Pete Lysak | Kollective | MAY 2026
In Episode 3 I described how to identify drivers and barriers across the guest journey and which data sources to use. Employee experience data was one of the four. This episode is about why it gets left out more than any other, and what that omission costs.
A friction point shows up at a specific touchpoint, and the CX team goes to work on it. They redesign the process, upgrade the system, retrain the staff. Three months later the numbers have not moved, and most teams conclude the problem is stubborn, or complicated, or just needs more time. What they have not asked is who is actually running that touchpoint, what their working day looks like, and whether anything about their situation gives them a realistic chance of delivering what the process was designed for.
The process tells people what to do. It tells you nothing about whether they can do it. Someone three weeks into a role, handling guests in their second language, measured on how fast they clear the queue, will find a way through the shift. That is not the same thing as meeting a service standard.
The data that explains the data
In the three-layer KPI framework from Episode 2, employee experience sits in layer three, the root cause layer. Put touchpoint satisfaction scores next to employee experience data from the same touchpoints and you start to see things that neither the CX team nor HR would find on their own.
Tenure is the most consistent variable we have come across. Touchpoints staffed heavily by people under 90 days in the role score 15 to 25% lower on guest satisfaction than the same touchpoints once that group has been there six months. That is a wide gap, and it rarely appears in a VoC dashboard because no one thought to look for it there. When a touchpoint underperforms, the instinct is to go straight to the journey design. The better question is often whether the people running it were ready when they first went live with guests.
Engagement data tells a related story, and this one caught the HR team at a mature organization we work with completely off guard. We matched their team-level engagement scores against guest satisfaction at the corresponding touchpoints and the correlation came back at 0.72. They had been collecting both sets of numbers for years, in separate systems, reported at separate meetings. The first time anyone laid them side by side, engagement turned into a leading indicator, a way of seeing where guest satisfaction was heading before the guest feedback arrived.
The third variable is the one most within a company’s own control, which makes it the most frustrating. Frontline KPIs that reward throughput with no counterweight for quality produce exactly what you would expect: staff move guests through as quickly as they can. A decision made somewhere in HR or operations resurfaces months later in the driver and barrier data as friction at those same touchpoints, guests who felt processed and rushed. The journey design did not cause that. The incentive structure did, and no amount of journey redesign will fix it.
Why the cross-reference never gets done
Guest satisfaction data sits with CX or marketing. Employee experience data sits with HR. They live in different systems, run on different reporting cycles, and answer to different people. Nothing brings them together by default, and usually no one holds the mandate to force it.
When a complaint comes in, the reflex is to hand it to whoever owns the channel, and each function works only with what it can see. That is quick, and it is wrong whenever the real cause sits on the other side of a functional boundary.
At a corporate startup we built the cross-reference in from the beginning. Every touchpoint score was tied to the team responsible for it, so when something dipped, the first questions were about people rather than process: who was on shift, how long they had been there, what their onboarding had actually covered. Process still mattered. We had simply learned that it was rarely where the answer turned out to be.
In the mature organization it had to be retrofitted into systems that had never spoken to each other. We ran the layer three analysis each quarter and took the results into a room with the heads of CX, HR, operations, and commercial in it together. Watching the correlation between their own team’s engagement scores and guest satisfaction at their own touchpoints did something no report had managed. HR began asking which touchpoints deserved priority for training. Operations began asking HR about staffing patterns at the touchpoints that kept slipping. Functions that had run in parallel for years suddenly had a reason to talk, because the data had handed them a problem they shared.
None of this happens as fast as the analysis does. Ask a function to share accountability for a number it used to own outright and it will push back, often for good reasons. The only approach that has worked for us is to begin at layer one: the customer outcome metrics up on the screen at every leadership meeting, shown before any departmental update, presented by the most senior person in the room. Not as a CX agenda item, just the numbers the whole business is judged on. Held there consistently, that framing slowly changes what people treat as someone else’s problem.
Pete Lysak is a Consulting Partner at Kollective, where he advises on Voice of Customer programs, customer journey analytics, and EX-CX alignment for airlines, telecoms, retail, hospitality, destination, and service businesses in the GCC, EMEA, and Asia. He has launched two startups (aviation and telco) in Saudi Arabia as a founding team member, led brand and marketing for a major European airline, and held VP/Managing Director roles at McCann across EMEA and Africa. He can be reached at pete@kollective.uk
Next: Episode 5 - Making it stick: governance, and why most VoC programs die in year two