Good leaders do not let organizations become directionless, paralyzed by indecision, or crippled through inaction. Good leaders are decisive.
But whilst even the best leaders do not always make the right decisions, they have frameworks that enable them to address and reduce uncertainty. Rather than delay hard calls unnecessarily or procrastinate in the face of difficult choices, good leaders increase their odds of making the right decisions by developing frameworks for:
- Gathering information. They ask the right questions that enable them to contextualize the choices they facing. The right questions are rarely obvious. And there is often a temptation only to use what data is already easily to hand. In practice, the right questions often demand new data or research, which is rarely costless to obtain.
- Understanding the shape of the distribution of likely outcomes. They consider the worst and best case outcomes, the probability of each, and the modally likely outcomes that are liable to result from the choices they make.
- Factoring in opportunity cost. What next best alternative is lost through action or inaction? Rather than model strategies in isolation, good leaders compare and contrast the potential impacts of various different courses of action.
- Accounting across the board. Rather than merely consider the direct financial impact, the very best leaders consider actions’ knock-on effects on human, social, and political capital as well. Embedding financial models within meta-models enables accounting for reputational and other non-pecuniary forms of capital side-by-side.
In short, it takes tremendous mental processing power to be a good leader. Informal mental modeling can only get leaders so far.
Recognizing that humans are boundedly rational, HASH is an open-source platform for rich descriptive simulation. Our work at HASH builds atop several key principles:
- real-time connectivity and modern analytics should result in rich information being readily available across the entire universe of domains within which modern enterprises operate;
- complex systems can be modeled more accurately when broken down into component parts and sub-systems;
- component parts and sub-systems may be reusable across simulations;
- abstraction on the frontend can help overcome cognitive barriers to processing information, and need not mean simplicity on the backend or a reduction in detail of an underlying model.