Cloud4Feed
Request a Demo
Customers Pricing
Request a Demo

Mystery shopping: what it is and how to use it with VoC

Customer feedback tells you how the experience felt. Mystery shopping audits whether the experience you designed is actually delivered, step by step, in every location. Here is how it works, where it pays off, where it falls short of real-customer feedback, and how to run both together.

Cloud4Feed Research Team Updated 10 min read
Quick answer

Mystery shopping sends trained evaluators, posing as customers, to check service and brand standards against a checklist. It is an audit tool: precise on standards, but small in sample and costly, so it complements VoC rather than replacing it.

Illustrative
Audit score = (Points achieved ÷ Points possible) × 100Try the calculator ↓

At a glance

Who
Trained evaluators posing as customers
Measures
Whether service and brand standards are delivered
Result
Compliance score, 0% to 100%
Sample
A few visits per location per period
Role
An audit that complements VoC

What is mystery shopping?

Mystery shopping is a structured audit in which trained evaluators, the mystery shoppers, act as ordinary customers and then report what happened against a set of agreed standards. A visit can take place in a store, a branch, a dealership, on the phone, by chat or on a website.

The method has been used since the 1940s, first by banks and retailers to check employee integrity. Today it is mainly used to answer one question that customer feedback cannot answer reliably on its own: are we doing what we said we would do?

That makes it different from Voice of the Customer (VoC). VoC measures how real customers perceive the experience; mystery shopping measures whether a defined standard was delivered. One is a perception measure, the other an audit.

How a mystery shopping programme works

  1. Define the standards. Turn your service model into observable criteria, decide which are critical, and give each one a weight in points.
  2. Write the scenario. Describe who the shopper is, what they want and what they will ask, so every location faces the same situation.
  3. Plan the visits. Decide how many visits each location gets per period, and spread them across days, times and staff.
  4. Brief and calibrate the shoppers. Train evaluators on the checklist and scenario so they score the same behaviour the same way.
  5. Visit and report. Shoppers complete the checklist and a short narrative straight after the visit, with evidence such as receipts where allowed.
  6. Check, score and share. Review each report for quality, calculate the score, and send results to the location and its manager quickly.

How to calculate the audit score

Audit score = (Points achieved ÷ Points possible) × 100

Criteria marked N/A, because the situation did not arise, are removed from the points possible.

  1. Give every criterion a number of points that reflects how important it is.
  2. Mark each criterion as met, not met or not applicable after the visit.
  3. Add up the points of the criteria that were met, and the points of all applicable criteria.
  4. Divide the first total by the second and multiply by 100.
Worked example8 criteria worth 100 points in total. A demonstration was not relevant for this visit, so its 10 points are N/A: 90 points possible.The criteria that were met are worth 65 points.Audit score = 65 ÷ 90 × 100 = 72.2%Report critical criteria separately: a missed legal disclosure should be visible even when the overall score looks fine.

Audit score calculator

Add your criteria, their points and the result of the visit. The score updates as you go.

Audit score
72.2% audit score
Points achieved
65
Points possible
90
Points excluded as N/A
10
Criteria not met
2

Mystery shopping vs VoC

The two methods look at the same experience from different sides. Mystery shopping is controlled and precise but small; VoC is broad and representative but cannot tell you exactly which step was skipped.

AspectMystery shoppingVoC (real customers)
Who gives the dataA trained evaluator posing as a customerReal customers
What it measuresWhether defined standards were delivered: behaviour, process, complianceHow customers perceived the experience and how loyal they are
Sample sizeSmall: usually one to a few visits per location per periodLarge: hundreds or thousands of responses
Cost per data pointHigh: every visit is recruited, briefed, paid and checkedLow once the programme is running
Kind of judgementA structured checklist, filled in by one observerMany personal opinions, aggregated
CoverageAny moment you can script, including pre-sale and competitorsOnly moments where customers are invited and choose to answer
TimingScheduled visits, reports within daysContinuous, close to real time
Best forAuditing standards, compliance and training needsUnderstanding satisfaction, loyalty and what drives them

Strengths of mystery shopping

What it does well

  • Checks exactly what you define: a greeting, a disclosure, a process step, a sales script.
  • Gives concrete, behaviour-level findings a store manager can coach on the same week.
  • Covers moments VoC rarely reaches: the pre-sale conversation, people who did not buy, phone and online enquiries.
  • Lets you visit competitors with the same checklist and compare like for like.
  • Produces documented evidence for brand, franchise and regulatory compliance.
  • Uses a controlled scenario, so locations are compared on the same situation rather than on different customer mixes.

Limitations compared with VoC

Every limitation below is a reason to keep mystery shopping in the role of an audit, and to let real-customer feedback lead. Each comes with the usual way to manage it.

