Because if the most important day at your company is someone’s first, then the second most important is their last.
My favorite scene in Moneyball is Brad Pitt teaching Jonah Hill how to fire a player. Not only does it give some good practical advice, but it also shows just how hard this conversation actually is; Hill’s character is entirely well-intentioned but still manages to bungle it. Firing is a skill worth learning to do correctly.
At Oceans, before anyone does a termination (whether for performance or gross negligence), they practice that call with me. We start by doing a roleplay, with the manager telling me that today will be my last day. If they start to go off track, I immediately stop it; you don’t want people to practice doing the wrong thing and then correct, because every repetition makes it harder to shake.
If I need to stop them, we discuss the right approach and then flip: I become the manager and model the right way to have the conversation. Then we flip again and the original manager practices until we get it right, which may require more flips.
And that’s OK. If you’re doing things right, terminations are the worst kind of event: rare and important. Rare and unimportant conversations are easy to ad-lib, frequent and important are easy to routinize.
As Brad Pitt says, the best way to fire someone is to deliver the minimum necessary information, while leaving the door open for further conversation. You can’t script it, because it comes off as inhumane, so you have to understand the basic building blocks and then be willing to adjust.
Start with the essential context setting: this is going to be a difficult conversation, you want to be respectful, you’re going to start with the most important facts. At Oceans, those facts are: today is your last day, we cannot confidentially keep you in this business role, what type of termination it is (this matter because gross negligence means you are ineligible to rejoin the company, while performance termination does allow you to return in a year and we’ve had several people do so), and that HR will be in touch with logistic details.
And that’s it. Open the floor for questions and hold space, but don’t force more than that; the essentials have been delivered.
Never impose feedback. Often, we feel the need to explain to people what they did wrong that resulted in the termination. But that is actually for our benefit: it makes us feel better and justified. Assuming that yours is a desirable workplace, you are delivering bad news to someone and even if they are eventually interested in feedback, they may not be in a place where they can hear it at the same time as processing that they no longer have a job.
At the same time, if they want feedback, you have to be ready to give it to them. Use behavioral examples and always tie them back to “and thus we cannot confidently have you in [business function] right now”; this helps depersonalize and sets the context as a business decision. If they don’t, you can offer to have a separate feedback session at a later time if they want.
If they try to give you feedback, redirect those to a separate exit interview; at Oceans, managers do the terminations but everyone has an exit interview with me where I gather feedback on how we could do better. You want this to be about them, not the company, but explicitly recognize that they will have an opportunity to share their thoughts.
Finally, try to end it on a positive and personal note. Because of the peak-end rule, the person’s subjective impression of the experience will be defined by the worst moment and the last one. At Oceans, that can include things like “Speaking on a personal level, I’d love to see you back here again in a year” or “I really appreciated [specific example] and I hope you continue that going forward.”
Firing is hard. It is worth practicing and improving. Ten years from now, I’m sure I’ll think about it differently than I do now. And that’s a good thing; of all the places to avoid foolish consistency, the second-most-important day is certainly one.
Michael Hallsworth reflects on the enduring success of the EAST framework, which encourages practitioners to make behaviors Easy, Attractive, Social, and Timely. Revisiting the framework a decade after its release, Hallsworth asks why such a simple tool has remained widely used despite a field that has grown more sophisticated, data-driven, and complex. His answer is that simplicity itself can be a strength—especially in applied behavioral science, where practitioners often face real-world constraints, limited time, and uncertain contexts.
Hallsworth contrasts this with a growing push in the field toward greater precision and complexity—using techniques like machine learning, adaptive trials, and highly detailed behavioral models. While these approaches aim to capture the messy realities of human behavior, they can also create problems. One risk is fragmentation: researchers constantly develop new frameworks and measures that few others adopt. Another is discouragement: overly complex models with dozens of variables can overwhelm practitioners who need tools that are practical and usable in everyday decision-making.
Instead, Hallsworth argues that simple frameworks can function as heuristics for applying behavioral insights, helping practitioners navigate uncertainty without getting lost in complexity. Acronym-based tools like EAST are memorable and accessible, allowing people to “muddle through” real problems while still applying behavioral science principles. Importantly, simplicity does not mean simplistic thinking. When designed well, simple tools can support sophisticated applications, much like well-designed objects that achieve versatility through minimalism.
Ultimately, Hallsworth suggests that effective behavioral science depends not only on better techniques but also on practical wisdom—the ability to make informed judgments about context, trade-offs, and what to exclude. Drawing on the concept of phronesis, he argues that experts must exercise judgment when applying frameworks and interpreting evidence. As the field evolves—especially with technologies like generative AI—this human capacity to evaluate, select, and apply simple tools wisely may become even more important than developing ever more complex models.
👉 Read the full piece here
You’re not “just” a behavioral scientist — you’re someone who knows human behavior is a competitive advantage. Here are 5 companies hiring people like you this week 👇
User Researcher - Our Future Health
Senior Researcher, Product Experience - Netflix
Human Factors Researcher - Apple
AI-Enabled Insights Lead - Airbnb
User Experience Research Senior Advisor - The Signa Group
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.