Tongkat capsules vary in weight. How significant is this?

You are always going to have a manufacturing tolerance. No machine is 100% accurate. Even manual filling will have manufacturing variance. It's how to set your targets and ensure that the entire batch meets your specs that matters. The USP acceptance criteria for capsules is based on a +- variance from the average weight. All the capsules have to be within 10% of the mean average gross weight to be approved. For this batch of tongkat ali, the mean average gross weight of the capsules was 269.31mg. This means that all capsules have to fall within the range of 242.38mg and 296.24mg. Our biggest issue with the USP guidelines is that it makes no reference to target, though. This is a big oversight in our eyes, so we have modified the USP guidelines to be more strict. Our guidelines state that the mean average gross weight needs to be at or above target to approve. Not sure why USP didn't include that, because it is a big deal in our eyes. If you think there is 200mg in a capsule, the mean average should be at or above that number. Who cares if all capsules are within 10% of each other if the average is less than target? It makes no sense to us, which is why we modified them to be more strict. The target mean average gross weight for this batch of tongkat is 250mg. Since the mean average gross weight measured at 269.31mg, and every capsule is within 10% of that mean, this batch is within spec.

So you can see that we err on the side of overfilling. We do that with all our manufacturing tolerances. I feel everyone should be getting at least what they expect. If they get slightly higher, that is much better than getting slightly lower. This is how we modify USP guidelines for a lot of things we do. We skew our specs and tolerances to the high side. Almost everyone else in the industry does NOT do this. They set their target at exact label claim, then just accept than half or so will be under. It's a lot more expensive to do it our way, because we have to add more raw material to ensure we are always over target. Sure, we are only talking 5-15% over, but that adds up over the years. Then we also add more capsules into each bottle to ensure nobody is ever short because of a manufacturing fill variance. This adds to the overage cost we build in, too. My ethos is that everyone should always be confident they are getting at least what we claim in each product, with an overage possible within reason. Obviously we are not going to overfill by 50%. However, a 5-15% is within reason in my eyes to ensure people are getting what they expect. I don't think many people here realize how uncommon that is. It's just not done in this industry. In fact, we often find significant under-filling on competitor products that we test. One L-theanine that's a top seller on Amazon literally has 50% of their label claim. Anyone taking that is getting half of what they expect. Does the fact that each capsule is within 10% of that mean matter to customers when they are getting half of what they purchased? That's the problem with some of the regulations in this industry. Technically passing USP capsule variance guidelines means nothing if every capsule is short.

On top of this, we assay the finished capsules to ensure they contain what they are supposed to. Almost nobody does this. Generally, if a capsule batch passes the base level USP capsule variance guidelines, the supplement brand passes that batch and moves on to the next. We actually assay the capsules via HPLC/UPLC to ensure that there is the proper amount of actives in it. A contract manufacturer could just add in more MCC or another filler to get it to meet the USP capsule guidelines. That doesn't mean that there is the amount of tongkat ali in each capsule that you expect. The only way to know for absolute certainty is to assay the capsules and measure the actives themselves. So our capsule approval process not only assesses fill rate and weight variance, but also directly measures the actives inside the capsules to ensure everyone is getting what they expect every time. We just skew that upwards, rather than downwards, to account for manufacturing variances.