Cost

Each visit has to be recruited, briefed, carried out, checked and paid for. The cost per data point is far higher than a VoC survey response.

How to manage it: Spend visits where a standard really matters: high-value sales, regulated steps, new processes.

Small sample

One visit is a snapshot of one employee, at one time, on one day. A single bad or good visit can swing a location’s score.

How to manage it: Read trends across several visits and roll results up by region or network before drawing conclusions.

Not a real customer

The evaluator knows they are testing and follows a script. Their needs, expectations and emotions are not those of your customers.

How to manage it: Use VoC to learn what real customers value, and build the scenarios from it.

Only finds what you ask

A checklist can only confirm or deny what is on it. It does not discover problems you have not thought of.

How to manage it: Update the checklist regularly from the topics that come up in VoC comments.

Observer judgement

Different evaluators can score the same behaviour differently, especially on soft criteria such as “friendly”.

How to manage it: Write observable yes/no criteria, train and calibrate evaluators, and review reports before release.

Detection and trust

Staff may recognise evaluators, and a programme used to punish individuals damages trust and invites gaming.

How to manage it: Rotate evaluators, use results for coaching, and tell staff the programme exists and what it checks.

Legal and ethical limits

Recording calls or visits, and using personal data, are regulated (for example under KVKK in Türkiye and GDPR in the EU).

How to manage it: Agree the rules with legal and HR before launch, and follow an industry code of conduct.

Where to use mystery shopping, and where not to

Use mystery shopping to…
  • audit brand and service standards across stores, branches and dealers
  • check a new process, campaign or script after rollout
  • evidence regulatory or contractual compliance, such as disclosures in financial sales
  • audit franchise and partner networks
  • benchmark competitors on the same scenario
  • find training needs at team and individual level
Use VoC or another method to…
  • understand loyalty and word of mouth: NPS
  • measure satisfaction with real experiences at scale: CSAT
  • measure how easy a task was for real customers: CES
  • decide which standards matter most to customers: MaxDiff

Sample checklist questions

Greeting

Was the shopper greeted within 30 seconds of entering the store?

Needs analysis

Did the advisor ask at least two questions to understand the shopper’s need?

Disclosure

Did the advisor explain the total price, including fees, before asking for a decision?

Environment

Were the shelves stocked and the sales area clean?

Narrative

Describe the interaction in your own words, including anything that helped or got in the way.

Tip: write criteria a stranger could check without interpretation. “Greeted within 30 seconds” can be audited; “friendly welcome” mostly measures the shopper’s mood.

Using mystery shopping alongside VoC

The strongest programmes connect the two. VoC, and a prioritisation method such as MaxDiff, show which moments matter most to customers; mystery shopping then audits whether those moments are delivered. Comparing a location’s audit score with its NPS or CSAT shows whether your standards are the right ones.

VoC score (NPS or CSAT) →
Check the standards Low audit, high VoC: customers are happy anyway. Is this standard really needed?
Standards working High audit, high VoC: protect and share good practice.
Fix execution Low audit, low VoC: standards are missed and customers notice. Coach and revisit.
Wrong standards High audit, low VoC: staff follow the checklist but customers are unhappy. Rewrite it from VoC.
LowHigh
Mystery shopping audit score →

Feed VoC comments back into the checklist every few cycles, and use mystery shopping to cover the gaps VoC leaves: the pre-sale conversation, prospects who did not buy, and competitors.

What is a good audit score?

A good score is the target you set for your own standards, not an external number: every company’s checklist and weights are different, so scores cannot be compared between companies. Many programmes set a target score for the network and a zero-tolerance rule for critical criteria.

Because each location receives only a few visits, judge it on the trend across several visits and on its gap to the network, not on one report. When most locations score near 100%, the checklist is probably too easy, or no longer measures what customers notice.

How to act on mystery shopping results

Share each report quickly

Send the visit report to the location within days, while staff still remember the shift. A finding that arrives weeks later is hard to learn from.

Coach, don’t punish

Use findings to coach teams and recognise good practice. Treat critical failures as cases with an owner and a deadline, in the same way you close the loop on a low VoC score.

Revisit and recalibrate

Schedule a follow-up visit after an action plan, and review the checklist against VoC every few cycles so the audit keeps measuring what customers value.

Mystery shopping FAQ

Yes, when it is run openly at programme level and within data protection and employment law. Employees should know that mystery shopping takes place and what it checks, even if they do not know when. Recording audio or video and processing personal data need a clear legal basis, so agree the rules with legal and HR before launch.

Cloud4Feed Research Team

Our research, analytics and CX team writes about what we learn running feedback programmes across 15+ industries.

About us →

New guides and research, once a month.

No sales emails. Unsubscribe anytime.

We use your details to respond to your request. See our Privacy Policy.

Ready to put this into practice?

A 45-minute walkthrough set up for your industry and your team.