Benchmarking your metrics against telecom industry standards
Two credible sources can put the same operator's churn four points apart, and neither is lying. One counts monthly disconnections on consumer mobile accounts; the other annualises across every SIM on the network, machine-to-machine included. Before you tell a board you are above or below par, you have to know what the benchmark counted and rebuild your own number the same way.
This lesson does not re-derive the metrics. It assumes the loaded acquisition figure the channel-cost lesson builds and the line-type value models the lifetime-value lesson sets out, and asks two different questions: whose numbers are you comparing yourself to, and what has to be fixed on your side before the comparison means anything.
Where the comparators actually come from
Four kinds of source, each with a different bias baked in.
- Regulators. Ofcom collects UK operator data under statutory information requests, so its pricing, switching and complaints series are closer to a census than a sample, and the definitions stay stable year to year. That makes them excellent for reading direction and poor for reading your own level, because the categories are regulatory, not commercial.
- Industry bodies. GSMA Intelligence aggregates operator-reported data across markets and draws the distinction most internal decks get wrong: connections (SIMs) versus unique subscribers. Global unique mobile subscribers sit in the region of 5.8 billion while SIM connections run far higher, because of multi-SIM behaviour and IoT. Divide revenue by the wrong one and your ARPU is off by tens of percent.
- Measurement firms. Opensignal builds network experience metrics from readings taken on consenting users' handsets. That tells you why a market's churn moves, not what its churn is, and the panel skews toward people with recent smartphones and an app installed, which is not the low-ARPU prepaid base.
- Vendors. Segment, a customer data platformcustomer data platformSoftware that unifies customer data from every source into one persistent profile that marketing, sales and service teams can act on.View full definition → (it sells the event-collection tooling that produces the funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition → numbers in the first place), publishes conversion benchmarks drawn from its own installed base. Companies that have bought a CDP are already digital-first and already instrumented. A vendor funnel benchmark describes the well-instrumented end of the market, not the middle of it.
A fifth source is free: operator earnings disclosures and investor packs. They are audited, but the marketing-spend allocation behind them is not standardised, and a converged national operator's blended figures average postpaid, prepaid, fixed and wholesale into one line that matches nobody's business.
Acquisition costAcquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → benchmarks and the denominator problem
Rough directional ranges (2024 to 2025, US and Europe, blended across published operator disclosures and analyst estimates):
| Segment | Estimated CAC range |
|---|---|
| US postpaid mobile | $300 to $500 per subscriber |
| European postpaid mobile | €150 to €300 per subscriber |
| US fixed broadband | $400 to $600 per subscriber |
| Prepaid mobile (US and Europe) | $50 to $150 per subscriber |
The postpaid/prepaid gap is structural: credit checks, device financing and retail staff time all sit inside the postpaid number.
The denominator moves more than the numerator. Take $2 million of quarterly acquisition spend, 8,000 gross adds and 3,500 disconnections in the same quarter.
- Per gross add: $2,000,000 / 8,000 = $250
- Per net add: $2,000,000 / 4,500 = $444
Same spend, same quarter, and one version sits comfortably under the postpaid benchmark while the other sits at the top of it. Published sources split roughly down the middle on which they use, and many do not say. If the source is silent, assume gross adds and note the assumption in the footnote of your own deck.
Two more definitional forks worth checking before you compare: whether handset subsidy is expensed at sale or amortised across the contract (European operators reporting a separate subscriber acquisition cost often do the former), and whether upgrades to existing customers are counted as acquisitions. Including upgrades can inflate your add count by a third in a saturated market, which flatters CAC and quietly ruins the churn comparison later.
Ratio benchmarks: matching both halves
The 3:1 LTVLTVLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition →:CAC threshold is quoted across subscription businesses, telecom included. Strong-retention postpaid books often land between 4:1 and 10:1; competitive MVNO and prepaid segmentssegmentsDividing a market into distinct groups of customers who share similar needs, characteristics or behaviours, so each group can be served with a tailored approach.View full definition → frequently sit nearer 2:1 or 3:1 on thinner margins and faster turnover. Below 3:1 you are buying customers who do not stay or do not spend; above 10:1 you are usually underspending on growth rather than winning.
Normalising the ratio means checking both halves come from the same world. Most published LTV figures use undiscounted gross margingross marginGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → with no tenure cap, which flatters fixed broadband (long tenure, high switching friction) and is close to meaningless for a multi-year low-ARPU IoT fleet. Use the line-type models the lifetime-value lesson sets out, then compare like against like: a converged household value against a converged household benchmark, never against a blended mobile one.
ARPU anchors, for context: US postpaid mobile sits around $45 to $55 monthly, European ARPU often €15 to €25, a gap driven by price competition rather than by anyone doing marketing better. Comparing a German ARPU to a US one at spot exchange rates, on a metric measured over twelve months, adds a currency error on top of a market error.
