Customer experience audit

012 min

Why a customer experience audit exists

Picture a head of customer experience at a subscription app. After every support chat, customers rate the help they received, and the average is high. Yet many new users cancel during their first month. The scores say customers are content. The cancellations say they are not.

Both can be true, because the survey only reaches people who contacted support. The head of customer experience needs a way to see what the survey does not reach. That is the job of a customer experience audit. Teams usually start one when escalations rise, satisfaction scores fall, or customers report different experiences on different channels.

022 min

Where the customer experience audit comes from

The audit that answers this problem has no single inventor. It is a practice built from older tools, each of which covered one part of what an audit now covers.

The oldest is mystery shopping, in which a person poses as a customer and records how staff treat them. The method became common in the 1940s and has changed little since. It shows how staff behave in one visit, but not what happens before or after it.

A modern audit uses all three ideas. It takes the walk-through from the customer's point of view from mystery shopping, the view of hidden work from the blueprint, and the single contact as the unit of study from Carlzon.

034 min

How a customer experience audit works

Those older tools each cover one part. An audit covers the whole journey, and it works in three moves.

First, it sets a scope. The team picks one question, such as why new users cancel in the first month, and decides which customers, channels and dates it covers. The team should include people from marketing, sales, product and support, because each of them owns some of the touchpoints.

Second, it maps the journey and lists every touchpoint. A journey map lines up the stages a customer goes through: learning about the product, buying, getting started, using it, getting help, and renewing or leaving. The team then lists the touchpoints in each stage. They fall into four kinds. Digital touchpoints include the website, email and online reviews. Human ones include phone calls and meetings. Physical ones include packaging and printed documents. Process ones include scheduling, billing and how problems are resolved.

Third, it gathers evidence from three directions and compares it with what the company says it delivers. The promise is written down first: what the website claims, what sales says, what the service is meant to be like. The evidence then shows what really happens.

No single row is enough, so the audit sets them side by side. Where all three point at the same step, the finding is strong. Where they disagree, the disagreement is the next thing to investigate. The team can also repeat the walk-through on a competitor, or compare today's scores with its own earlier ones, to see how a result compares.

Why an audit needs several auditors

A walk-through means an auditor goes through the journey as a customer would: signing up, using the product, asking for help, cancelling. When the auditor hides who they are, the method is called mystery shopping or secret shopping. Secret shopping on its own checks service standards and staff behaviour, not the full experience.

One auditor is not enough. The Nielsen Norman Group advises that three to five people should review the same interface independently, because each reviewer misses some problems. That advice was written for screens, as in heuristic evaluation, but the reasoning holds for a journey too. Have several people go through it alone, then compare what each one noticed.

How an audit listens to customers

Surveys give numbers. The three common ones are Net Promoter Score, which asks how likely a customer is to recommend the company, Customer Satisfaction Score, which asks how satisfied they are, and Customer Effort Score, which asks how hard something was. Used alone, each hides something, so audits read them together. Interviews give reasons. The useful measure for interviews is saturation, the point where new interviews stop revealing new themes, not a fixed number.

Choose interviewees on purpose. Include customers who succeeded and customers who struggled, high-value and low-value ones, and people who have already left. The people who left are the ones a survey of current customers never hears from, which is a case of survivorship bias.

042 min

How a customer experience audit ranks its findings

An audit usually ends with more problems than a team can fix at once. Ranking decides which ones come first.

For an audit, this means one problem at the end of a journey, such as a hard cancellation, can outweigh several small problems earlier. It also means auditors should note the strongest good and bad moments, not average every touchpoint evenly. A point of friction near the end costs more than the same friction in the middle.

How an audit becomes a plan

Teams then sort the ranked problems by how much effort each fix needs. Cheap fixes with a clear benefit go first, because they show progress early. Larger changes go on a roadmap. Every finding gets an owner, a date and a way to tell whether the fix worked. The value of the audit lies in that plan, not in the report.

Last, the team books the next audit. One audit shows one moment in time, so it cannot show whether things improved. Repeating it after major changes, and when warning signs return, turns the audit into a way of tracking change.

052 min

Customer experience audit compared with similar reviews

Several reviews sound alike and answer different questions. A UX audit examines how usable the interface is, while a customer experience audit also covers support, billing, marketing and every other point of contact. A customer service audit looks at support and service interactions only. Each of the others in the table is a narrower review, or a tool that an audit may use.

The journey map is the closest neighbour. An audit often produces one, but a map drawn from the team's own memory records what the team believes. The audit tests that belief.

061 min

When a customer experience audit goes wrong

An audit goes wrong in a few predictable ways, and each one has a visible sign.

• No scope. Without a clear question, the team collects data that does not connect, and ends with recommendations nobody can act on.
• A one-time project. A single audit shows one moment in time. It cannot show whether a fix worked or a new problem appeared.
• Numbers without reasons. Scores show what is happening, not why. An audit that has no interviews or walk-throughs cannot explain its own scores.
• Findings kept in one department. Touchpoints belong to different teams. If only one team reads the report, the other teams' problems stay as they were.
• Reviewing your own work. An internal team knows the product well but may defend it. An outside reviewer is more objective but costs more. A mix of an internal expert and an outside specialist gives both.

071 min

Running a customer experience audit

Return to the subscription app. The head of customer experience limits the audit to one question: why do new users cancel in their first month? The scope covers sign-up, onboarding, first use, billing and cancellation. Two auditors who have never used the app sign up and go through the first month alone. The team reads the cancellation records and interviews users who left.

The chat survey stays in the audit as one source. The team notes what it measures: only the moments after a customer asked for help. Anything that confused a customer without making them contact support is invisible to it.

If you are starting your own audit, begin with one journey, one question and two or three people from different teams. Write down what the company promises at each step before you look at what happens. Then rank the findings by severity, give each one an owner, and book the next audit before you close this one.

?4 questions

Questions people ask

§12 sources

Sources