Ballistix

Cost accounting has something interesting in common with the prisoner's dilemma. The optimal strategy for the prisoner's dilemma depends on whether you play the game once (defect) or multiple times (cooperate). When it comes to management decisions, cost allocation makes sense if used for a single decision, but it dramatically reduces the organization's profitability when used across multiple decisions. This is because the impact of a decision on indirect costs is very likely to be bimodal. The decision will either cause no increment in indirect costs or trigger a considerable increase. Now, if you consider these two outcomes, the first is much more probable (most organizations carry much more excess capacity than they care to admit), but the second will have a much larger (negative) impact on profitability. Cost accounting (as it's commonly used) is weighted toward the second outcome. For example, it makes it much less likely that a manager would say "yes" to a request for a volume discount. This decision, in isolation, is defensible. But, imagine what happens if managers process 100 similar requests every month and refuse all of them. We now have an organization that is stagnant and that, in due course, will almost certainly become unprofitable. I'm sad to say that this is not just a thought experiment. We have worked with a number of organizations that make decisions like this and then build sales teams, hoping that salespeople will be the solution to their problem! For more, read my latest mini-book. https://a.co/d/2YXqeZp

The conflict of visions between DOGE and federal government departments is an even more extreme version of the conflict that ensued after Musk acquired Twitter. The key point is that there are two visions of how organizations should be designed and how they should be managed. There's conflict because folks with one vision have a tendency to regard those with the other as primatives! My new book (An Ode to Speed) introduces the "operating system" that hyper-growth organizations run. The central difference is that traditional organizations are optimized for cost (or efficiency), whereas hyper-growth organizations are optimized for speed. This is not a trivial distinction. The alternative optimization parameter results in a radically different approach to management decisions (an approach that conflicts with cost accounting) and in a different mental model of the organization that lives in the heads of managers. My book (released just days ago): https://a.co/d/8bX9Otz

The first phase of the transition to a speed-based operating system is to change the focus of every department from cost to speed. You need every manager to maximize the speed at which value is generated within their department (while, of course, exercising reasonable control over costs). This requires that every department move to a primary time-based metric. It also requires that work is packetized so that the rate of value generation can be readily measured. This is a radical change to most departments' mode of operation. With an efficiency focus, managers will attempt to keep their departments consistently busy. But when the focus shifts to speed, managers will need their teams to sprint when work appears and then relax when it’s done.

Grok 3 is breathtakingly good. I just asked this question. "Does the word calculating still apply if you're making a prediction? For example, there's a difference between calculating how much profit I made last month and estimating how much I will make next month." The answer starts with. You’re digging into a subtle but interesting distinction here, and it’s worth unpacking. The word "calculating" can apply to both scenarios—past profit and future estimates—but its fit depends on context, precision, and intent. Let’s break it down. The rest of the answer is exhaustive and impeccably reasoned. The conclusion is. Yes, "calculating" can apply to predictions if you’re using a systematic, numerical process—it’s not just for settled figures. But if you want to signal uncertainty, "estimating" or "forecasting" might dodge the impression of false precision. In everyday use, though, people won’t bat an eye at "calculate" for either. Does that settle it for you, or want to poke at it more? Try it for yourself!