I walked into a plant on a Tuesday morning and asked for their pre op sanitation records. The QA manager opened a filing cabinet and started digging. She pulled out one folder, then another, then a third. She was shuffling through loose sheets, checking dates, squinting at handwriting. That was the moment I knew this was going to be a long day.
I spent ten years as a USDA Consumer Safety Inspector. I walked into hundreds of plants. Big ones, small ones, brand new operations, and facilities that had been running since before HACCP was a word anyone used. And over all those years, I learned something that most operators do not hear until it is too late: inspectors are not looking at your equipment first. We are not walking the floor first. We are looking at your records. The records tell us whether your system actually works or whether it just looks like it does on paper.
This post is the conversation I wish I could have had with every plant manager before I showed up with my clipboard. These are the five categories of records that FSIS inspectors prioritize, what we are actually looking for inside them, where plants fail most often, and what you can do today to tighten things up.
Pre Operational Sanitation Records
This is usually the first thing I looked at. Before the line starts running, someone at your plant is supposed to walk the facility and verify that every food contact surface has been cleaned and sanitized. That verification needs to be documented. Every single day.
Here is what the textbook will tell you: the record should show that pre op sanitation was conducted, that it was effective, and that any corrective actions were taken if something was not right. That is accurate. But here is what I was actually looking for when I opened that binder.
I was looking for gaps. I was looking for days where no record existed at all. I was looking for records where every single entry said "acceptable" for six straight months with zero deviations ever noted. That is not a clean plant. That is a plant where nobody is actually looking. Real sanitation programs find problems. A record that never shows a corrective action is a record I do not trust.
The most common failure mode I saw was not filthy equipment. It was missing documentation. The plant did the work. They cleaned everything. But nobody wrote it down, or they wrote it down on a loose sheet that got lost, or the person who was supposed to sign it was out sick and nobody covered. When I ask for your pre op sanitation logs and what FSIS expects to see in them and you cannot produce them within minutes, that is a problem. That is the start of a Noncompliance Record.
What you can do today: make sure your pre op sanitation records are completed before the line starts, every day, with no exceptions. Make sure they include the date, the name of the person who conducted the check, what was inspected, what was found, and what was done about it. And make sure someone reviews those records regularly, not just files them.
CCP Monitoring Logs
Critical Control Points are the backbone of your HACCP plan. They are the steps in your process where you are actively preventing, eliminating, or reducing a food safety hazard to an acceptable level. And CCP monitoring logs are the proof that you are doing it.
When I pulled CCP monitoring records, I was not just checking that someone filled in the numbers. I was looking at the pattern. Are the values consistent? Do they make sense for this process? Is there variation that tells me someone is actually measuring, or does every entry say the exact same thing day after day? If your internal cooking temperature log shows 165 degrees for every single batch for three months straight, I am skeptical. Real processes have variation. Real monitoring captures that variation.
The failure mode here is twofold. First, plants record values late. The monitoring is supposed to happen at the frequency your HACCP plan specifies, but in practice, someone fills in the log at the end of the shift from memory. That is not monitoring. That is creative writing. Second, when a value does fall out of range, the response is often missing or incomplete. You had a CCP deviation. What did you do? Where is the corrective action? If the log shows a temperature of 140 degrees and the next entry is just back to normal with no explanation, that tells me your system broke down.
What you can do today: make sure your CCP monitoring is happening in real time, at the frequency your plan requires. Make sure the person doing the monitoring understands what a critical limit is and what to do when one is breached. And make sure every deviation has a documented corrective action attached to it. No gaps. No blank lines. No retroactive entries.
Corrective Action Documentation
This is where I saw the most damage during FSIS inspections. Not because plants were hiding problems, but because they genuinely did not understand what a complete corrective action record looks like.
When a CCP deviation occurs, or when any part of your food safety system fails, you are required to take corrective action. That means four things, and all four need to be documented. You need to identify the cause of the deviation. You need to show what you did to bring the process back into control. You need to show what you did with the affected product. And you need to show what you are doing to prevent it from happening again. 9 CFR 417.3 — Corrective Actions is the regulation behind this.
