In brief
- Customer experience quality is falling, not rising. In Forrester's 2025 US index, four times as many brands lost ground as gained it, and the declines trace back to a handful of repeatable causes.
- Root cause analysis is not driver analysis. Drivers tell you what generally moves satisfaction; RCA explains a specific drop and names the lever that, if fixed, prevents it from recurring.
- The payoff is leverage. Measured at the journey level rather than the touchpoint, the few causes that matter are far more predictive of who stays, who churns, and what to fund first.
A satisfaction score drops ten points in one quarter, in one segment. The reflex is to do something visible: retrain the agents, rewrite the email, add a callback. A quarter later the score is still down, the team is busier, and no one can say which of the three fixes worked. None of them did, because none of them touched the cause.
Root cause analysis exists to break that loop. It answers one question: why did this happen, specifically, here, now. Then it names the single change that makes the problem go away and keeps it away.
The problem is getting worse, not better
This is not an academic concern. Across the market, experience quality is sliding, and the slide is concentrated enough to diagnose.
Forrester traces the decline to a short list of recurring causes: weaker employee experience, waning customer obsession, disappointing technology rollouts, and economic pressure that makes customers question the value they get. That is what a root cause looks like at market scale. The score is the symptom. The cause is upstream, and it repeats.
Exhibit 1
In 2025, declines outnumbered gains four to one
Root cause analysis is not driver analysis
The two are routinely confused, and the confusion is expensive. Driver analysis ranks the factors that generally move satisfaction or retention across your base. Root cause analysis investigates a specific event: this drop, this segment, this touchpoint, this month. Drivers tell you what to improve in the abstract. RCA tells you what went wrong in the case in front of you, and what to do about it.
The two are most powerful together. Driver analysis narrows the search space. RCA confirms which driver actually caused this incident, then quantifies the fix.
A score tells you something is wrong. Only the cause tells you what to do, and whether it will work twice.
Measure the journey, not the moment
The biggest reason RCA fails is that teams diagnose at the wrong altitude. They examine the touchpoint where the score dipped, when the cause usually sits in the journey around it. A clean handoff to billing matters less than whether the customer ever reaches a resolved bill.
The evidence for working at the journey level is direct.
When you connect survey signal, behavioral data, and operational records across the journey, the noise collapses. A few moments turn out to explain most of the variation in whether a customer stays or leaves. Those moments are where the cause lives.
A disciplined way to find the cause
RCA is a method, not an instinct. The teams that get repeatable answers run it the same way every time.
- State the problem precisely. Not “satisfaction is down,” but “NPS fell 10 points in Q3, concentrated in new commercial accounts.” A vague problem produces a vague cause.
- Establish the timeline. When did it start, and what changed in that window: a pricing move, a release, a policy, a staffing shift. Causes have start dates.
- Segment until the pattern sharpens. If the drop is everywhere, you have not segmented enough. The cause hides where the effect is concentrated.
- Form hypotheses, then test them. Do detractors mention onboarding more than promoters? Did the dip follow the release? Let the data falsify your favourite theory.
- Name the root cause and price the fix. Land on the one factor that, if changed, prevents recurrence. Then estimate the value of fixing it, so the recommendation arrives as a business case, not an opinion.
The discipline matters because the alternative is guessing, and guessing is what fills roadmaps with fixes that never move the number.
Why the cause beats the symptom
Treating symptoms feels like progress and produces motion without movement. Finding the cause does the opposite: less activity, more effect, and a problem that does not come back next quarter wearing a different mask. It also builds an organizational habit of asking why, so recurring issues get retired at the root instead of being managed forever.
To see how we run this with clients, explore our Root Cause Analysis and Experience to Impact work, or browse the case studies.
Sources
- Forrester, "Forrester's 2025 Global Customer Experience Index Rankings," forrester.com.
- McKinsey & Company, "From touchpoints to journeys: Seeing the world as customers do," mckinsey.com.