Mental models I use for optimisation: Engines

For most situations, the first and most fundamental question asked has always been: What should I be optimizing for?

The next question asked is: How should I be optimizing for *this*?

It is not enough to simply say I have access to *this*, or I am slowly developing access to *this* as a function of time. There is a more strategic and structured approach to growing *this*.

Engines

This approach is developing an Engine. An Engine is something that works for you to achieve the core essence of a goal akin to an automated flywheel. An Engine should be something that can be activated, or de-activated according to demand.

An Engine can be broken down into the following constituents:

• Inbound channels
• Outbound channels
• Resources to sustain its operations
• Processes to sustain its operations

What does the practical end-state look like?

In the context of the work I do, it appears to me that there are 5 main levers to optimize for, which I believe are also generally applicable.

These levers are:

Information | Processes |Network | Brand | Capital

Let us run through a practical situation. In the context of one of the 5 optimization levers above, information, the following questions arise:

• What is my Engine for inbound information?
• How are the different stakeholders in my Engine incentivized?

Scott Adams, the author of Dilbert, makes a similar point in his article, “Goals vs. Systems”. However, as anything can be considered a system, it is important to go deeper. Taking the specific example he cites, his system of blogging daily in order to one day be able to depend on writing as a source of living, we see that he mentions the R&D process for experimenting with different voices (in this semantic, “Process optimization”), and how it leads Scott to get speaking requests and Wall Street Journal articles.

Scott’s concept of the system can be refined further; How exactly did these request(s) come in? What was the efficiency of the different channels? Was he too focused on only inbound requests?

When it comes to Engines, it is too broad to judge it as a whole (Imagine Scott trying to understand the efficiency of all of his blogging effort). Instead, by understanding that a system itself also has underlying levers, we are in a better position to understand its deficiencies and strengths, and optimize accordingly.

Think of an Engine like any other system. We can take an actual car engine for example. There are inputs, outputs, resources to sustain its form, and a design or process to ensure that all its constituent mechanisms work together, ideally in an automated fashion. I personally think of Engines in a similar way to e-commerce marketing funnels.

There is a great scene in the TV show, Billions, where the main antagonist asks: ‘Who is Bobby Axelrod?’.

Bobby Axelrod is the main protagonist, a billionaire hedge fund manager who outmaneuvers his competition with ease.

A line stands out:

“When you are Bobby Axelrod, you don’t chase information. Information flows to you.”

The focus on inbound information flow, and the subsequent explanation of how Bobby Axelrod got access to what he wanted (a private concert) reveals the focus on an Engine that is independent to a specific time-pressured goal.

Axelrod planted seeds for information flow (See above for Process -> Incentive Alignment) to achieve the core essence of a goal, presumably happiness through rock concerts.

A Plan can be irrelevant. Engines are a foundation.

An Engine is a directional foundation for resource gathering. Plans, in contrast, tend to have a binary outcome. You succeed or you fail in the achievement of your goal. A more focused Plan is frequently used and required on top of an Engine.

An Engine allows for goalposts to shift with changing externalities, allowing you to adapt to fluid situations.

Especially if you are in a state of affairs where your goals are suitably macro, or not well-defined, a focus on an Engine is a more logical choice.

The opportunity for cross-disciplinary resource sharing or synergistic movements between the various engines above is also a topic I enjoy diving deeply into, for another time.

Thanks to Bjorn Lindfors, Alin Dobrea and Annette Lim for their comments/edits.