9 replies to this topic

[SerenityNow!]

Hello All,

 

My inquiry is regarding COA's and whether we can create our own COA based off an external, 3rd party laboratory results. 

 

Some added context as to why I am asking this:

 

First off, we test our product to customer specifications.  Although it is generally the same product, not all product is tested equally.  If product doesn't meet certain parameters, it may be rejected for one customer, but be well within spec for another.  In these cases, we may have to submit again for analysis that may not have been tested the first time. 

 

We are currently going through a rather rough transition with out COA's. As we have had some growth within the company, we have more personnel collecting samples and sending them off to the lab. The issue at hand is that there are many mistakes as to what the product is being tested for.  Some samples are getting tested for things that weren't required, others are missing tests then we end up having to resend product which create delays and an not to mention added costs. Some are improperly labeled resulting in errors on the COA. One process needed to be sent to the lab about 14 times.  Why 14 times? It became so convoluted I couldn't really tell you, nonetheless it happened and we see so many retests due to mistakes.  I imagine the costs are extremely high.

 

We would like to take results from the COA's with data provided and issue them ourselves, in some cases.  That way, there aren't 14 COA's for one batch. We simply want to report the requested tests with out extra or missing data.

 

I know the real issue is to fix the root cause of the issue, but this has not been something that our FS department has been unable  to do as we are not given the authority over the personnel collecting/sending samples (crazy, yes).  The unusual part is that, Food Safety is handled by the Food Safety Director and quality technicians are under the Site Director.  For whatever reason, Food Safety and Quality have been separated and this has been grandfathered in to the facility and mindset.  Due to the transition, since about February, we have had to deal with the mistakes "as they learn".  Problem is that instead of improving, its just become status quo and ok to carry on this way.  When the mistakes happen, only then is is thrown to FS to talk to the lab and explain to the customer.

 

Ok, I got carried away, but can we issue our own COA's based off valid data?

[Scampi]

hmmmmmmm 

 

So you want to test product multiple times when it doesn't meet 1st consumer requirements??

 

Would it not make more sense to test for ALL parameters and then have a form that you tick that tells you which customers you can sell it to?  Less confusing for everyone, one lab analysis to track and YES you can then generate your own CoA

[jfrey123]

Who creates the COA at your business?  If it's your FS department, you would think that your department should have some control over it...

 

From what it sounds like, you guys are encountering three problems based on current practices:

1) Delays in shipping when material doesn't meet first customer spec and new product has to be found/tested.

2) Delays and incurred costs in retesting product lots that weren't included in previous tests.

3) Increased overhead in processing the extra tests.

 

I landed on Scampi's idea too, test everything for everything right off the bat so you know who it can be sold to if it doesn't meet the first customer's requirements.  I know that becomes expensive, but you could take the time to review the past 6 months of COA's and put together a report detailing the costs of the retests and overhead associated.  Based just off your description, there's likely a cost savings to testing upfront versus holding the inventory and waiting for retests.  Not to mention improved turnaround times to your customers' orders.  I'd call this option 1.

 

I like to present extra options when I think the bosses are likely to balk at my first choice.  So as a second option, maybe take the top 5-10 customers (or however many you need to account for the top 50% of your sales) and make a list of all of their testing requirements.  Make it standard practice to test for all of their requirements, and correlate those results into a standardized COA.  Sure, the COA will list things that some customers won't need, but it makes more sense to me than to create customer specific COA's every single time.

[kingstudruler1]

Yes, you can create COAS based of third party results.  I've never used and rarely see a coa that was created by a third party lab.  I've always created my own based off of internal or third-party data.    There are no regulations or standards for COA creation.   The caveat to this is "what is the customer's expectation?"

Edited by Charles.C, 09 July 2023 - 06:05 AM.
moved

[SerenityNow!]

hmmmmmmm 

 

So you want to test product multiple times when it doesn't meet 1st consumer requirements??

 

Would it not make more sense to test for ALL parameters and then have a form that you tick that tells you which customers you can sell it to?  Less confusing for everyone, one lab analysis to track and YES you can then generate your own CoA

 

Hi Scampi,

 

I may not have explained properly.  For instance, Customer 1 requires various tests. With the exception of moisture, all other results met their requirements. They require 4.5% moisture and the result is 5%.  Customer 2 will accept up to 6% (industry standard).  So in that case, the product wouldn't meet C1's requirements but meets C2's.  However C2 may have additional tests required that either weren't tested the first go or they may require multiple of the same test (common with several customers).  So in this case we would want to combine results on our own COA from the first and second set of results.  Just wanted to clarify that bit.

