The Customer Journey Doesn’t Care Who Owns the Platform
Every system can report success, and the customer still doesn't get help. Why CX leaders need to test the full journey, not just each platform's uptime.
Topics
Let me start with a fairly ordinary customer situation. You receive an alert about a suspicious transaction, so you call the number on the back of your card. You authenticate through the IVR, select the fraud option, and hear that your call is being transferred.
Then the line goes silent.
Inside the organisation, everything might still look fine. The carrier says the call connected. The interactive voice response (IVR) says authentication worked. The routing platform says it issued the transfer. The contact centre says agents were available. Every team has a green dashboard, yet you still have a customer who couldn’t get help.
I have spent much of my career in core telephony, and this is one of the most persistent problems I have seen. Customer journeys rarely fail in a neat place. They fail at the handoff, where one platform stops looking and the next starts.
Breaking Down Silos is Only Half the Job
There’s lots of positive work happening around organisational silos. Marketing, digital, operations, IT, and customer service are being brought together around shared outcomes. A 2025 CX Dive article on CX councils made the case for connecting teams so that what a brand promises in one channel is delivered in another.
I agree with this, but I think we have to take it a step further. What about the technical silos, for example? A call can pass through a local number, several carriers, an IVR, speech recognition, routing logic, customer data systems, a contact centre platform, and perhaps an AI agent nowadays. A different team or supplier may own every stage.
The customer doesn’t know that, nor should they have to. They called one brand. To them, it’s all just one conversation.
The Genesys State of Customer Experience 2025 reflects this challenge. Customers want connected experiences, while many CX leaders are still working with disconnected technology, data, and departments. So yes, align the organisation, but don’t assume that means the live journey works.
The Problem Lives Between the Dashboards
Take a global customer call. The number must work in the customer’s country and network. The voice may traverse several carriers. The correct prompt must play. The IVR must understand the customer. Routing must be accurate, the transfer must complete, and the customer’s context should follow them. That’s a pretty long communication chain.
A number may work with one carrier but fail with another. The wrong language prompt may be uploaded. A customer might press one and go nowhere. A voice agent may understand a clean lab recording but struggle with someone calling from a car, a busy store, or a weak mobile connection.
Even take me, right? I speak English with an Irish accent (Google the Waterford accent if you’re very curious). A voice agent trained on narrower acoustic conditions may interpret me differently than it would someone in London, Shanghai, or New York.
These are real customer conditions.
Each platform can still report success within its own boundary. The carrier connected the call. The IVR accepted the input. The routing engine sent the instruction. But did the customer reach the right destination? Did they repeat themselves? Was the issue resolved? That’s where the brand damage tends to live.
Uptime is not Resolution
I’m not arguing against dashboards. We need availability and operational reporting. We do, however, need to be honest about what those metrics tell us. A dashboard reports what happened inside a system. It doesn’t always tell us what happened to the customer.
The Qualtrics Contact Centre Trends 2025 research highlights continuing frustration around wait times, reaching the right person, and resolving issues during the first interaction.
Those are journey outcomes. Uptime alone cannot explain them. The question has moved on from “Did the platform respond?” to “Did the customer get what they needed, and did they trust us more or less afterwards?” A connected call is not necessarily a successful call. An interaction that ends is not necessarily resolved.
Journey Maps Need Live Evidence
Most large organisations have journey maps that show how an interaction is meant to flow across teams and channels. But a map is still a design. It doesn’t prove that the experience works today from Japan, Germany, Guatemala, or anywhere else customers are calling from.
Forrester’s 2026 guidance on customer journey management argues that businesses need to move from mapping journeys to managing them as a measurable discipline linked to operational decisions and business outcomes.
Teams need shared evidence that the whole journey worked. That means looking at routing accuracy, transfer completion, audio quality, context preservation, customer effort, and final resolution alongside availability.
When everyone sees the same end-to-end result, the conversation changes. We stop asking which team owns the failure and start asking where the customer journey stopped.
Test From the Outside In
I think the most practical way to understand a journey is to experience it the way the customer does. For voice, that means generating real calls from the countries and networks where customers are located. It means navigating the live IVR, validating prompts and languages, testing speech and keypad inputs, following the routing path, and confirming that the transfer reaches the intended destination.
The test should not stop because one platform reports success. It should continue through to resolution.
Pre-launch testing is necessary. But it’s table stakes, not a permanent safety net. Carrier routes change. Vendors release updates. Prompts are replaced. Routing rules are amended. Contact centres migrate, as we all know. AI agents evolve (continuously, as we all know!).
Metrigy’s 2026 research on CX assurance for AI-powered contact centres makes the same wider point: organisations need continuous testing before deployment and observability once systems are live.
My advice is simple. Take one of your highest-volume or highest-risk journeys. Test it in the production environment across two or three key markets. Use real-world conditions and follow them to the final outcome. Don’t assume that because it worked in the lab, it works everywhere.
The Brand Owns the Result
A business can outsource its carriers, IVR, contact centre platform, and AI technology. It simply cannot outsource the customer’s perception of the experience. When a call fails, customers won’t blame the local carrier or routing vendor. They’ll blame the company they called.
Breaking down organisational silos helps teams share responsibility. Testing from the outside in gives them a shared view of reality.
We need both.
The customer journey doesn’t care who owns the platform. It only cares whether the next step works.
ALSO READ: Why Retail Media is Moving Beyond Attribution to Incrementality
