Seen as a customer. Judged as an owner.

Seven steps

• 01
  Define the class. We establish three things before any visit: what the business intends to be, what it charges, and what its customers currently believe it is. We assess against the promise, but we test the promise too. An owner's ambition and the market's view are not always the same.
• 02
  Visit unannounced. Visits are made as an ordinary customer, without warning to staff. One visit is a first reading, not a verdict: a full assessment covers different days, times and customer journeys before any broad conclusion is drawn.
• 03
  Record the evidence. Every observation is recorded with its time, its touchpoint, what was expected and what was found, and a confidence label: verified, repeated or indicative. Only verified or repeated observations support a formal finding. Photographs are taken only where permitted and never of staff or customers' faces; conversations are not recorded.
• 04
  Grade each touchpoint. Each touchpoint is graded against a written rubric for its industry and class, with the evidence attached. Critical failures in hygiene, safety or law are reported separately and immediately.
• 05
  Report. For each finding: what was seen, the evidence, why it matters to the customer, its likely commercial consequence (stated as likely, not as measured), the smallest change that would lift it, and its priority by cost, effort and effect. Important findings are reviewed a second time against the evidence before the report is issued.
• 06
  Set the standard. If the client continues, the findings become a written perception standard: the objects, words, timings and sensory choices for each touchpoint, with a named owner for each. It is built around the team, budget and working day the business actually has. A standard a team cannot keep is not a standard.
• 07
  Keep it. Return visits, unannounced. They ask whether the original gap has closed, whether the standard holds across visits, and whether anything new has slipped. The progress report shows what improved, what did not, and what needs attention next.

On AI

Judgement is human. Memory is not.

A person walking through a business notices forty things and remembers twelve. AI helps us record every observation in the moment, sort it by touchpoint, compare it with earlier visits and draft the report for review.

AI never decides a business's class, never assigns a grade, and never fills in what was not observed. A well-written report is not evidence. The evidence is.

Independence

We never sell a grade.

Assessment and improvement work are scoped and billed separately, and an assessment can be bought on its own. A client who has paid for improvement work can still receive an unchanged or lower grade on a return visit, with the evidence for it. What we sell is the expertise to earn a better grade.

Beyond mystery shopping

Good mystery shopping firms build custom criteria and benchmark branches well, and for checking consistency across many sites they are the right choice. Our work starts a step earlier and ends a step later: