Marketing teams usually design the beginning of a relationship with almost obsessive care: awareness, the first visit, the first lead, onboarding, the welcome sequence, the first purchase. The end of the relationship is treated very differently. It is often left to support scripts, cancellation forms, refund rules, claims teams, legal language, or a queue that nobody in marketing wants to own.
That is the useful provocation in Adrian Swinscoe’s July 2 CMSWire article about customer journey endings. The piece uses Zurich Insurance’s bereavement-support service as the example, but the lesson is broader than insurance. Any company that sells subscriptions, services, ecommerce, financial products, software, memberships, or long B2B projects has endings: cancellations, returns, failed renewals, complaints, disputed payments, account closures, contract exits, and moments when a customer is simply exhausted.
The uncomfortable part is that those moments may shape memory more strongly than the polished beginning. A brand can spend heavily on acquisition and still lose trust at the point where the customer needs clarity, dignity, and help.
Why endings deserve design, not leftovers
CMSWire frames this through the idea of endineering: the deliberate design of how a customer relationship closes or enters a painful final stage. The concept matters because most organizations are much better at onboarding than offboarding. They know how to welcome a customer, but not how to let a customer leave, recover from a failure, or complete a difficult process without feeling abandoned.
Zurich’s example is powerful because the company does not treat bereavement support as a minor service improvement. It pairs AI-driven logistics and secure digital tools with human care managers, keeping automation in the administrative layer and people in the emotional layer. That distinction is important. It shows a mature use of technology: AI reduces burden, but it does not pretend to perform empathy.
For marketers, the broader principle is simple. The end of a journey is not a back-office corner. It is a brand moment. In categories where stress, money, trust, or grief are involved, the way the relationship ends can become the story people tell about the company.
The CMO problem: poor endings distort growth metrics
A weak ending does not only create a bad service experience. It distorts the economics that marketing reports later. If cancellations are confusing, returns are hostile, complaint handling is slow, or account closure feels punitive, the business pays through support cost, negative reviews, lower referral intent, weaker reactivation, and a lower-quality demand base.
This is why a CMO should not treat ending design as a soft customer-experience theme. It belongs in the operating model. A customer who leaves cleanly may come back when the need returns. A customer who leaves angry becomes more expensive to reacquire, and may influence other buyers long after the immediate revenue is gone.
The source article also points to the commercial case behind Zurich’s approach: research cited by CMSWire says 57% of consumers would choose a life insurer based on bereavement support, while Zurich’s own research found that many consumers would avoid brands that fail to show empathy. Even if those numbers are category-specific, the signal is not. Compassion, clarity, and operational design can affect consideration, not only satisfaction.
What should be measured
The practical move is to give endings the same discipline that marketing gives acquisition. For subscription businesses, that may mean cancellation completion time, save-offer acceptance, repeat contact rate, churn reason quality, reactivation rate, and the share of customers who leave without an unresolved complaint. For ecommerce, it may mean refund friction, return cycle time, cost per return, exchange recovery, review sentiment after returns, and repeat purchase after a problem.
For B2B and service companies, look at handoff quality, project closure satisfaction, time to resolve billing disputes, renewal-loss reasons, escalation frequency, and whether an exiting customer remains a future advocate. None of these metrics should live only in support dashboards. They should be visible to marketing, product, operations, and finance because each team influences the final experience.
The owner also needs to be explicit. If everyone partly owns the end of the journey, nobody really owns it. A serious model names the accountable team, defines the budget, sets the service standard, and reviews the data regularly. Otherwise the final experience becomes the place where organizational silos are exposed to the customer.
How to redesign the end
Start by mapping the moments the brand would rather ignore: cancellation, refund, complaint, claim, account closure, failed payment, downgrade, contract exit, death of an account holder, return after disappointment, or a transition to another provider. Then ask what the customer must understand, what they must do, how much effort it takes, and where anxiety enters the process.
From there, reduce cognitive load. The customer should see the next step, the timeline, the responsible contact, the documents required, the status of the case, and the likely outcome. This is not decorative UX. It is trust infrastructure.
Finally, decide where automation helps and where it should stop. Automate status, routing, document collection, reminders, and routine checks. Keep human judgment for emotional, ambiguous, high-value, or vulnerable situations. Zurich’s case is interesting precisely because it avoids the lazy answer that every service moment should become a chatbot.
The real takeaway
The end of the customer journey deserves a budget, an owner, and KPIs because it is not merely the end of revenue. It is often the beginning of the next reputation cycle: will the customer return, recommend, warn others, complain publicly, or quietly exclude the brand from future consideration?
For a CMO, the decision question is direct: if the company is spending more to acquire demand while leaving the final experience underfunded, is marketing actually growing the business, or just compensating for preventable trust leakage?
