Find the friction between telecom touchpoints
Customers experience a telecom service as a sequence: choosing a plan, activating it, paying, getting help and deciding whether to stay. Research should show where that sequence becomes difficult, rather than collapse every moment into one overall rating.
Map the moments before writing the survey
Create a journey map with a row for each meaningful event. Record what the customer wanted to achieve, the channel available, the information required and the expected outcome. Distinguish a mobile plan from home connectivity, prepaid from postpaid and a first purchase from a renewal.
A customer who has never used a support channel should not rate its performance. Build an eligibility question for each event and route the interview accordingly. Use a reference period that people can realistically remember; agree it with the business before fieldwork.
Onboarding
Explore plan understanding, document steps, activation and the first successful use. Ask where customers needed help and whether they understood what would happen next.
Billing
Separate bill readability, unexpected charges and payment completion. A payment problem and a pricing objection require different operational responses.
Support
Capture the reason for contact, channels attempted and whether the issue was resolved. Ask about repeat contacts without assuming the customer used only one channel.
Renewal
Explore whether the existing plan still meets needs, what alternatives were considered and which changes might affect the next decision.
Combine depth with measurement
Interviews can help the team discover how customers describe a problem. A subsequent survey can measure how often defined experiences are reported within the recruited sample. The two approaches answer related but different questions.
Use a small qualitative discovery phase to refine the journey language, then pilot the structured questionnaire. Do not label a survey question “network quality” when customers might mean coverage, speed, call stability or a device problem. Ask about the relevant experience separately.
The most useful journey research identifies the moment, the customer need and the operational owner of the problem.
Plan the audience and collection channels
Create a sample matrix covering the customer groups that matter to the decision: product, tenure, geography and recent contact experience. Avoid an overcomplicated matrix that leaves many tiny cells. If the main priority is support, allocate enough interviews to recent support users while explaining how that allocation affects overall reporting.
Choose collection methods around reach and the research task. Telephone interviewing can clarify a complex sequence; online surveys can suit customers comfortable with digital channels. Mixing methods requires attention to wording, presentation and the interpretation of differences. See our comparison of CAPI, CATI and CAWI research methods.
Deliver a journey improvement map
Report each touchpoint with its respondent base, issue frequency, customer explanation and suggested next investigation. Keep observed evidence separate from proposed fixes. If customers report an activation delay, research can identify the experience; it cannot establish the underlying technical cause without operational verification.
- Show the route customers actually followed, including repeated contacts.
- Separate results for people who experienced the touchpoint from the total sample.
- Prioritise problems by severity, reach and practical ability to act.
- Repeat core measures after a service change using a comparable design.
A worked example: tracing an activation problem
Consider a hypothetical customer who buys a new connectivity package online, receives a confirmation message and then calls support because activation is unclear. An overall satisfaction question may capture frustration, but it cannot explain whether the problem came from missing instructions, a delay or a failed activation step. The study needs a sequence of questions that reconstructs the journey.
Start by establishing the product and purchase channel. Ask whether the customer understood the expected activation timing, which instructions they received and what happened when they tried to use the service. Only ask about a support contact if the customer made one. If they used several channels, record the sequence rather than force them to choose a single “main” channel.
The resulting journey can distinguish a completed activation with confusing communication from an activation that remained incomplete. Those findings should lead to different operational investigations. The first may require reviewing the explanation or message timing; the second may require checking the fulfilment process. Research identifies the reported experience and does not independently diagnose a technical failure.
This scenario is an illustrative research design, not a Census project outcome. It shows why the interview should describe what happened before asking respondents to evaluate it. A useful questionnaire allows for a successful journey too, rather than assume every new customer encountered a problem.
Questions that separate the telecom experience
For onboarding, ask “Which of the following steps did you complete?” before asking how clear each step was. The options should match the actual customer process. If an identity check, delivery or installation step is relevant only to one product, keep it in that product route. Do not use the same generic list for every service.
For billing, ask about the latest bill or a defined billing period. Separate understanding the bill from agreeing with the amount. “Was any charge different from what you expected?” can provide a diagnostic route, followed by a question about the type of unexpected charge. Avoid implying that an unexpected charge was incorrect; the customer may have misunderstood an agreed condition.
For support, ask whether the customer achieved the purpose of the contact, how many contacts were needed and whether the next steps were clear. A polite interaction can still leave a problem unresolved. Keep courtesy, clarity and resolution as separate measures if each has a different operational owner.
For digital self-service, identify the task: paying a bill, changing a plan, checking usage or finding help. Ask whether the task was completed and which step caused difficulty. General questions about “the app” can mix unrelated experiences. Use the respondent’s reported task to determine the follow-up questions.
Connect journey evidence without exposing unnecessary data
Agree in advance whether survey findings will be compared with operational information. The analysis might use broad groups such as product type or contact channel, rather than require every customer’s complete account history. Define the information needed for the decision and the permitted handling process before collection.
If a respondent reports multiple contacts, preserve that sequence in the dataset. The analysis should be able to distinguish one unresolved issue from several unrelated issues. Stable question IDs and clear answer codes help analysts reconstruct the route without manually interpreting every interview.
When reporting a journey, show the basis for each stage. Some customers may never reach a later step, while others may skip a stage because it was not relevant. Missing and not-applicable responses should not be silently treated as a poor experience. Explain the rules used to construct each measure.
Prioritise changes and plan the next wave
Use a touchpoint matrix with customer goal, reported problem, affected audience, evidence and proposed next investigation. Include the number of respondents who experienced each touchpoint. Compare common issues with their impact on task completion so that the team does not prioritise only the easiest metric to display.
A practical workshop can ask the product, billing and support teams to review the same journey. Each team may see a different part of the problem. Assign responsibility for checking the underlying process, and state which proposed changes can be tested. Research findings should guide that investigation without pretending that a survey alone has verified its cause.
After a process change, preserve the main eligibility rules and journey measures. Add a separate module about the revised experience if needed. Document any change in recruitment, product coverage or survey mode. An apparent improvement should be assessed in the context of both the service change and the research design.
Frequently asked questions
Is an overall satisfaction score enough?
It is a useful summary but usually does not identify the step requiring action. Add relevant touchpoint and outcome measures, then connect them to the customer’s actual route. Keep the questionnaire focused on experiences that respondents can recall and that the business can investigate.
Should prepaid and postpaid customers be combined?
They can share a reporting framework, but their tasks and billing experiences may differ. Define which comparisons matter and plan enough responses for them. Report product bases clearly and avoid interpreting a combined result as though every customer followed the same journey.
Can research prove that a network caused a complaint?
A survey can describe a reported connectivity experience. Establishing a technical cause may require device, coverage or operational evidence beyond the interview. Keep reported experience, possible explanation and verified cause separate, especially when presenting recommendations to technical teams.
How can customers who stopped using the service be included?
Define a relevant former-customer group and an appropriate recruitment route. Explain its coverage limitations and distinguish reasons for leaving from current-user experience. Former users may provide useful evidence, but their answers should not be mixed into a current-customer measure without a clear reporting rule.
Continue your research planning
Turn the questionnaire into an action plan
Share your audience, markets and decisions with Census Market Research to discuss a suitable study design.
Discuss your research briefCover image: AI-generated editorial illustration. It does not depict an actual client project.