# Smoothing things over before an inspection is the riskiest strategy. Three months before the date, everyone knows where the weak points are. The temptation is to make them disappear from the surface. **That is exactly what turns an observation into a major finding** — because a hidden gap, once discovered, stops being a technical gap and becomes a question of reliability. Two questions An inspector wants to know, first, whether you know your own weaknesses, and whether you have a credible plan to address them. The rest of the interview turns on the answer to those two questions. The same technical fact is graded differently depending on whether it was presented or discovered.The same gapknown, traced, plannedMinor observationdiscovered in the roomMajor finding A constant observation across inspections supported in regulated industry. A site that answers "yes, we know, here is since when, here is the analysis and here is where we stand" is in a radically different position from one discovering the topic in the room. **The first demonstrates a system that works; the second demonstrates a system that does not see.** The same technical gap then leads to two very different conclusions. "Never volunteer anything that was not asked for." As a matter of form the advice makes sense — answer the question put, do not wander. It becomes toxic when it turns into a policy of concealment. **An experienced inspector finds the grey areas: they are visible in the indicators, in review dates, in the findings of the annual quality review.** Letting them be discovered rather than presenting them gains you nothing, and costs you the only asset that matters in that room: credibility. ## Four weeks that change the outcome. ### 1\. The honest inventory List what will not hold: procedures overdue for review, late CAPAs, expired qualifications, untreated recurring deviations. This inventory does not leave the site, but it must be complete. A known and planned weakness can be defended; one discovered in the room cannot. ### 2\. Triage by risk Not everything can be fixed in four weeks, and trying to fix everything guarantees fixing the important things badly. What touches the product and the patient comes before what touches formality. The rest goes into a dated plan with a named owner — which is already an acceptable answer. ### 3\. Preparing the teams Not learning answers, but practising how to answer: say what you know, say you do not know when that is the case, and know who to refer to. An operator reciting a script is spotted immediately, and it triggers the next question. An operator explaining their workstation in their own words reassures. ### 4\. The site's narrative Being able to set out in ten minutes where the quality system stands, what has improved since the last inspection, what is still open and why. That narrative frames how every subsequent observation is interpreted. Without it, each gap is read in isolation, and therefore at its worst. ## The inspector is not an adversary. They are a partner, provided you hide nothing. Authorities are open to innovation and simplification when you set out the reasoning, the data and what you deliberately changed. **What they do not forgive is opacity.** A site that explains why it removed a control, with its risk analysis, will be questioned on the substance. A site whose control vanished without a trace will be questioned on its reliability — and that is no longer the same inspection. ## An inspection decided on substance. Severity The same technical fact can be graded a minor observation or a major finding, depending on whether it reveals a system failure. Transparency acts precisely on that grading. Duration An inspector in doubt widens the sample. A legible site shortens the inspection, because each check confirms what was announced rather than opening a new line of enquiry. Follow-up An action plan presented before it is demanded becomes the basis of the post-inspection response. You gain weeks on the next cycle, at the moment the organisation is most stretched. If the deadline is short and the system genuinely fragile, the right call is usually **to bring in help early rather than late**: four weeks is enough to change the preparation, two days before arrival is only enough to tidy the offices. ## Go further. [ ### Remove without losing the trail What makes a reduction defensible. Read →](/en/blog/documentation-simplification-auditable-traceability/)[ ### The CAPA that holds The first interview topic. Read →](/en/blog/capa-effectiveness-corrective-action/)[ ### The gemba walk See the workstation before the inspector does. Read →](/en/blog/gemba-walk-quality-culture-gmp/) ## Preparation, plainly. Should you volunteer an unresolved gap? + If your system knows about it and it is tracked in a plan, yes: presenting it in the site narrative places it in a controlled frame. If your system never detected it, the question is not whether to disclose it — your detection system is the real topic, and it will be seen. How long does preparation take? + Four to six weeks allow an inventory, risk-based triage, treatment of the critical points and team rehearsal. Below two weeks, most of the gain shifts to interview preparation and the coherence of the narrative — still useful, but it does not rescue a file. Should operators rehearse? + Train them, yes; have them learn answers, no. A recited script is detected in two questions and triggers suspicion. The useful training is about posture: answer what is asked, own up to not knowing, know who fetches the information. And if the inspection is announced for next week? + You do not rewrite a system in a week, and you should not try. You build the narrative, you check that the documents to be presented exist in their current version, you prepare the interviews, and you settle the action plan that will be presented. That is the format of our emergency support. ## Do you know what the inspector will find? If the answer is yes and the plan exists, you are ready. If not, four weeks can still change the outcome.