ISP Customer Retention Workflow: One Case Owner, One Recovery Plan

Why a closed ticket is not the same as a retained customer

Ofcom’s Comparing Customer Service report, reissued on 18 September 2026, describes UK customer experiences in 2024. It reports that 44% of fixed broadband complainants said their complaint was completely resolved at first contact, and 58% were satisfied with complaint handling. These are service-experience measures, not FTTH churn rates.

A separate Airties-commissioned Qualtrics survey of 2,130 US and UK Wi-Fi-connected respondents in Q3 2025 found that 85% of US switchers and 88% of UK switchers had complained before leaving. The implication is worth testing locally: the complaint journey can offer a recovery opportunity before cancellation.

The Ofcom reissue corrected certain fault data; it is not a new 2026 customer survey. The Airties findings are vendor-sponsored and self-reported. Neither source proves that this workflow reduces churn or establishes an Indonesian SLA. Source links and dates appear below.

Connect the workflow without creating another departmental queue

Create the retention case in the system your teams already use, and link existing fault tickets, billing records and commercial notes. Avoid asking customers to repeat information solely because responsibility changes.

1. Detect and register. Accept a repeat complaint, disputed bill, failed payment, unresolved activation or cancellation request as a trigger for investigation.
2. Assign and acknowledge. Name the case owner, confirm the problem and record the next customer update.
3. Diagnose and coordinate. Route the service, billing or plan-fit work while retaining one shared history.
4. Approve the commercial response. Correct errors first, then assess whether a discretionary offer is appropriate and affordable.
5. Verify and classify. Record resolved, monitoring, unreachable, offer declined or cancellation completed. Do not force every case into a “saved” status.
6. Review lasting outcomes. Follow repeat complaints, SLA performance and retention after the intervention.

Required case fields are customer/account reference, trigger, severity, consented contact channel, case owner, departmental task owners, original deadline, last update, next callback, approved offer cost and closure evidence. Restrict access and keep unnecessary personal details out of shared notes.

A RACI matrix with one accountable role for each activity

RACI means Responsible for doing the work, Accountable for the result, Consulted before decisions and Informed about progress. The case owner is a named customer-care or retention specialist, not a generic mailbox. A/R means the same role owns both delivery and accountability.

RACI legend: R = Responsible, performs the task. A = Accountable, owns the outcome and final decision. C = Consulted, provides input before action or a decision. I = Informed, receives updates on progress and outcomes. A/R means the same role is both accountable and responsible.

The Sales lead is accountable for the offer policy; documented agent and manager authority handles routine approvals. Department leads provide named task owners and completion evidence. The case owner obtains those inputs and keeps the customer informed.

Escalation SLAs: define response, updates and missed deadlines

The following service-level agreement targets are a proposed pilot policy, not industry benchmarks, regulatory requirements or promises of restoration. P1 runs continuously with an on-call rota. P2 and P3 use staffed hours of 08:00 to 20:00 daily in the customer’s local service time; a business day means one such staffed day.

Start the clock when the trigger is recorded. A transfer does not reset it. Capture an unavailable customer’s preferred callback separately; do not silently stop internal escalation. For major incidents, link each case to the incident and use the incident’s verified update, avoiding duplicate calls or contradictory repair estimates.

Escalation is complete only when a receiving owner accepts the task and records the next action. If they have not accepted within 30 elapsed minutes for P1, or 2 staffed hours for P2/P3, escalate to the on-call senior manager or retention lead respectively. Record the breach even if the customer later stays.

Only Operations confirms a restoration estimate. If the cause is unknown, promise an update time and a diagnostic action instead of inventing a repair deadline.

For the CMO: align the customer message with the recovery stage

Marketing owns approved message templates and lifecycle rules. The case owner delivers individual recovery communication using verified inputs. Suppress optional upgrade and acquisition-style promotions for customers with an open P1/P2 recovery case; keep essential billing, service and security notices active.

Retention content should give a useful next step. An appointment reminder, bill explanation or home Wi-Fi check can remove uncertainty. A generic message about how much the provider values customers cannot replace a missed callback.

For the Chief Sales Officer: approve offers without hiding the problem

Set approval limits on the total discretionary cost per account, not just the monthly discount. Include the entire offer duration, credits, subsidized equipment and incremental visits. The limits below are illustrative USD pilot settings, subject to your margin policy.

A USD 5.00 monthly discount for six months costs USD 30.00 and needs manager approval. Add USD 8.00 equipment subsidy and the total becomes USD 38.00, requiring director approval. Do not split offers across tickets to bypass limits.

Correcting a wrong invoice is not a discretionary retention incentive. Apply billing corrections and any required compensation under their own policies. A low estimated customer value never justifies leaving a service fault unresolved.

Use scripts that explain the next action

These are illustrative scripts, not messages already sent. Agents must select times they can meet within the applicable SLA, confirm the customer’s local time and preferred channel, and check the actual approved terms.

The assigned case owner retains callback accountability even if a sales agent presents the offer. If the customer cannot be reached, make up to two attempts at the agreed times over two staffed days, then record “unreachable” and provide a permitted follow-up route. Unreachable is not saved.

Weekly save-rate review: define the denominator first

Keep proactive recovery and cancellation saves in separate cohorts. A customer who never requested cancellation is not automatically a saved cancellation. Use a weekly meeting to identify broken handoffs and review cohorts whose follow-up windows have matured.

• 30-day verified save rate: cancellation-intent cases where the request was withdrawn and the customer remains active 30 days after intervention, divided by all eligible cancellation-intent cases entering the same cohort. Include failed contacts and declines; publish exclusion rules.
• Callback SLA attainment: callbacks completed by the original agreed deadline divided by all callbacks due in the period. Show missed attempts and customer-requested reschedules separately.
• Repeat-issue rate: verified recovery cases reopened for the same problem within 30 days divided by recovery cases with a complete 30-day observation window.
• Commercial value: retained recurring revenue and intervention cost by cohort, with incremental contribution assessed against a suitable comparison group where feasible.

For example, 60 verified saves from 100 eligible cancellation-intent cases means a 60% observed save rate. If 10 customers could not be reached, retain them in the denominator rather than reporting 60 out of 90. This is a descriptive result, not proof that the workflow caused all 60 to stay.

Review 90-day retention and offer expiry as those cohorts mature. Marketing examines message timing, Sales examines offer cost and conversion, Billing checks disputed charges, and Operations examines recurring faults. Assign every improvement a named owner and due date before the meeting closes.

Test the workflow on one customer journey

An illustrative FTTH customer reports repeated evening disconnections and threatens to leave. The case owner classifies P2, links the earlier ticket and arranges a callback within the proposed SLA. Marketing pauses optional upgrade messages. Operations investigates whether the access line, Wi-Fi or another factor explains the symptom; Billing checks a disputed charge in parallel.

The bill is corrected under billing policy. Operations supplies its findings and a recovery action. Only then does Sales evaluate whether a discretionary offer would help. The case owner confirms the original symptom after the action, records the customer’s decision and keeps unresolved tasks open.

Contrast that with a discount sent before diagnosis: the customer pays less but the problem may remain. For the broader retention context, read our perspective on price, reliability and Wi-Fi experience in FTTH retention.

Start with a four-week operational pilot in one service area. During week one, agree the RACI, clocks and case fields. During week two, rehearse handoffs and scripts. During weeks three and four, review live cases daily for overdue updates. Continue tracking 30-day and 90-day cohorts after the pilot; those outcomes cannot all be known within four weeks.

Frequently asked questions