Amazon forces sellers to divulge COGS, won't reimburse other costs

Suppose there are 100 suppliers. Or even just two that aren't colluding with one another. The retailer goes to each of them and plays them against each other. They all want the contract as long as they can eek out at least a small margin and the winner is the one willing to offer the lowest price.

What are the possible ways that this doesn't result in the winner having a razor-thin margin whether you know their costs or not?

Possibility one, there is only one supplier or there are a small number who are colluding. Knowing their costs doesn't help because you have no other means to get the product than paying the asking price, since there are no non-colluding suppliers.

Possibility two, one of the suppliers has lower costs than the others. Their costs are $5 when the next lowest is $20, so they can bid $20, have a 75% margin and nobody can beat it because at $20 the competition's margins are already zero. But this is the same as the first one. Even if you know their costs are $5, there is no one else who will give you the product for less than $20 and then they don't have to do it either.

> And I’d defer to the person who did this work, it sounds like successfully, versus someone whose opinion is very broad and I’m not sure what evidence or experience they are basing their commentary on.

Management at different companies has all kinds of opinions on how to do something. Just because somebody did something somewhere doesn't mean it's a useful strategy, only that a manager somewhere thought it would be.

The right answer here is to have the person claiming inside knowledge describe the bad assumptions and why they weren't true in that particular case. Because that's how you distinguish that case from the one where a bad manager is just assigning arbitrary work.

More than that, you need that information to do something useful with it anyway.

Suppose the answer is that they're doing the research to see if they should enter the market for manufacturing the product themselves. Then this works (a) only in the case where the existing market is uncompetitive because otherwise the existing suppliers already have razor-thin margins and there is nothing to extract and (b) only if you're big enough to justify building an entire factory and buying out all of its output yourself. But under those circumstances knowing the cost of production could get colluding suppliers to cut you a better deal to prevent you from entering the market. Moreover, the cost is further justified in that case because if they won't cut you a deal you'd have had to do that research to start manufacturing it yourself anyway.

Without knowing that context the information is misleading. Because (a) implies that anyone doing it are actually the good guys because the counterparty is a price-fixing cartel or monopolist, and (b) implies that it's a waste of resources for smaller entities to pay to do the calculation when they can't enter the market themselves or credibly threaten to in order to get the deal. And even if you are big enough, you would still have to know that making that threat is what you have to do with the information.

So just saying they used it is useless until they tell you how, and then you can assess their claims by reasoning about what they tell you.

> We can't just take general principles and assume they hold true in specific circumstances - that is what makes this kind of work so hard.

General principles apply unless there is an exception. But if someone is claiming there is an exception, the first thing they should be expected to do is spell out how the exception works.