Churn benchmarks: monthly, annual and per what
Directional monthly estimates:
- US postpaid mobile: 1% to 1.5% (strong performance)
- US prepaid mobile: 3% to 5%, structurally higher with no contractual lock-in
- European postpaid mobile: 1% to 2%, with wide country variation
- Fixed broadband, US and Europe: 1% to 2%, held down by installation friction
Annualising is where the comparisons break. Multiplying 1.5% by twelve gives 18%; compounding correctly gives 16.6%. A 1.4 point gap is enough to move you from second quartile to third against a published table, and it is entirely an arithmetic artefact.
Regulation moves the base rate. The European Electronic Communications Code capped consumer contract terms and simplified switching across member states. In the UK, mobile customers have been able to switch by text since 2017, and the One Touch Switch process for broadband and landline went live in September 2024, with Ofcom publishing the switching and complaints data that shows the effect. A churn benchmark drawn from a market with one-touch switching is not a target for a market without it.
Funnel conversion benchmarks
Rough conversion estimates for the path from awareness through coverage check, plan selection, identity or credit check and activation:
- Website visit to lead (plan configuration started): 2% to 5%
- Lead to activated subscriber, postpaid: 20% to 35%, with credit-check drop-off doing most of the damage
- Store walk-in to activation: 40% to 60%, on self-selected high-intent traffic
- eSIM or trial to paid activation: 15% to 30%
Check the denominator here too. A rate computed on sessions is roughly half the same rate computed on unique visitors, and analytics vendors default differently. If your digital funnel converts leads at 12% against a 20% to 35% range, the useful question is where between credit check, plan complexity and checkout you lose people who had already declared intent.
Knowledge check
1. A regional ISP reports a CAC well below the industry average for fixed broadband. What is the most important next step before celebrating this as a strength?
2. Why is comparing an MVNO's CAC directly to a major MNO's CAC misleading?
3. A telecom analyst wants to benchmark their company's funnel conversion rate against 'the industry average.' What should they do first to make the comparison meaningful?
4. Select ALL correct answers about factors that must be matched before comparing a company's metric to a published telecom benchmark.
Select all the correct answers.
5. Select ALL correct answers about why telecom benchmarking is more complex than simply comparing a single published industry figure.
Select all the correct answers.
Where benchmarks mislead you
Averages hide mix. An operator with several million IoT SIMs on its network reports a blended ARPU per connection that no consumer marketer should ever benchmark against. The same arithmetic runs the other way on churn: cellular IoT lines barely move, so they flatter a blended churn figure while the consumer base is bleeding.
Definitions drift between sources and between years. Operators change how they report subscriber counts after acquisitions, and regulators revise category boundaries. Reading a five-year trend from one source is safer than reading one year across three sources.
Benchmarks lag. Analyst estimates are typically 6 to 18 months old when you read them, and Ofcom's market reporting is annual. In a segment moving toward eSIM-first onboarding, last year's activation benchmark may already describe a different product.
The failure mode that costs most is quiet: picking the source that makes you look best, then tying a bonus to it. Set a target of beating a $300 postpaid CAC benchmark and the fastest legitimate route is to shift mix toward cheap gross adds, which lands the marketing bonus in Q1 and blows up the churn line in Q3. Pick the comparator before you pick the number you want it to say.
🎬 [VIDEO: "How Telecom Companies Make Money (and Lose Customers)" - youtube.com - search for GSMA or industry analyst explainer videos breaking down telecom unit economics and churn drivers, useful for a visual walkthrough of ARPU, churn and LTV mechanics]
A simple benchmarking checklist
- Confirm segment match (postpaid, prepaid, fixed, converged, MVNO) and line type.
- Confirm geography and regulatory regime, including switching rules and contract-length caps.
- Recompute your metric with the source's formula: gross versus net adds, connections versus unique subscribers, monthly versus compounded annual churn, sessions versus users.
- Check the publication date and treat anything older than 18 months as directional.
- Triangulate across at least two source types, ideally one regulatory or industry-body series and one operator disclosure. If they disagree by more than a few points, the definitions differ; find out how before you pick a side.
Key takeaways
- Know your source type before you trust the number: regulators give near-census stability, industry bodies give cross-market comparability, measurement firms explain experience, vendor benchmarks describe their own installed base.
- The denominator decides the verdict. $2m over 8,000 gross adds is $250; over 4,500 net adds it is $444, from identical spend.
- Compound churn properly: 1.5% monthly is 16.6% a year, not 18%, and the difference is enough to move your quartile.
- Match both halves of a ratio to the same line type and margin base, and never benchmark consumer mobile against blended figures that include IoT SIMs.
- Benchmarks are for direction (in range, wildly off, suspiciously good). Tying incentives to one you did not normalise buys you a good quarter and a bad year.