Data and the Capital Cycle: How Koch Became an Empire - Commoncog

How Koch Industries became an empire. We draw on ideas from both the Becoming Data Driven series and the Capital Expertise series.

This is a companion discussion topic for the original entry at

Very much enjoyed the layout of these posts, combining the cases with the general newsletter. In particular the ability to jump out to the original cases was an excellent experience, even on mobile.

On the case itself. One point that caught my eye is the importance of operational efficiency during challenging market cycles.

Having strong operations gives you the strategic breathing space to do fun stuff. Something I think is missed when people discuss moats, and how you form them.

It reminds me of the Lessons From Titans book, where they made the claim that operational control was critical for industrial Titans to succeed over the long term. Very unlike the Porter idea that operations can never be the source of competitive advantage.

Need to think more about this to be sure of what I think.


This was a really interesting observation. Of course, I’m biased to agree with you, but let me think out loud a bit.

Apparently in Amazon Bezos liked to go around saying “don’t make the easy things hard” — especially when it came to things like instrumentation and access to operational metrics. (Presumably he picked this up from the ex-manufacturing, ex-six-sigma folk who joined in the early 2000s). But I’ve always attributed Amazon’s focus on operational rigour to its low margin business model.

Does anyone here know of a (presumably high margin) software company with high operational rigour?

Off the top of my head, I’d bet Netflix:

  • They have released a few reliability tools, e.g. Hysterix
  • They invented (or at least popularized) Chaos Engineering
  • Their tech blog covers things like 0-downtime cross-regional failures
  • They famously run on AWS. Being the size they are, they’ve almost certainly rubbed shoulders with some of AWS’ talent, who may well have passed on some sage advice.
  • I assume they have decent margins? They’re used as an example of scale economies shared in Helmer’s 7 Powers. That may be changing with competition from Disney+, Hulu, Apple TV, Prime Video, etc.

Outside of “technical” operations, I’d be making a guess. I don’t have any insight into how they measure and manage their business operations.

Google is also in play for technical ops, since they literally wrote the book on Site Reliability Engineering and helped popularize Objectives and Key Results (OKRs). OKRs are another version/refinement of management by objective. Their ads business seems to be high-margin, if memory serves.

Which leads me to another guess: maybe Intel? They have to run a large foundries without losing money, but they also produce a lot of software. A lot of their hardware needs firmware to work, SSDs need wear-levelling algorithms, graphics cards need drivers, etc. Since Grove wrote High Output Management, which includes a fair few operations tips, I think some probably survived there. Biggest drawback – maybe not what you’d call a “software” company, and I don’t know how good their margins are.

And my final guess is DPD, the parcel company. Again, not your traditional software business, but they use a lot of software in their hubs. Most notably, their Predict service, which gives customers a 1-hour time window in which they’ll deliver your parcel. DPD is consistently ranked as one of the best parcel companies in the UK, and no-one has replicated it yet. I do not know how good their margins are, so … possible candidate? Disclosure: I did some contract software work for DPD “back in the day.”

Edit: Motorola! They invented Six Sigma in the 80s. Like Intel they produce hardware and the software for that hardware, so they’re sort-of a software company. No idea if they’re high-margin, or if they were during their hey-day.