My top 3 favorite mental models are incentives, inversion, and second order thinking. I've found these three to be my holy grail - the most versatile and useful in my work. Together, they form a powerful framework: incentives help me understand what drives behavior and how to influence behavioral decisions, inversion reveals blind spots by considering what could go wrong, and second-order thinking maps out the cascading consequences of any decision.
Less reactivity, and more logic. Music to my ears.
Second-order thinking is often treated as a buzzword in tech circles, with many AI-powered startups claiming it as a core value, but most people are actually pretty bad at utilizing it effectively. They focus on immediate metrics or make decisions based purely on best-case scenarios, missing the ripple effects that follow. Many also make super reactive decisions instead of playing the strategic and slower long-game.
Tisk tisk.
So here's why I love second-order thinking so much and how to actually use it.
What is Second-Order Thinking
Second-order thinking is simply asking the question "and then what?" It's thinking long-term, into the unknown and mysterious future with a deeper understanding of the consequences of the consequences. This kind of thinking is really important, especially when making big decisions as this allows you to identify and mitigate risks before they become real-life problems.
This forward thinking engages your prefrontal cortex - the brain region responsible for planning and decision-making. Research suggests that when you mentally rehearse different scenarios ahead of time, you're better equipped to handle challenges when they arise. Your brain has already considered various possibilities, so you're more likely to respond thoughtfully rather than reactively. This can lead to less amygdala activation (your fight-or-flight response) during actual crises, helping you stay calmer and think more clearly when things don't go according to plan.
Second-order thinking requires you to understand why things are the way they are in their current state before making changes to fully understand what all of the potential short and long-term consequences may be. It's also a great way to slow your thinking down to keep your reactive decision-making tendencies in check.
Why is it useful?
Understanding the concept is one thing, but applying it systematically is where the real value lies. I employ this mental model quite frequently in my work building early-stage startup operations. It's particularly important when I'm coming into a situation where there is a lot of chaos that I need to tame or messy processes that need cleaning up.
Here's my framework:
Understand the current state: I won't tear any fence down before understanding why it's there in the first place. This will allow me to understand the original desired outcome and requirements, which informs what I'm going to build to make it better or tear it down completely.
Map out my options: I like using a decision journal template from Farnam Street for this exercise, and for more complex problems or processes, I'll use Miro to make a mind map and really get deep into brainstorming.
Evaluate my options: Elaborate on each option in terms of the potential risks and consequences. This is where true second-order thinking comes into play, and even past that, 3rd, 4th, etc. order thinking. I want to answer "and then what" until I can't answer it further.
Pick the best option: Self-explanatory. Highly recommend having your own decision-making framework to help in this step. Once you have it, the more you do it, the more effective and efficient it gets.
Here’s a concrete example from my work: I was running multiple functions at an early-stage startup, and every team was submitting all of their requests directly to me - everything from IT issues to HR requests to process questions.
I had become my own worst nightmare - the single point of failure, the bottleneck, the human speed bump.
First-order thinking says: "Just handle requests faster or tell people to figure it out themselves." Quick and dirty and makes my life a tiny bit easier in the immediate future.
But second-order thinking asks: "And then what?" If I just work faster, I'll burn out and the bottleneck will get worse. If I tell people to handle things themselves, what happens? Critical requests get ignored, teams duplicate efforts solving the same problems, and inconsistent solutions create bigger operational issues down the line. Then what? Productivity drops, frustration builds, and I end up spending even more time fixing the chaos.
Instead, I designed and built a help desk system in JIRA that categorized requests by type and automatically routed them to the right people or resources. The tool was easy to train our small team on and had an intuitive, easy to understand interface. Non-urgent requests got self-service solutions, while truly urgent items still came to me but with proper context and prioritization. This solution emerged only by thinking through what would actually happen if I took the obvious first-order approaches. Sure, it took more work and time upfront, but was one of the best solutions I built and was even able to use the framework at multiple other startups, making my life easier even outside of that role.
The best solution usually requires a bit more time and work, but will almost always have greater long-term benefits. Short-term gain usually ends up in long-term pain.
Other Use Cases
HR Investigations (and just HR in general): When an employee violates company policy, first-order thinking focuses on immediate discipline. Second-order thinking asks: How will this decision affect team morale? Will it encourage or discourage reporting future issues? What precedent does this set? For instance, being too lenient might signal that rules don't matter, leading to more violations. But being too harsh without context might create a culture of fear where people hide mistakes instead of learning from them.
Building a company: It's impossible to predict the future, which makes predicting success of a new business idea fundamentally uncertain. Second-order thinking, however, does make it easier to think through your strategic moves and paint a somewhat clearer picture of your outcomes. If you introduce a new feature, how will your users respond? What does x action do to your brand perception long-term?
Mass layoffs or AI replacement: A real-world example we see constantly is companies replacing people in droves with AI or conducting large layoffs to cut costs. First-order thinking says this saves money immediately - lower payroll, reduced benefits, streamlined operations. But second-order thinking reveals the hidden costs. And then what? Remaining employees become overworked and stressed. And then what? Quality drops, burnout increases, and institutional knowledge walks out the door. And then what? When business picks back up, you need to rehire and retrain new people at higher market rates. And then what? The damage to company culture and employee trust means top talent leaves, recruitment becomes harder, and you've created a cycle where you're constantly rebuilding your team instead of building on existing expertise.
Data Privacy and Security: I like doing an exercise in live workshops with teams to understand the larger impacts of simple mistakes, such as falling for phishing. We start with clicking a malicious link, then ask "and then what?" The malware installs. And then what? It accesses your email. And then what? It finds customer data. And then what? Data gets stolen and sold. And then what? Regulatory fines, customer lawsuits, reputation damage. One click of the mouse suddenly becomes a company-threatening event.
The Bottom Line
Second-order thinking works on it’s own, and even better when combined with other mental models, such as understanding incentives (what motivates the key players?) and inversion (what could go wrong?). This trio has become indispensable in my work because they force me to see the complete picture rather than just the immediate problem in front of me.
The next time you face a significant decision, try this: write down your initial solution, then ask "and then what?" at least three times. You might be surprised by what you discover - and what disasters you avoid. Slowing down your thinking will not only improve your outcomes, but train your brain to make less reactive and more thoughtful decisions.