 

Thank you for your advice, we definitely need a new way to approach this as what is in current practice isn't ideal.

[FSQA]

IMHO the bigger issue is that multiple people are sending the same samples.

Before implementing any major changes, you would need to create a centralized system of "all samples" going out to 3rd party labs.

 

You can use it on a software (if using one) or can be done on an internal spreadsheet.

Every sample going out to the lab need to be registered in the same, across all departments. This will give visibility to all and would mitigate repetitive samples going out for the same testing.

 

You can create your own COAs based on 3rd party lab data/COA.

Edited by Charles.C, 09 July 2023 - 06:05 AM.
moved

[SerenityNow!]

Who creates the COA at your business?  If it's your FS department, you would think that your department should have some control over it...

 

From what it sounds like, you guys are encountering three problems based on current practices:

1) Delays in shipping when material doesn't meet first customer spec and new product has to be found/tested.

2) Delays and incurred costs in retesting product lots that weren't included in previous tests.

3) Increased overhead in processing the extra tests.

 

I landed on Scampi's idea too, test everything for everything right off the bat so you know who it can be sold to if it doesn't meet the first customer's requirements.  I know that becomes expensive, but you could take the time to review the past 6 months of COA's and put together a report detailing the costs of the retests and overhead associated.  Based just off your description, there's likely a cost savings to testing upfront versus holding the inventory and waiting for retests.  Not to mention improved turnaround times to your customers' orders.  I'd call this option 1.

 

I like to present extra options when I think the bosses are likely to balk at my first choice.  So as a second option, maybe take the top 5-10 customers (or however many you need to account for the top 50% of your sales) and make a list of all of their testing requirements.  Make it standard practice to test for all of their requirements, and correlate those results into a standardized COA.  Sure, the COA will list things that some customers won't need, but it makes more sense to me than to create customer specific COA's every single time.

 

Thanks for the feedback.  What both you and Scampi presented make sense.  As to who creates COA's here; this company has only ever presented the customer with COA's directly from the 3rd party lab. They have never been created internally.  FS reviews them and takes action where/when needed regarding the hold/release process.

 

Customers don't have requirements as to who it comes from, but are accustomed to receiving the third party COA.  It seems this became the standard as there can easily be anywhere from 100-300 COA's per week. Due to the volume, I don't think it would be likely that we'd switch over to internal issued COA's for all results, just for certain situations.  Creating a standard set of tests as you suggested, seems to make the most sense in our scenario.

 

I appreciate your insight and look forward to presenting some options to the team here in the future. 

[Charles.C]

Please do not double post topics. Produces confusion and wastes people's time.

[lkosler]

Interesting - I understand what you are going through.  Up until about 3 years ago we were doing what you describe for our customer base.  However, it becomes very cumbersome and expensive to customize all of the variations of the COA.  We eventually decided to pare down the COA and only report the mandatory specifications from the Food and Chemical Codex monograph.  We explained to the customers that the additional testing we provided was expensive and slowing down the production of the material.  The product we sell is used in food, animal feed, pharma, cosmetics and has wide ranging uses in industrial products.  There was a plethora of tests that companies wanted.  We were doing all of these tests to fit into each arena.  The FCC monograph is what we decided the conformance testing would follow. 

 

Most of the pushback from the proposed change came from the sales team as they try to convince everyone that the only way the customer will take the material is if we test everything for the customer.  That is not true - the customers still need the material and will figure out how to test for what they need or adjust their formulations for something else.  Thankfully the product we provide is a low cost commodity compared to other ingredients that could be used instead of it. 

 

COA's - customers will need to be notified about the change in COA and this is not an uncommon occurrence in the food grade industry.  As long as you are notifying the customer of the change and then when the change occurs include the letter with the COA for a month or two as a reminder.  Some customers pushed back but had to figure out the best way for them to cope with the change.  

[AJL]

Just my two cents, our lab had an online registration system for samples - you could track each and every step after registering electronically. Helps to avoid errors and streamline. You could customize a group of tests and call that 'customer A' for example. So you just needed to tick off once when choosing which tests to send for. If you can find a lab that does this, that will help for a start. And some training for your staff.... good luck!!!