leave but were retained through a save. Averaging over a growing base shrinks the denominator. And counting gross churn while adding heavy acquisition masks net losses. When we rebuild churn from customer-level events, reported figures typically understate true voluntary churn by 15 to 30 percent. Treat any benchmark that does not state its definition as marketing, including ours: ours is monthly voluntary cancellations over average active subscribers.

What the top decile does differently

Comparing top-decile operators to the median across our base, three practices separate them, and none is a loyalty programme:

1They measure churn as a decision, not an event. Risk is scored continuously from engagement and billing signals, so intervention happens weeks before the cancellation call. The modelling practices are in our machine learning churn prediction article.
2They separate voluntary from involuntary. Payment failures go to dunning, not to the retention team, and the two are never pooled into one churn number.
3They prove saves with holdouts. Every retention offer runs against an untreated control. In our datasets, 20 to 40 percent of counted saves at median operators are customers who would have stayed anyway; top-decile operators know their real number because they measure it.

Using the benchmarks

A churn rate is only actionable against your own mix: prepaid share, bundle penetration and market maturity move the number more than any programme. Compare like for like, segment first, and read churn next to ARPU, because the operators growing fastest pair low churn with rising revenue per line, as we cover in how to increase ARPU in telecom.

Markin benchmarks churn and ARPU at customer level for operators and acts on the gaps with continuous decisioning. See the telecom industry view.

Frequently asked

Questions readers ask about this.

What is a good churn rate for a telecom operator?: On postpaid mobile, monthly voluntary churn at or below 1.1 percent is median and 0.7 percent is top decile. Prepaid runs far higher, with a 4.4 percent median and 2.9 percent top decile, because there is no contract to hold the base.
Why is prepaid churn so much higher than postpaid?: Prepaid has no contract, so churn is dormancy rather than a stated cancellation. The measured rate also depends heavily on the inactivity threshold: operators defining churn at 90 days of inactivity report materially lower numbers than those using 30 days.
Do bundles reduce telecom churn?: Converged bundles show roughly a third lower churn than mobile-only at the median, but part of that gap is selection: households that bundle were already more stable. Claims about bundle uplift need a holdout rather than a before-and-after comparison.
Why is reported churn lower than actual churn?: Three accounting choices flatter public figures: measuring disconnections rather than cancellation requests, averaging over a growing subscriber base, and reporting gross churn while heavy acquisition masks net losses.

Related resources

Solutions

Industries

Telco

See it in the product

This runs in Markin today.

The same loops this note describes run 24/7 against your customer base. Watch the workspace decide, experiment and execute 1:1.

Explore demo Or book a demo with the team

Vocabulary

The terms behind this page.

Definitions an answer engine can quote, each one a page of its own.

Full glossary

Compare

How this compares to what you already run.

Neutral reads on the categories buyers evaluate against a decision layer.

All comparisons

Keep reading.

All field notes

ResearchJul 27, 2026

Churn prevention economics in streaming and telecom

PlaybooksAug 07, 2026

Customer churn prediction: models, features and incrementality

GuidesAug 21, 2026

Voluntary vs involuntary churn: two problems, two fixes

The 2026 benchmark set

Monthly voluntary churn, active subscriber basis, trailing twelve months:

1Postpaid mobile: 1.1 percent median, 0.7 percent top decile. Contract structure keeps the floor low. The spread between median and best is driven less by network quality than by save-desk discipline and contract renewal timing, both of which are decisioning problems.
2Prepaid mobile: 4.4 percent median, 2.9 percent top decile. No contract means churn is a behavioural signal, not an event: most prepaid churn is silent dormancy rather than a stated exit. Operators that define churn as 90 days of inactivity report lower numbers than those using 30, so comparisons across operators need the definition stated.
3Converged bundles: 1.6 percent median, 1.0 percent top decile. Adding broadband or TV to a mobile line cuts churn by roughly a third at the median. The caveat is selection: households that bundle were already more stable, so bundle uplift claims need a holdout, not a before-and-after.
4MVNOs: 2.6 percent median. Price-led positioning attracts price-led customers. The MVNOs beating the median do it with onboarding quality and first-90-day engagement, not with deeper discounts.

Why operator-reported churn is lower

What the top decile does differently

Comparing top-decile operators to the median across our base, three practices separate them, and none is a loyalty programme:

1They measure churn as a decision, not an event. Risk is scored continuously from engagement and billing signals, so intervention happens weeks before the cancellation call. The modelling practices are in our machine learning churn prediction article.
2They separate voluntary from involuntary. Payment failures go to dunning, not to the retention team, and the two are never pooled into one churn number.
3They prove saves with holdouts. Every retention offer runs against an untreated control. In our datasets, 20 to 40 percent of counted saves at median operators are customers who would have stayed anyway; top-decile operators know their real number because they measure it.

Using the benchmarks

Markin benchmarks churn and ARPU at customer level for operators and acts on the gaps with continuous decisioning. See the telecom industry view.

Telecom churn rate benchmarks, 2026

The 2026 benchmark set

Why operator-reported churn is lower

What the top decile does differently

Using the benchmarks

Questions readers ask about this.

Bring this into your workspace.

This runs in Markin today.

The terms behind this page.

How this compares to what you already run.

Keep reading.

Churn prevention economics in streaming and telecom

Customer churn prediction: models, features and incrementality

Voluntary vs involuntary churn: two problems, two fixes

The 2026 benchmark set

Why operator-reported churn is lower

What the top decile does differently

Using the benchmarks