Bram De VosA customer experience audit is a structured, evidence-based review of what customers actually go through at every touchpoint, set against what the organisation intended to deliver. Done well, it produces a prioritised list of the moments that cost revenue, the moments that earn loyalty, and the blind spots in between.
Key takeaways:
- An audit closes the gap between the experience a company designs and the one customers live through. In Bain & Company's delivery gap study, 80% of companies said they delivered a superior experience; 8% of their customers agreed.
- The audit works in four layers: touchpoints, the processes behind them, the people delivering them, and the feedback loop that should catch problems. Most disappointing audits never leave the first layer.
- Pair every experience metric (NPS, CSAT, CES) with the operational number that explains it. A score alone is a symptom; the pair is a finding.
- Audit the silence, not just the complaint log: the customers who wanted to complain and could not are your unhappiest segment, and most unhappy customers never complain at all.
- An audit ends with owners, deadlines and a 90-day plan, or it was a workshop.
There is a quiet distance between the customer experience a company designs and the one its customers actually have. Scores arrive on a dashboard, complaints arrive in a queue, and somewhere between the two, the experience customers live through can go unexamined for months.
That distance has been measured. In Bain & Company's well-known delivery gap study, 80% of companies said they delivered a superior experience; 8% of their customers agreed. Twenty years on, the pattern persists: in PwC's 2025 Customer Experience Survey, nine out of ten executives said customer loyalty had grown, while four in ten consumers said the same.
A customer experience audit closes that gap. This guide walks through the eight steps, the conditions that decide whether the audit changes anything, and the mistakes that quietly ruin the exercise.
What a customer experience audit is
Think of the difference between a thermometer and a physical examination. Your NPS or CSAT dashboard is the thermometer: it tells you whether the patient has a fever. The audit is the examination: it finds out why, and what to do about it.
Concretely, a CX audit is a point-in-time assessment across four layers: the customer-facing touchpoints, the processes and policies behind them, the frontline people delivering them, and the feedback loop that should catch problems. Most audits that disappoint stay in the first layer; the experience at the counter is produced three layers deep.
The payoff is practical: you learn where customers drop out and why, which fixes will move your scores, and where next year's CX budget should go, the raw material for a customer experience strategy.
Before you start: three conditions
Three things decide whether an audit changes anything, and all three are settled before the work begins:
- An executive sponsor who can act across silos.
- A defined scope.
- Reserved capacity to act on the findings.
An audit with no landing strip for its conclusions is an expensive way to confirm what everyone suspected.
Step 1: define the scope
Pick one or two journeys, a segment and a time window. "The onboarding journey for new SME customers over the past six months" produces findings someone can own; "our customer experience" produces a 60-page report and a long silence. Choose the journey that carries the most revenue or churn risk, or the one where complaints already cluster. Narrow and deep beats broad and shallow.
Step 2: map the journey as the customer lives it
Map every touchpoint in the chosen journey: every channel, every handover, every moment the customer has to wait or decide. Two rules make the map useful:
- Map from the customer's side, in the customer's order. Internal process maps follow departments; customers do not.
- Include the gaps. The silence between order confirmation and shipping notice is a touchpoint, and some of the worst moments in a journey are the ones where nothing happens.
Keep it to one page; a map you can hold in one glance is a working tool.
Step 3: walk the journey yourself
Nothing in the audit replaces first-hand experience. Lean practitioners call it going to gemba: go to the place where the work actually happens and watch. Buy your own product. Call support with a genuine question, at 4:55 pm on a Friday. Submit a complaint. Try to cancel. Take timestamped notes: how long each step took, what you had to figure out, what you were promised and when it arrived.
The small frictions insiders have learned to ignore are exactly what customers stumble on. Michel Stevens, course director at CXM Academy, tells a story on our CX Matters podcast about sitting with the customer service team of a utilities company that was drowning in calls about house moves. All the information customers needed was available online, but it lived in a document the company had named, in all seriousness, the document of abandonment. No customer would ever search for that. Nobody inside had noticed for years, because as he puts it, the last thing a fish discovers is the water it swims in.
That is why fresh eyes matter. Let a few colleagues outside the CX team do the same walk, and listen hardest when they trip over something everyone else considers normal.
Step 4: collect the quantitative evidence
Gather two kinds of metrics for the scoped journey and read them together:
- Experience metrics: NPS, CSAT and CES. These are perception metrics: how the journey felt.
- Operational metrics: first response time, resolution time, repeat-contact rate, delivery versus promise, drop-off, churn by cohort. These describe what actually happened, and further downstream sit the outcome metrics they feed: retention, share of wallet, revenue.
The pairing is the point: a dip in CSAT after delivery means little on its own; next to a rise in delivery delays it is a finding.
