Blog · How-to guides · Telecommunications
Telecom marketing analytics compares acquisition spending with activations, early cancellations and realized contribution. Use equally mature customer cohorts to see which channels produce durable results.
Telecom marketing analytics compares campaign spending with activated services, early cancellations and the contribution those services produce. The useful comparison follows customers acquired in the same period through the same length of time. A low activation cost means little if the services disappear before the acquisition spending is recovered.
A US connectivity provider might sell mobile lines, home internet subscriptions and business accounts. Those units cannot share an acquisition denominator without explanation. Decide whether the report counts accounts or services, then keep that choice through spending, activations, cancellations and billing.
For this example, one activation means one new residential internet service that became billable. A replacement router is outside the measure. So are an existing customer's plan change and an unfulfilled order. These events can matter operationally, but they do not represent the same acquisition outcome.
Keep the difference between a campaign cohort and the installed customer base visible. Verizon's annual report defines its wireless churn measure using disconnects and an average connection base, illustrating why a published churn ratio needs its specific denominator (Verizon annual report). The early cancellation rate below instead follows only the services acquired in a stated cohort. It is not a substitute for a companywide monthly churn measure.
Covirage is not affiliated with Verizon.
The broader account questions belong with the B2B telecom sales director's analysis. This report answers the narrower marketing question: what happened to the services acquired through each channel?
Use the marketing spending export, the activation register and a billing export that includes credits and cancellations. The analysis needs stable service identifiers to connect the records without counting a service more than once.
| Export | Fields to retain | Check before comparing |
|---|---|---|
| Marketing spending | Campaign, channel, service period, spending | Spending reconciles to the approved campaign cost report |
| Activation register | Service ID, acquisition channel, activation date, offer | Only qualifying new billable services remain |
| Cancellation register | Service ID, effective cancellation date | Each cancellation belongs to an activation in the cohort |
| Billing and direct costs | Service ID, service period, net revenue, direct cost | Credits and costs use the same observation window |
Give an unresolved channel its own category. Do not silently move it into the largest campaign. Where several touches compete for credit, use the same documented rule across cohorts; marketing attribution models covers that separate choice.
An activation cohort and a campaign spending period may differ. A campaign can generate orders that activate later. Document how its cost is assigned, and keep any unallocated spending visible rather than making the service cost appear lower by dropping it.
An illustrative US provider follows its January 2026 acquisition cohort for each service's first 90 days. All services have completed that window before the report is prepared. Net service revenue includes credits; direct service costs use the same policy for each channel. Acquisition spending includes media and channel commissions, with neither cost included again in direct service costs.
| Channel | Acquisition spend (USD) | Activated services | Canceled within 90 days | Net service revenue (USD) | Direct service costs (USD) |
|---|---|---|---|---|---|
| Paid search | 48,000 | 600 | 90 | 90,000 | 42,000 |
| Retail partner | 36,000 | 480 | 24 | 76,800 | 33,600 |
| Direct mail | 24,000 | 300 | 30 | 46,500 | 21,000 |
| Total | 108,000 | 1,380 | 144 | 213,300 | 96,600 |
The calculation follows these definitions:
Cost per activation = Acquisition spend / Activated services
Early cancellation rate = Services canceled within the window / Activated services
Cost per retained service = Acquisition spend / (Activated services − Canceled services)
Contribution after acquisition spend = Net service revenue − Direct service costs − Acquisition spend
Here, a retained service has not canceled by the end of its observation window. A service that cancels and returns requires an explicit reinstatement rule before using this simplified identity. Amounts per service below are rounded to cents; cancellation rates are rounded to one decimal place.
| Channel | Cost per activation (USD) | Early cancellation rate | Retained services | Cost per retained service (USD) | Contribution after acquisition spend (USD) |
|---|---|---|---|---|---|
| Paid search | 80.00 | 15.0% | 510 | 94.12 | 0 |
| Retail partner | 75.00 | 5.0% | 456 | 78.95 | 7,200 |
| Direct mail | 80.00 | 10.0% | 270 | 88.89 | 1,500 |
| Total | 78.26 | 10.4% | 1,236 | 87.38 | 8,700 |
Paid search costs 48,000 / 600 = $80.00 per activation. After 90 cancellations, that same spending supports 510 retained services, so 48,000 / 510 = $94.12 per retained service. Retail partner acquisition is cheaper on both measures in this cohort, and its contribution after acquisition spending is 76,800 − 33,600 − 36,000 = $7,200.
For a fuller acquisition cost policy, including the treatment of shared sales costs, see customer acquisition cost. The table uses a campaign cost boundary rather than claiming to include every company acquisition expense.
The service identity must hold: 1,380 activations − 144 cancellations = 1,236 retained services. The financial identity must also hold: $213,300 net revenue − $96,600 direct service costs − $108,000 acquisition spending = $8,700. Both are exact before rounding.
Calculate the overall cost per retained service from the summed amounts: 108,000 / 1,236 = $87.38. Averaging channel costs without their service counts would answer a different question. Tie the revenue and costs back to the included billing rows and explain any services without a complete financial match.
This is realized contribution during the observation window, not lifetime value. A service may remain active but unprofitable, and longer-lived services may recover acquisition costs after this window. Do not project that recovery from retention alone.
Marketing owns campaign definitions, channel assignments and proposed spending changes. Billing operations owns activation and cancellation status. Finance owns the revenue and cost boundary and signs off on the reconciliation. The commercial lead decides whether the evidence warrants a larger controlled test or a change in acquisition mix.
Retail partner looks promising here, but its customers might buy different offers or live in different areas. Compare those segments before treating the channel difference as an acquisition effect. If the question is whether extra spending produces extra services, design a suitable controlled experiment; NIST's explanation of randomized designs describes random assignment to experimental units. Attributed outcomes alone do not answer that causal question.
A telecom analytics dashboard can provide the recurring overview. Keep this cohort comparison behind it so a changing customer mix does not masquerade as campaign improvement.
Bring a sample of your own campaign spending, activation, cancellation and billing exports to Covirage. Covirage maps the columns, and its tools compute and check the cohort comparisons; the external AI model explains the results with every figure traced to its rows. Create a free Covirage account, which takes about a minute, then set up your business and bring your file.
See Covirage for telecoms for the supporting industry context.
Either can work, but the unit must stay consistent. A household can add several lines under one account. State whether spending is divided by new accounts, activated lines or retained lines, and keep upgrades and replacements outside a new-customer measure.
Choose a window that covers the business decision, including early cancellations and promotional credits. Compare cohorts only after each has completed that window. Longer retention and payback questions need additional follow-up rather than an assumption that early results continue unchanged.
A low cost per activation can coexist with frequent cancellations, large incentives or weak contribution. Compare retained services and realized contribution using the same cost policy. Differences can also reflect product, geography and customer mix rather than the channel itself.