Meridian Water: abandonment in the smart-meter journey
Meridian Water serves 400,000 domestic customers. Its new self-install smart-meter app loses 62% of first-time users at the pairing step, and the contact centre has seen an 18% rise in call volume. A customer journey map exists, but it was produced by marketing, ends at app download, and has never been shown to the technology teams. Support tickets are closed on resolution with no follow-up, so the only customer sentiment data is a quarterly brand survey that does not mention meters. Field engineers report that customers often had already phoned the contact centre before their appointment, but no one can quantify it. The COO wants a fix within one quarter and has asked for "a better app".
1. How would you decide where to act first? Set out the evidence you would gather and what would change your mind.
Read the worked answer
The COO's framing - "a better app" - is a hypothesis, and the 62% abandonment is an outcome with several plausible causes: an unclear pairing instruction, a device or handset compatibility problem, an expectation gap at point of sale, or an unresolved earlier step. Diagnosing from the funnel alone would be guessing.
Gather, in this order:
- Step-level funnel data from the app, including error codes at pairing and device/OS segments, so you can see whether failure is concentrated.
- Structured contact reasons from the contact centre, coded to journey steps, to find where customers surface before they give up.
- Short interviews or a small closed-loop callback with 20–30 abandoners - effort and emotional data, not satisfaction scores.
- A service blueprint behind the journey map, showing frontstage actions, backstage systems and the engineer handoff. The marketing map must be extended to include installation and first bill, since value is realised there.
What would change my mind: if abandoners cluster on one handset model or one region, the cause is technical and the fix is a device fix, not a journey redesign. If they cluster by tenure or channel of purchase, it is a communication and expectation problem at the point of sale. Fix the moments of truth that carry the most volume and the most emotion first - that is pairing today, unless the data says otherwise.
2. Design the feedback and measurement approach so this does not recur.
Read the worked answer
The gap is that Meridian measures activity (tickets closed, app downloads) and not outcome or effort. A workable design:
- Close the loop at the point of failure. Trigger a two-question callback or in-app prompt only to users who abandon or contact support within 24 hours - this is where the actionable signal is. Keep it short enough that response rates stay meaningful.
- Track customer effort for the journey, not the ticket. A single effort question immediately after pairing success or a resolved ticket, reported by journey step. Ticket-level satisfaction will stay high because resolved tickets are self-selecting survivors.
- Add behavioural metrics to the journey map: abandonment by step, repeat contacts within 7 days, engineer visits preceded by a support contact, time from install to first successful reading. These are leading indicators and are harder to argue with than a quarterly brand survey.
- Attribute back to the blueprint. Every metric sits against a frontstage or backstage component, so a drop points at a team and a system.
Governance matters as much as instrumentation: agree who owns the end-to-end journey metric (the whole journey, not one channel) and review it monthly with marketing, contact centre and field operations in the same room. Otherwise the same data is re-litigated.
3. Marketing owns the journey map and technology owns the app. What would you do about that?
Read the worked answer
Treat it as a stakeholder and decision-rights problem rather than a documentation problem. Marketing owning the map is not wrong - it holds brand and customer research - but a map that stops at download and is invisible to delivery teams cannot influence design, so it becomes decoration.
The intervention:
- Make the map a shared artefact with a named owner for the end-to-end journey, and explicit contributors from contact centre, field operations and technology. Ownership of the map should sit where end-to-end outcome accountability sits; if no such role exists, propose one rather than assuming it.
- Run one joint workshop built on real abandoner evidence, not opinion. Shared data changes the conversation faster than shared frameworks.
- Agree a small joint scorecard - four or five journey metrics reviewed monthly by all three functions - so the relationship has a recurring reason to exist.
- Clarify who decides what: marketing decides proposition and messaging at the moments of truth; technology decides implementation and instrumentation; someone decides trade-offs between them.
Watch for the political trap: the COO's one-quarter deadline will tempt everyone to fix the app screen and declare victory. Anchor the commitment to journey abandonment through to first successful reading, so a prettier screen with the same 62% drop-off counts as failure.