To understand your business, you have to understand your costs--at a total, average, and incremental level. Many small businesses that would otherwise be successful fail for lack of this understanding. Think of a restaurant that clearly has a sufficient level of customers but that nonetheless closes its doors permanently.
To be sure, many businesses manage to stay alive and in some cases thrive despite anyone understanding their fundamentals, but these examples are rare and fleeting. More often what seems to be operators flying blind are actually people with some combination of magnificent instincts and great muscle memory honed by years at the treacherous helm.
Truly understanding the underlying drivers of costs unlocks the ability to guide all manner of vital decisions: what to make and how much, how to shrink when necessary, where you can discount and where you cannot, where to expand operations and what to expect from growth, etc.
Cost per the relevant unit is as essential as it is boring to all but a few of us, an elite group who relish mastering the concept. The good news is there are only two difficult parts to that equation, but the bad news is the same.
Cost is an elusive concept. Defining cost properly is where economics meets accounting. Here we must understand variable versus fixed costs, marginal versus average versus total costs, the thresholds of cost expansion/contraction (At what point of production must we build an entirely new production plant? If we shutdown a production line, how are costs affected?), cost drivers and probable variances, etc.
Defining the unit(s) that is relevant is the art of cost accounting. For example in hotels it starts with the units for sale as seen in the revenue metric REVPAR (revenue per available room). While there are other important metrics beyond that one in that particular industry, the available room is foundational since that is the essence of what a hotel is selling.*
In defining the relevant units we must understand if the units are variable or fixed, if the units are fully or partially or not at all under our control, how the units behave independent of cost (think of seasonality, etc.), when certain costs (or revenues) do and do not apply to various units. In the last case consider a restaurant where the cost applied per table seat available might be separated from the cost applied per bar seat available even though we would still want to look at them in totality. Hence a mythical unit might be created to synthetically mimic the real units on an aggregate basis--e.g., cost per customer spot available.
Not complicated enough? Add in a dimension of time. Open a restaurant an hour longer each night--do your relevant units change? Cost certainly will. Depends on how you define units given the objective you're trying to actually measure.
From cost we next need to know revenue from which point we can understand profit. Allow me to illustrate using a personal anecdote from my time as a financial analyst at a newspaper.
One of the reasons I was hired was to understand the cost and revenue drivers as profit margins had begun shrinking in the industry. I like to say that a fat profit margin hides a lot of bad decisions. The newspaper industry was no exception. Quite a few things were able to be tried that once a thorough analysis was conducted turned out to not be as successful as expected or believed to be. This isn't a bad indictment per se. Success in business is built on a mountain of well-placed failures.
Of all the things I was asked to do, I was never explicitly asked to determine the specific attribution of the company's profits--meaning what lines of business were profitable. Yet this seemed a natural thing to want to know. In fact it fascinated me. To everyone else it was obvious or uninteresting. They simply "knew" what was profitable. Profits were so big, heck, everything was profitable, right? That was intuitive to some but not to me. My intuition was the opposite--I knew that it would be highly unusual for everything that went into a bundled product to be profitable in the sense of direct attribution.
There seemed to be two biases at work: a fear of knowing the answer (what if my area isn't profitable?) and a lack of critical thinking (look at how much revenue this generates/this is an essential part of the business; it must be profitable).
Once I achieved a strong understanding of the company's cost, it became apparent that everything simply could not be profitable. There were vast differences in revenue by various business lines but very little differences in properly allocated costs. Applying revenue minus cost (i.e., profit) business line by business line would "use up" the profit before all the areas were covered. Before mentioning the areas that were profitable, a caveat is needed. The newspaper was a bundled product meaning everyone got the main section, the sports section, the monthly special sections, the inserts, the classifieds, etc. A point of near religious dogma in the industry was how vital nearly all of these components were to a successful bundle. So I was both risking heresy as well as producing an analysis that the very cost-conscious management team might misinterpret much to its own demise. Loss leaders is a very real and healthy business practice as is cross subsidization. But these concepts can also be co-opted to excuse bad mistakes were money is lost for no actual indirect gain.
Everything wasn't profitable. Out of nearly a hundred different product lines and sublines, only four areas accounted for 100% of the profit of the business: preprint inserts, national ROP ads (when a company like American Airlines ran a full-page newsprint ad), color ink (a big upsale item), and employment ads in the classified section. This meant all the other ads in the main, sports, business, and other newsprint sections including special sections were losing money. All the rest of the classified section outside of employment ads were losing money--ad areas like traditional for-sale listings, automotive ads, real estate ads, etc. The circulation revenue was not covering costs. Subscribers were not paying enough to cover the cost of delivery much less any content production.
The entirety of this analysis was not a complete surprise to the seasoned people at the helm of the paper, but the details were revealing and eye-opening. To repeat and be fair, this did not mean that people like the guys selling automotive ads weren't adding value--they certainly were. But what they were adding was content value much like the guys writing the sports columns. The upshot was that a limited decision-making mantra could have been "How will it help increase preprints, national ROP, color, or employment ads?" If the answer was "it wouldn't", the right decision would be to reject the proposal.
Knowing cost is not easy; so a good deal of respect is owed the business people of the world who work hard to master it. This might start with the cost accountants I was slighting in the prior post, and it certainly ends with all those entrepreneurs, business middle managers, and captains of industry toiling away so cost is never unknown.
*There are always exceptions. Some hotels are selling experiences outside of the room itself. The room might not correspond tightly to variable costs or to revenue. However, usually even when there are multiple lines of business (think Las Vegas hotels), these still are broken down per available room as it corresponds to cost and revenue as tightly as any other unit.