I work for Ba Feers, which you probably know is the largest diamond miner in the world; our revenues are high enough that a small change in price leads to a big change in the $ value on the bottom line. We do change prices monthly, but there are a lot of different constraints on our business that make pricing at Ba Feers a lot more challenging than pricing at a "widget factory" in the textbooks. First of all, we're the overwhelmingly dominant player in the industry, which means we can't take prices from the market. There simply isn't a market.
Secondly, our clients are incredibly reluctant to reduce the quantity of diamonds that they buy, partly because they have orders to fulfil that they don't want to lose, and partly because they don't want to lose privileged access to Ba Feers's goods (which they believe they risk losing if they don't maintain high purchases from month to month). They would rather take a hit to their bottom line themselves than refuse goods, but we want them to make normal economic returns, in order that they can continue to buy from us in the future, make investments (and therefore have more money to spend on our diamonds), and engage in marketing themselves (a chore that Ba Feers has historically taken on itself, rather famously).
Thirdly, it's really hard to switch a mine on and off; it takes up to 6 months to get diamonds from the mine to the client, so we have a strong imperative to maintain sales volumes and respond to changes in demand by changing prices.
And finally, our mines come predominately from Botswana, whose government has a 50% shareholding in the mining operation there. The Botswana government wants stable revenues, because the gov'ts revenues are highly dependent on diamonds, so they don't want to have to change their public spending plans. Although we control our own prices, they have a big influence over our pricing decisions.
All of these different constraints make it difficult to get pricing right. We can't just raise prices until we lose sales, and then figure out by trial and error where the optimum price is, because our clients just take the hit themselves. Instead, I have financial models of a generic manufacturing company that try to predict what margin this generic company would achieve on our goods, and plan prices backwards from that, so that they receive normal economic returns. We can't really do "cost-plus" pricing, because in cost-plus, you set your price as a fixed number, and then change volume in response to changes in demand. But we can't do that because it would either mean running higher stock levels (which means higher capital requirements => lower return on capital, and also if we drop prices, we'd have to write down our stock), or turning our mines on and off (which is physically impossible).
So we have to kind of "guess" what our prices should be, which is what I try to do. I already mentioned the financial models of a generic manufacturer, but legal restraints (we're a big bad monopoly!) forbid us from actually asking for key details from clients that allow us to do that. We also have only a vague idea of what our goods are manufactured into, and how much they sell them for in the end, which obviously affects the manufacturer's profitability; I use a statistical model to predict that, too. We also auction a small amount of our goods; I analyse the results of those auctions (or rather, the bidding behaviour of clients during the auction) to extrapolate prices for the bulk of our goods.
All this stuff sounds like it should take a lot of time, but it doesn't really. The models, once set up, run themselves. I spend most of my time figuring out how to make better use of the existing data, where to find new, more reliable sources of data, whether there are other indicators of future short-term demand that we haven't thought of, and so on. But because we don't know precisely what our profit maximising price should be, have basically none of the normal feedback mechanisms for finding it out, and have an incredibly strong imperative to maintain sales volumes, that offers a lot of scope for increasing profit through more accurate pricing.
Now, those are the commercial issues. However, I haven't even gone into the technical issues... You see, even if we did know exactly what price we should put on our goods, there are a myriad technical and logistical issues preventing us from doing that. It's far too boring to go into, but very often, it's almost impossible to sell our goods at exactly the prices that we want to sell them for. You can get them pretty close; how close you get them depends on your ingenuity and skill with the tools you have available. And a lot of my time is spent improving the tools and creating new tools to help us get closer to the price that we want to set our goods at. The closer you get them to the desired price, of course, the more money the company makes, so the creation of those tools, and my ability to use them, directly impact the bottom line. Of course, there are process enhancements in all parts of all businesses, but as I said, because I'm in a role with such a direct impact on the bottom line, those process enhancements go a lot further than they would in some other part of the business.