When to need to mystery shop to examine real touchpoints.

Mystery shopping is a method of placing a trained person in the customer's position, going through real touchpoints, and recording what happens based on measurable criteria. It does not ask customers how they feel after returning home, nor does it inquire about what the team did. It records what actually happens, in a real environment, at that moment. This is why Sinh Vũ incorporates this tool into the audit of touchpoints when experiences depend on people.

Why internal reports and surveys are not enough

Reports from the team reflect what they observe or want to convey. Surveys reflect what customers remember and are willing to write down. Both have value, but both filter reality through an intermediary layer. Conducting anonymous trials of that intermediary layer by being present at the touchpoint, recording based on specific criteria, not relying on memory or willingness to recount.

When diagnosing a brand, the most dangerous thing is not the lack of data but having data that does not reflect reality yet is used to draw conclusions. Conducting anonymous purchases is a way to verify whether the existing data is accurate.

Situations where mystery shopping should be used

• Series of multiple points of sale need to measure consistency. When there are many branches or points of sale, experiences at each location can differ significantly, which aggregated reports may not reveal. Mystery shopping helps measure service standards at each point and compare them.
• Suspect that the real experience is very different from what the team reports. If you sense there is an issue but internal data always looks good, conducting a test purchase is a benchmarking tool to obtain concrete evidence, not subjective feedback.
• Want to measure service compliance periodically. For example: do employees greet customers within 30 seconds, do they introduce products according to the process? These criteria need to be observed directly and cannot be asked of customers afterward.
• Currently building or standardizing the experience before expansion. Before scaling the model, it is essential to know where the current standards stand. Conduct trials for a realistic picture to standardize correctly.

When there's no need to buy a sample, or it can be replaced in another way

What needs to be determined beforehand is: how much the experience at the touchpoint depends on people and operational processes. The more it relies on people, the more valuable a trial purchase becomes.

Common mistakes that cause trial purchase data to be useless

• Ambiguous scoring sheet. Questions like "Are the staff friendly?" allow each buyer to rate differently. Criteria need to be specific, measurable, for example: "Staff greet within 30 seconds of customer entering, yes or no." Without clear criteria, data cannot be compared across sessions or locations.
• Only try once and then conclude. A single trial purchase may not be representative. Multiple cycles are needed to see patterns and distinguish systemic issues from isolated incidents.
• Taking the feedback of test buyers to represent the entire customer segment. Test buyers come with their own missions, criteria, and perspectives. They are not typical customers. Test purchase data measures service standards, not the overall emotional response of customers.
• Skipping data quality checks. Raw data from test buyers needs to be verified before analysis. Skipping this step can lead to biased data influencing conclusions unknowingly.

An effective trial purchase program requires a structured evaluation form, specific measurable criteria, and a quality control step for the data before analysis.

Intouch Insight, What is Mystery Shopping

The perspective of Sinh Vũ when incorporating trial purchases into the audit

In the touchpoint audit, Sinh Vũ collects real samples from various sources. When experiences depend on people, anonymous purchase trials are included to see what customers encounter in reality, not just what the team describes. This approach adheres to measurable criteria, labeling by layer and channel, to identify where the experience breaks down with specific evidence rather than vague comments.

One thing Sinh Vũ often observes: trial purchase results are often not entirely "bad." They show where consistency exists, where there are deviations by person, and where there are deviations by point of sale. From there, you can identify which areas to standardize first, rather than retraining everything based on intuition.