There are several misconceptions about what acceptance inspection is, what it is used for and even if it is valid in today’s world of third-party audits and government oversight.  It went out of vogue about 15 to 20 years ago. The basic argument was that a quick spot check was sufficient as the vast majority of producers were putting out extremely high-quality product and the cost saving of rejecting substandard product just wasn’t there anymore. 

When I was a veterinary food inspector for the U.S. Army, I spent a lot of time doing acceptance inspections on receipt product for military use.  And while the procedures have changed over the years the basic premise is still the same: checking what you are receiving to verify that it meets your specifications and requirements, i.e., getting what you paid for. 

So why don’t companies use acceptance inspection anymore?  To be honest, it comes down to cost.  You have to pay someone to do the actual inspection, coordinate the load, develop the specifications, etc.  When you consistently have minor or no findings it isn’t worth the cost to do the inspection, and with the push for quality production over the last 25 to 30 years, in most cases it wasn’t worth the cost.

Times change.

I have been asked about pass/fail and or AQL acceptance inspections many times over the last several months. The No. 1 complaint is repeated findings of contamination.

How many recalls have been caused by changes in a supplier’s ingredients that weren’t noticed until it was too late?  Not to mention the level of contamination that is being found in general trim by grinders and other processors as well as miscut/off-spec product being shipped. We all know margins are tight. Profits come through pushing more product out the door. And with more and more involvement in the food industry by large investment groups, the pressure to increase profits has gotten to the point that quality is slipping across the board, or at least to the extent that it has become noticeable across the market.

The rumblings of late are about how raw material quality has dropped while costs have increased.  There is also the requirement to report contamination issues to the government, be it a U.S. Department of Agriculture (USDA) District Office or the Food and Drug Administration (FDA) hotline.  The other grumble being heard is that, even when reported, nothing gets done to fix the issues.  The end result is that folks are again looking at doing acceptance inspections when they receive products, and/or are requiring inspections to be done before products are shipped to them.

While you could rely on second- and third-party audits in place of acceptance inspections, those audits only give you a snapshot of how your suppliers operate.  To have an integrated approach, one that protects your bottom line, you need to be able to correlate the audit to actual performance. Using your acceptance inspections as part of that correlation data can be extremely useful.

So how do you develop an acceptance inspection program?  While there are a lot of ways I prefer to avoid re-inventing the wheel. The current ANSI/ASQ sampling plans for attributes are based on the old Military Standard 105E. The benefit of them is that they have been found to be valid sampling plans in court (they are legally and statistically sound).  They have the added benefit of being flexible and can be used to determine sample sizes as well as give guidance on what to do if you have failures, consistently high quality product, etc.

The basic gist of it is, you start a new vendor in a tightened sampling plan, then, as they prove themselves, you drop to normal. With a good history, you drop to a reduced sampling plan.  In the event they have failures, you reverse it and move back to tightened, etc. 

Here are some key components to acceptance sampling.

1. Your standards have to be just that, standards you support either contractually or through other means such as IMPS criteria.
2. Your sampling has to be consistent in its process and procedures.
3. You have to take the appropriate actions based on your findings.

The first step is developing your standards. 

What are your general requirements? These are usually things that apply to all products, for example free of contamination, proper temperature control, labeled weight is the actual weight, etc.

What are product-specific requirements? These are detailed and verifiable requirements such as course ground 80/20 meat is actually only 20 percent fat. You can get even more detailed and say you only accept product that is within 1 percent of your fat requirement.  Maybe you have a particle size requirement, or a packaging requirement.  Whatever your requirements are, they have to be clear and communicated to your vendors.

The second step is developing your sampling plans.  An easy place to get started is SQC online (www.sqconline.com). It has put together multiple sampling plan schemes. Simply select the one that best applies to you.  Alternately, you can develop your own based on your experience; however, make sure the scheme you develop is statistically valid for the type of inspection/product inspection you are performing.  You then need to train your team on how to perform the inspections. They must be proficient at the task and consistent in their implementation.

The final piece to the acceptance inspection puzzle is actually doing something with the data. If you are receiving non-conforming product, do you request a price adjustment from the vendor each time it doesn’t conform? Is the non-conforming product automatically rejected?  How does the vendor’s history of acceptance dictate additional contracts? 

A good use of the acceptance data is when it is tied to other performance measures such as final yield, product cost, third-party audits, etc. When you pull all of the data together, you can quickly determine which of your vendors is giving you the best “real” value, not just the lowest item cost. Another way of thinking about it is, if you lose 10 percent of product from Vendor A  because they are sending in poorly trimmed loins when compared to Vendor B, you haven’t just lost 10 percent of value. You have to calculate the lost man-hours for the additional trimming, the increased cost to bring in makeup product, etc. In the end, Vendor B comes in at 2 percent more expensive per item up front but ends up generating 5 percent more net profit.  Acceptance inspection can assist you in making these calculations.

Acceptance inspection can be used by very large companies or very small.  Although a small company won’t have the clout of a large company, it can use the data generated to make informed buying decisions. One of our smaller clients told its broker it wanted different product and why. The broker was able to get the product from a different source at approximately the same cost.  The big difference was that the new product didn’t require near the trimming, which resulted in increased profits for the smaller operator. I am also aware of a very large client who recently implemented acceptance sampling on receipt. It subsequently resulted in a major change in the way its vendor manufactured the source material. In the end, it was a win for everyone.

So what is the real downside of acceptance sampling?  As I mentioned earlier, cost.  You have to calculate the cost of implementing acceptance sampling against the net gain and determine whether it is worth it at your establishment. Whereas three years ago it may not have been worth the cost, if you now find your profits dropping because of issues with your source material, it may be worth revisiting.  NP