The patterns below are the GMP failures that recur most persistently across food manufacturing environments — the ones that appear in audit after audit, across different sites, different products and different scales of operation. [ADD INFORMATION: Dhanu Audit's own audit data — nonconformity frequency by category across completed audits — to replace industry-pattern framing with proprietary findings]
What makes these five worth examining is not that they are unusual. It is that they are almost always known. In most cases the site is aware of the weakness and has normalised it. The audit does not reveal a hidden problem; it removes the permission to keep tolerating a visible one.
1. Cleaning performed but never verified
The most common GMP finding is not dirty equipment. It is clean equipment with no evidence that the cleaning process is effective.
Typical picture: a documented cleaning procedure exists, a cleaning schedule is completed and signed, and the equipment looks visually clean. What is missing is any objective verification — no swabbing program, no ATP results, no microbiological verification, no validation that the chemical concentration and contact time specified actually remove the soil present.
Visual inspection confirms the absence of visible residue. It says nothing about microbiological status, allergen residue, or biofilm in a location nobody can see. A cleaning program that relies solely on visual checks is monitoring appearance, not cleanliness.
What good looks like: a defined verification method appropriate to the risk, performed at a stated frequency, with results trended rather than merely filed. Failures trigger investigation, not just a re-clean.
2. Maintenance deferred and engineered around
Flaking paint above an open product line. A damaged door seal patched with tape. A cable tie holding a guard in place. A temporary repair that has been in service for two years.
Maintenance nonconformities persist because each individual instance appears minor and the production cost of addressing it appears high. The systemic problem is that temporary repairs accumulate, and the site loses the ability to distinguish between a controlled temporary measure and a permanent fixture that happens to be made of tape.
This category also covers equipment condition that creates cleanability problems: worn surfaces that can no longer be cleaned effectively, hollow legs and open frames that harbour soil, and food contact surfaces with cracks, pitting or degraded seals.
What good looks like: a maintenance system that records temporary repairs with an expiry date and escalates them when that date passes. Hygiene-critical maintenance prioritised on food safety risk, not production convenience.
3. Personal hygiene — the gap between policy and practice
The policy is comprehensive. The practice is variable. This is one of the clearest examples of why auditing documentation alone produces a misleading result.
Common observations: jewellery beyond policy allowance, hand washing performed inconsistently at entry, PPE worn incorrectly in ways that defeat its purpose, movement between raw and ready-to-eat areas without the required changeover, visitors and contractors held to a lower standard than staff.
The finding is rarely "there is no hygiene policy." It is that the policy is not being applied uniformly, and that internal monitoring has either not detected this or has detected it and not escalated it.
What good looks like: hygiene compliance monitored as an operational measure with results trended, contractors and visitors managed to the same standard as staff, and non-compliance addressed consistently regardless of who is involved.
4. Pest management — monitoring without analysis
Pest control is one of the most consistently documented GMP areas and one of the most consistently under-analysed.
Typical picture: a contractor attends monthly, devices are checked, a service report is filed, and the reports accumulate in a folder. What is absent is the site's own analysis — trending of activity by location, identification of entry points or harbourage, correlation with structural or housekeeping issues, and evidence that recommendations from previous reports were acted upon.
A stack of service reports showing repeated activity in the same location, with no corresponding action, is not evidence of pest control. It is documented evidence of an uncontrolled pest issue.
What good looks like: the site owning the pest program rather than delegating it to the contractor. Activity trended by location over time, contractor recommendations tracked as actions with due dates, and structural causes addressed rather than devices simply re-baited.
5. Allergen control — segregation assumed, changeover unvalidated
Allergen findings carry disproportionate consequence because the failure mode is undeclared allergen in finished product, which is a recall and a direct consumer safety risk.
Recurring weaknesses: changeover cleaning procedures that have never been validated as effective for allergen removal, shared equipment where segregation relies on scheduling that is not always achievable, storage where allergenic materials sit above non-allergenic ones, rework handling that loses allergen identity, and label verification that confirms the label is present rather than that it is correct for the product actually in the pack.
Precautionary labelling is frequently used to cover control weakness rather than genuine unavoidable cross-contact — which transfers risk to the consumer and, increasingly, attracts regulatory and customer attention.
What good looks like: validated changeover cleaning with objective evidence of allergen removal, rework controlled with full allergen identity maintained, and label verification that checks label-to-product match at the point of packing.
The pattern underneath all five
Each of these categories shares a structure: an activity is performed and documented, but its effectiveness is never established. Cleaning is done but not verified. Maintenance is logged but temporary repairs are not tracked to resolution. Hygiene is policed but not measured. Pest devices are checked but activity is not analysed. Allergen changeover is performed but not validated.
This is the same gap described in the validation article, appearing in a different part of the system. Documentation of activity is not evidence of control.
Practical takeaways
- Audit your own verification, not your own compliance. For each GMP program, ask what objective evidence exists that it is effective — separate from evidence that it was performed.
- Put an expiry date on every temporary repair. If it passes, it escalates automatically.
- Trend what you already collect. Most sites hold enough pest, hygiene and cleaning data to see the pattern. Very few analyse it.
- Validate allergen changeover cleaning specifically. General cleaning validation does not establish allergen removal.
- Watch for normalisation. If a finding has appeared in two consecutive audits, the issue is not the finding. It is the corrective action process.