Take a moment to muse on the following thought experiment:

Yesterday there was 10 stacks of Whiptail for 60g/stack and 5 stacks of Cinderbloom for 55g/stack. You bought them all for 875g and made 100 Mythical Mana Potions using all of the herbs, with the intention of selling them for 25g apiece. Today, the price of Whiptail is 110g/stack and Cinderbloom is 90g/stack.

Question: Is there a difference in profit between a sold Mana Potion yesterday and a sold Mana Potion today?

I find this thought experiment fascinating because A) it’s based on true events, and B) my answer oscillates wildly even when I bring in no new information. To save you the time of mathing it out, Mana Potions made at yesterday’s herb prices cost 8.75g each to create while potions at today’s prices cost 15.5g. So if I sold 10 yesterday and 10 today at 25g apiece, how much profit did I make? 308g (16.25 * 20 – 5%) or 247g (16.25 * 10 + 9.5 * 10 – 5%)? What would be your answer?

The answer you choose largely has to do with how you view investing and opportunity cost. I think a lot of people would choose 308g because it “makes sense” in a static investment scenario, especially if you made all the potions yesterday. What does it matter the price of herbs today if you used them all up the day before? Well… it matters quite a lot, actually. If I had resold the herbs the next day instead of making potions, I would have gained 6.75g worth of profit per “potion” compared with the 9.5g profit of today. In other words, 70% of the profit for 0% of the effort (since you don’t even need Alchemy to warehouse herbs for a day), and probably 0% of the risk (the market for herbs > market for mana potions).

Passing on the cost to the customer frequently does not work because, rightfully so, the customer doesn’t care about your costs. I sold 20 potions at 25g apiece over two days; what possible expectation do I have that even 20 more will sell at 31.75g? Well, I could be optimistic and perhaps post during server raid days, relying on cyclical desperation.

One of the most fascinating Wikipedia pages I have come across is the one on Sunk Costs. It is fascinating precisely because it does not make any sense… until it does… and then the sense that it makes slips your mind even as you are thinking about it. No, seriously, go read it.

Buying cheap herbs are not considered “sunk costs” because the herbs can be resold. Once the potions are made however, the herbs can be considered sunk as you cannot un-potion them. I bring it up at all because, conceptually, I largely consider my own investments as sunk costs as soon as I make them. In terms of the thought experiment, then, my answer is (usually) the surprise third one: 500g.

This is not to say that opportunity cost is zero or that I do not make strategic purchases with an eye on profit margins; obviously I care about these things. Rather, I endeavor to not let past investments (unduly) influence future investment and/or the prices I set today*. This mindset is made easier by default as I have no real way of tracking the returns of each individual gold piece (there are addons which will do this though) invested. All I have is Auditor, which keeps the bottom line in plain view – as long as the number keeps going up, I am content with letting experiments like making Mythical Mana Potions potentially run at a loss. It would be miles easier (and more profitable) to have simply milled and inked those 15 stacks of herbs, for example.

It would have been easier, but way less fun.

*Incidentally, that is precisely the reason I stopped using Auctioneer and it’s GetAll and historical pricing features: they are largely useless trivia. Obsidium Ore went for 75g/stack two weeks ago? Who cares? It is either a good price today or it isn’t. What is a “good price?” Can you buy it now and turn it into something that will sell for more now? Good price! Things get fuzzier once the time horizon shifts further back, like buying the Bracer Enchant patterns in anticipation for cheaper Maelstrom Crystals in 4.1 (and subsequent demand spikes). Then again, how is the historical price supposed to help you there?