Why Dispensaries Fail Compliance Checks Even When Their METRC Reports Look Correct

The METRC report is clean. Every package balance ties out. The audit trail runs back 90 days without a gap. You checked it yourself this morning.

An hour into the inspection, the officer has found a discrepancy in your vault.

This happens more than most compliance managers want to admit, and it doesn't mean the METRC record was wrong. It means the physical inventory drifted from the METRC record after the last time someone reconciled them.

How compliance checks work

State cannabis regulators don't announce inspections in most jurisdictions. An officer arrives, identifies themselves, and begins work. Standard procedure involves pulling your license records from the state compliance system, selecting a sample of package IDs or bin locations for physical verification, and counting the physical product.

The officer is comparing what METRC says you have against what's physically in the room. They're not reading your SOPs. They're counting grams.

If the physical count disagrees with METRC by more than the allowable tolerance for your state, you have a discrepancy. What happens next depends on the magnitude of the discrepancy, whether you have documentation of your reconciliation process, and whether the officer has seen similar patterns in your facility before.

A first discrepancy with documentation that explains it is a note in your file. A pattern of discrepancies without documentation is a violation.

Why METRC can be clean and the physical count can still be wrong

METRC reflects what was reported to it. It doesn't reflect what physically happened in your facility since the last time someone reconciled the two.

The window between your last reconciliation and today's inspection is where the discrepancy lives.

Here's the most common version of this. Your compliance manager ran a cycle count last week. It took most of the day. Everything tied out. A METRC adjustment was made on two packages with minor variances. The record is clean.

Since then, 600 sale transactions have processed through the POS. A vendor delivery came in. Two packages were processed and subdivided. One product moved from storage to the sales floor.

Each of those events was recorded in METRC. Each recording depended on a human entering the correct weight or quantity at the point of transaction. A budtender who pulled 3.4g instead of 3.5g. A receiving entry that rounded to the nearest gram. A package subdivision that was off by one unit because the scale read slightly different on a different surface.

By today, your METRC record and your physical inventory are not the same thing they were last week. They're close. But close is the space where inspection discrepancies live.

The most common sources of drift

POS weight variance. Most dispensary POS systems record the sale weight entered by the budtender. If the budtender enters the weight from the scale verbally rather than reading the exact number, that weight may be rounded. Over hundreds of transactions, rounded weights compound into discrepancy.

Package subdivision errors. When a bulk package is subdivided into smaller packages, the total weight of the subdivided packages must equal the original package weight in METRC. If the subdivisions don't sum correctly, METRC and physical reality diverge immediately. This is one of the most common sources of inspection failures.

Transfer entry lag. Product arrives. The METRC entry happens at the end of the shift. For several hours, you have physical product that METRC hasn't recorded yet.

Destruction weight estimates. Destruction entries in METRC require a weight. If that weight is estimated rather than measured on a calibrated scale, the estimate introduces error.

What a continuous verification layer changes

An inspection that arrives during normal operating hours and finds discrepancies is an inspection that found you in the drift window. The last count showed clean. The next count would also show clean. But right now, in this moment, the weights don't match.

A continuous weight-based verification layer changes the drift window from weeks to minutes.

When a scale sits under each storage container and reports real-time weight, the system knows what's in each container at every moment. When a transaction is recorded, the expected weight change is compared to the actual weight change immediately. A variance gets flagged within seconds, not discovered weeks later.

The inspector who arrives at a facility running continuous verification faces a different audit. The records reflect current physical reality, not last week's snapshot. Discrepancies that exist are documented, investigated, and already have a corrective note in the audit trail.

That's the difference between a discrepancy that triggers a violation and a discrepancy that results in a note and a question.

A specific scenario

A dispensary in a state with strict METRC tolerances was getting flagged in compliance checks for vault discrepancies. They were running cycle counts every two weeks, their compliance manager was thorough, and they couldn't identify a systematic error.

After reviewing their workflow, they found the source: package subdivisions in the back room were being weighed on a bench scale that was 18 inches off-center from its level position. The scale read consistently, but it was reading slightly low. Over time, the subdivision entries were consistently 2 to 4 grams below the physical weight of what was actually going into each package.

Once the scale was recalibrated and repositioned, the discrepancies stopped. But they only found the scale problem because they deployed a continuous verification station in the subdivision room and saw the pattern in the data.

Without continuous monitoring, that error would have kept compounding until the next inspection found it.

What to do before the next inspection

Run a weight audit of your current METRC balances against physical inventory. Don't wait for the inspector to run it for you. Count your highest-value packages first.

Check your scale calibration dates. Every scale used for METRC entries should have a current calibration certificate. Most state programs have specific requirements on calibration frequency.

Review your package subdivision records for the past 90 days. If subdivisions exist where the subdivided weights don't sum to the parent package weight within tolerance, investigate and document.

Learn more about Cloudbox continuous inventory verification for cannabis retail at cloudboxapp.com/cannabis, review the compliance integration documentation at cloudboxapp.com/compliance, or contact the team at cloudboxapp.com/contact-us to discuss your specific compliance risk profile.