Two warnings from practice: distrust averages (four hours on average can hide a fifth of customers waiting two days) and distrust trends without context (a stable NPS can mask a shrinking response rate). If the numbers feel shaky, you are in broad company: in McKinsey's survey of more than 260 CX leaders, only 15% said they were fully satisfied with how their company measures CX, and just 6% were confident their measurement supports real decisions.
Step 5: listen to what customers actually say
The scores tell you where it hurts; the words tell you why. Four sources:
- Open feedback at scale. The open answers behind your NPS and CSAT surveys are the richest audit material you own. Classified into topics and linked to scores, they show which themes drive satisfaction and which create detractors; at volume, a voice of customer platform such as Hello Customer does that classification for you.
- Interviews. Eight to twelve customers in depth: recent buyers, long-time customers, and above all the people who left. Churned customers have no reason to be polite.
- Support tickets and call logs. Contact reasons map where the experience fails; every repeat contact is a broken promise with a timestamp.
- Public reviews. They skew extreme, which is their value: these are the moments that made someone go public.
Audit the silence, not just the complaint log
A good complaints process invites a dangerous conclusion: no complaint, no problem. The numbers say otherwise. In research shared during a webinar with us, covering Belgian banking and insurance, the average NPS sat at 15. Among the customers who had filed a complaint it was 14. But among the 4% who said they should have complained and never did, it was minus 32: the unhappiest group in the entire base was invisible to the complaint process. Another slice of customers, roughly one in eleven, simply leaves without ever saying a word.
For the audit, the instruction is simple: treat declining response rates and quiet drop-off as findings in their own right, and make giving feedback effortless everywhere. The customers who could not be bothered to complain are not neutral; they are gone.
And one habit beats every report format. When leadership sees a chart, they nod; when they read one raw verbatim from one real customer, they lean in. Put actual customer words in the audit, unpolished.
Step 6: audit the machinery behind the experience
For every weak moment found so far, trace the cause through three questions:
- Is it a process or policy? Many "experience problems" are policies working exactly as designed, for the company's convenience rather than the customer's.
- Is it a people or tooling issue? Sit with frontline agents for half a day; they usually know precisely where the experience breaks.
- Is it the feedback loop itself? Check whether feedback reaches the teams who can act on it, and whether anyone closes the loop with the customer.
Step 7: score and prioritise
Score each finding on impact (how many customers it touches, how strongly it relates to scores and churn) and effort to fix. A key driver analysis makes the first half of that objective: it ranks which themes actually move your metric, so you fix what demonstrably drives the scores, even when something else is louder in the boardroom. Quick wins go into the next sprint, structural fixes onto the roadmap with an owner and a budget, and low-impact items are written down and consciously parked.
Step 8: turn findings into a 90-day plan
An audit ends with commitments. For each priority finding, write down the owner, the deadline, the fix and the metric that will show whether it worked, and summarise the whole audit on one page for leadership. Then review monthly. That discipline is what separates an audit from a workshop that everyone remembers fondly and nobody acts on.
How often should you audit?
A workable rhythm:
- A full audit once a year.
- A focused review each quarter, on the journey with the most friction.
- Continuous listening in between: the audit is a photograph, your voice of customer programme is the film.
If the organisation changes fundamentally (a merger, a replatforming, a new service model), audit sooner: the old findings describe a company that no longer exists.
Five mistakes that quietly ruin a CX audit
- Auditing only the digital journey. Customers experience your channels as one brand, and the audit should too.
- Mistaking a metrics report for an audit. If the output contains no verbatims, no first-hand observation and no root causes, it was a reporting exercise.
- Skipping the customers who left. Auditing only current customers is asking the survivors how the voyage went.
- Letting it end as a slide deck. No owners, no deadlines, no metrics: no audit.
- Leaving the frontline out. Agents hold years of unrecorded findings; involving them also builds buy-in for the fixes.
Frequently asked questions
How long does a customer experience audit take?
Four to eight weeks for one or two journeys; company-wide audits take a quarter or more, which is an argument for scoping tightly.
Who should run it: an internal team or an external party?
Internal teams know the context and cost less; external auditors bring fresh eyes and independence. A pragmatic middle ground: run it internally and let an outsider challenge the conclusions.
What is the difference between a CX audit and customer journey mapping?
Journey mapping is one instrument inside the audit. The map describes the experience; the audit measures it, judges it against evidence and ends in a prioritised plan.
Which metrics matter most in a customer experience audit?
Whichever pair proves your point per moment: an experience metric (NPS, CSAT or CES) next to the operational metric that explains it.
What does a CX audit cost?
Run internally, the main cost is time: several weeks of one lead plus contributions across teams. External audits range from a few thousand euros to far more. The more useful question is the cost of the unfound problems, usually paid in churn.
The audit is the honest mirror
Every organisation carries a story about its own customer experience, and the story is always kinder than the reality. An audit replaces the story with evidence: this is what customers meet, this is what it does to them, and this is what fixing it is worth. The organisations that improve fastest are the ones that look at themselves most honestly, and most often.