Most plants get the first two. Something went wrong, they fixed it, they wrote that down. But the product disposition piece and the preventive measure piece are where records fall apart. I would see corrective action logs that said "adjusted temperature" but had no mention of what happened to the product that was processed during the deviation. Was it held? Was it reworked? Was it destroyed? That matters. That is what I am looking for. That is what separates a plant that has its system under control from a plant that is reacting and hoping nobody asks questions.
What you can do today: build a corrective action form that forces all four elements. Do not leave it to memory. Do not leave it to whoever happens to be on the floor that day. Create a structure that requires cause, correction, product disposition, and preventive measure. Every time.
HACCP Plan Reassessment Records
This one catches people off guard. Most plant operators know they need a HACCP plan. Most of them have one. But a surprising number of them have not reassessed that plan in years. And when I ask for reassessment records, the room gets quiet.
FSIS requires that your HACCP plan be reassessed at least annually, and also whenever there is a change that could affect your hazard analysis. That includes new products, new equipment, new suppliers, new processes, regulatory changes, or a pattern of recurring deviations. The reassessment does not have to be complicated. But it does have to exist, and it has to be documented. 9 CFR 417.4 — Validation, Verification, Reassessment spells it out.
The failure mode here is simple neglect. The plan was written three years ago. It worked. Nobody touched it. Meanwhile the plant added two new product lines, switched suppliers on a key ingredient, and started running a second shift. None of that triggered a reassessment because nobody had a system to flag it. When I pull your reassessment records and the last entry is from 2021, that is a finding. And it is the kind of finding that calls your entire HACCP system into question, because if you are not reassessing, how do I know your hazard analysis is still valid?
What you can do today: put a recurring date on the calendar for your annual HACCP reassessment. Document it even if nothing changes. Write down that you reviewed the plan, considered whether any conditions have changed, and concluded that the plan remains adequate. That record is your proof. And any time you make a process change during the year, document the reassessment decision alongside it.
Written Prerequisite Programs and SOPs
Prerequisite programs are the foundation your entire HACCP system sits on. These include your Sanitation Standard Operating Procedures, your pest control program, your supplier approval process, your employee hygiene practices, your allergen control plan, and anything else that supports food safety outside of your CCPs. Inspectors look at these because if the foundation is weak, the HACCP plan built on top of it does not matter.
What I was looking for here was not just whether the written programs existed. I was looking for whether they matched reality. Does your SSOP say you sanitize food contact surfaces with a 200 ppm chlorine solution? Is that what is actually happening on the floor? Does your pest control program specify monthly inspections by a licensed operator? Can you show me the last three reports? SSOPs Made Simple: what every USDA plant needs to know walks through this in more detail.
The most common failure mode in this category is drift. The written program was accurate when it was created. But over time, the plant changed how it operates and nobody updated the document. The SOP says one thing. The floor does another. That gap is where noncompliance lives. And it is one of the easiest things for an inspector to spot, because all I have to do is read your program and then walk your floor. If those two things do not match, we have a problem.
What you can do today: pull every written prerequisite program and SOP you have. Read them. Walk the floor while you read them. Ask yourself whether the document describes what is actually happening today, not what was happening when it was written. Update anything that has drifted. And then build a review cycle so you are doing this regularly, not just when an inspector forces the question.
The Bottom Line
The plants that pass FSIS inspections consistently are not the ones with the fanciest equipment or the biggest budgets. They are not the ones with the thickest HACCP manuals or the most expensive consultants. They are the ones where someone actually looks at the records every single day. Someone reviews the pre op logs before the line starts. Someone checks the CCP monitoring sheets before the shift ends. Someone makes sure the corrective action file is complete before it gets filed away. The discipline is in the daily review, not in the binder on the shelf.
After ten years of walking into plants and seeing the same failures repeat at facility after facility, I knew the problem was not the people. It was the system they were stuck using. Paper records that got lost. Logs that nobody reviewed until an inspector forced the question. Values that drifted out of range for hours before anyone noticed. That is the problem I set out to solve.
John Parisi is the cofounder of U.S. AgriDocs and a former USDA Consumer Safety Inspector with ten years of federal inspection experience.