Investigating customer loss means examining the customer’s task, the service delivered and the evidence of departure. Mystery shopping can test possible service weaknesses; it cannot by itself identify which real customers have left or establish their reasons.
A useful investigation considers several explanations rather than assuming every departure was caused by price or poor service. Needs can change, a relationship can end naturally, or a competing offer can be more suitable. Focus on the avoidable problems the organisation can substantiate and address.
Start with the evidence of loss
Define what “leaving” means for the business: cancellation, non-renewal, inactivity, reduced purchasing or another recorded outcome. Choose a suitable period and comparison group. A quiet account may need investigation, but silence alone does not establish dissatisfaction or a decision to switch.
The published abstract of Susan Keaveney’s 1995 exploratory study of service switching describes multiple categories of reported switching reasons. It supports investigating more than one explanation; it does not provide a current Australian churn benchmark or a universal ranking of causes.
Turn possible causes into testable questions
| Possible issue to investigate | What mystery shopping can observe | What requires additional evidence |
|---|---|---|
| Unclear value or conditions | Whether an employee explains price, relevant conditions and the proposed solution. | Whether actual customers considered the offer poor value and whether that influenced departure. |
| Unnecessary effort | Transfers, inconsistent instructions, missing ownership and repeated steps within the authorised scenario. | The frequency and consequence of those barriers across real customer journeys. |
| Broken promises | Whether a specific authorised commitment is explained and completed within the observed task. | Its wider prevalence and relationship with cancellations or repeat purchasing. |
| Weak recovery | Listening, explanation, a suitable referral and ownership in an approved problem scenario. | Actual complaint resolution, perceived fairness and subsequent customer behaviour. |
| Poor fit | Whether needs are explored and relevant options or limitations are explained. | Whether a different product or provider genuinely better meets the customer’s needs. |
A fictional accumulated-friction journey
A customer receives an unclear quotation, calls to clarify it, gets conflicting instructions and must chase a promised update. The customer later chooses an alternative. This fictional sequence suggests questions about consistency and ownership. It does not prove which incident determined the choice, or that a discount would have prevented it.
A controlled assessment can test whether the quotation and handover weaknesses recur under the defined conditions. Exit feedback can explore the customer’s explanation. Account and service records can establish timing and recorded actions. Those sources should inform one investigation, while retaining their different limitations.
Choose the right evidence for the question
| Evidence source | Useful contribution | Limit |
|---|---|---|
| Mystery shopping | Tests defined behaviours and task delivery in controlled scenarios. | Does not sample the attitudes or subsequent purchases of the actual customer population. |
| Exit feedback or interviews | Explores customers’ stated reasons and experiences. | Non-response, recall and a simplified stated reason may affect interpretation. |
| Complaints and service records | Shows recorded issues, actions, repeated contact and completion. | Customers who do not complain may be absent; records may be incomplete. |
| Transactions or account data | Establishes recorded renewal, purchasing, inactivity or cancellation patterns. | Shows what changed, not necessarily why it changed. |
Avoid turning an observation into a retention claim
Finding a missed follow-up is a reason to investigate and improve that process. It is not proof that the business lost a sale, that a customer would have stayed, or that correcting the issue creates a particular revenue uplift.
Similarly, repeat purchasing does not on its own establish preference or trust. An organisation may retain business because of convenience, limited alternatives or an ongoing contract. Read service observations alongside customer feedback and behaviour, rather than using any one measure as a complete loyalty score.
Close the investigation with a proportionate action
Select a supported problem, name the process owner, agree a workable change and define adoption evidence. Retest the behaviour with comparable scenarios. Review actual customer outcomes over a suitable period, accounting for changes in product, price, staffing and market conditions before attributing an effect.
Use the action-plan guide to assign the response, the complaint-handling guide for recovery observations, and the measurement comparison to keep observation and customer opinion distinct.
Sources and scope
Prepared by Secret Customer Australia · Published 13 November 2026. Prepared as a public adaptation of Why Customers Leave Businesses (SC-RD-002) and selected Drivers of Customer Loyalty (SC-RD-003) concepts. The investigation tables are Secret Customer guidance; the journey is fictional. The linked 1995 study’s publisher abstract was checked on 13 November 2026; no current national survey or client-retention result is claimed.
Browse the Secret Customer Academy or discuss your objectives through our mystery shopping services.
