Time is the wrong trigger for most behaviors
In L-0425, you learned about time-based triggers — alarms, calendar reminders, scheduled blocks. They work when the behavior genuinely belongs at a fixed moment: take your medication at 8:00 a.m., join the standup at 9:15. But most of the behaviors you want to trigger do not belong to a specific time. They belong to a specific moment in a sequence of events.
Consider: you want to review your priorities before starting reactive work each morning. If you set a 9:00 a.m. alarm, it fires while you are still commuting on some days and interrupts deep focus on others. The alarm is anchored to the clock, but the behavior is anchored to a transition — the moment you arrive at your workspace and are about to begin your day. What you actually need is not a time-based trigger. You need an event-based trigger: "When I sit down and open my laptop, I review my three priorities before opening email."
Event-based triggers fire when a specific, observable event occurs rather than when the clock hits a particular number. They are the most natural and often the most reliable form of trigger you can design, because they synchronize your agents with the actual flow of your day rather than with an arbitrary schedule imposed on top of it.
The science of event-based prospective memory
Prospective memory is the cognitive function responsible for remembering to do something in the future — not remembering facts or past events, but remembering intentions. In 1990, Gilles Einstein and Mark McDaniel published a landmark study that drew a fundamental distinction between two types of prospective memory: time-based and event-based (Einstein and McDaniel, 1990).
Time-based prospective memory requires you to remember to do something at a particular time or after a particular interval has elapsed. Call your dentist at 2:00 p.m. Take the laundry out in thirty minutes. This form of prospective memory is demanding because it requires self-initiated monitoring — you must periodically check the time and compare it against your stored intention. There is no external signal to remind you. You must generate the reminder yourself.
Event-based prospective memory works differently. You form an intention tied to an external event: "When I pass the pharmacy, pick up the prescription." "When I see my colleague at lunch, ask about the project timeline." The critical difference is that the environment itself provides the cue. You do not need to monitor anything. The event occurs, you perceive it, and that perception triggers retrieval of your stored intention.
Einstein and McDaniel's research, and the extensive body of work that followed it, consistently shows that event-based prospective memory is easier and more reliable than time-based prospective memory (Einstein and McDaniel, 2005). The reason is architectural: event-based retrieval can occur relatively spontaneously when you encounter the target event, especially when the event is distinct and the association between the event and the intended action is strong. McDaniel and Einstein called this the reflexive-associative hypothesis — if the link between the cue event and the intended action is well-encoded, encountering the cue reflexively brings the intention to mind without requiring deliberate monitoring.
This is not a minor difference. It is a fundamental asymmetry in how your memory system works. Time-based triggers require you to do cognitive work — monitoring, checking, comparing. Event-based triggers let the environment do the work for you. The event shows up, and your intention comes along with it.
Implementation intentions: the if-then format
In 1999, psychologist Peter Gollwitzer formalized a technique that harnesses event-based prospective memory for behavior change. He called it implementation intentions — pre-decisions about when, where, and how you will act, structured in an if-then format: "If situation X arises, then I will perform behavior Y" (Gollwitzer, 1999).
The distinction between a goal intention and an implementation intention is critical. A goal intention is "I intend to exercise more." An implementation intention is "When I finish my last meeting of the day, I will change into running shoes and go for a twenty-minute run." The goal intention specifies what you want. The implementation intention specifies the event that triggers action and the exact action you will take.
The effect sizes are substantial and well-replicated. A meta-analysis of 94 independent tests found that implementation intentions produced a medium-to-large effect on goal attainment (d = 0.65) compared to goal intentions alone (Gollwitzer and Sheeran, 2006). A more recent and comprehensive meta-analysis covering 642 independent tests confirmed the finding: implementation intentions were effective for cognitive, affective, and behavioral outcomes, with effect sizes ranging from d = 0.27 to d = 0.66, and the effects held across different populations and over long-term follow-ups (Thurn, Wolff, and Gollwitzer, 2024).
Why does the if-then format work so well? Because it exploits the same mechanism that makes event-based prospective memory reliable. When you form an implementation intention, you create a strong mental association between a specific situational cue (the "if") and a specific response (the "then"). This association heightens the accessibility of the cue — you become more likely to notice it when it occurs — and it automates the response, reducing the deliberation required to act. In Gollwitzer's terms, implementation intentions delegate the control of behavior from the self to the environment. The situation triggers the action, not your willpower.
This is exactly what you are doing when you design an event-based trigger for a personal agent. You are forming an implementation intention. The trigger is the "if." The agent's action is the "then." The event-based format makes the trigger reliable because it leverages a retrieval mechanism your brain already handles well.
Habit stacking: chaining new behaviors to existing events
James Clear, in Atomic Habits (2018), popularized a specific application of implementation intentions that he calls habit stacking. The formula is: "After [CURRENT HABIT], I will [NEW HABIT]." Instead of linking a new behavior to a time or a location, you link it to an existing behavior that already occurs reliably in your day.
"After I pour my morning coffee, I will write down my top three priorities." "After I close my laptop at the end of the workday, I will review what I accomplished." "After I sit down for dinner, I will name one thing I am grateful for."
Habit stacking works because your existing habits are already encoded as reliable events in your daily sequence. They happen automatically, without deliberation. When you stack a new behavior on top of an existing one, the existing habit becomes the event-based trigger for the new one. You do not need an alarm. You do not need a reminder app. The completion of one behavior is itself the cue for the next.
BJ Fogg, the Stanford behavior scientist whose research underpins much of Clear's framework, calls the existing behavior an "anchor moment" — a reliable routine behavior that serves as a prompt for a new target behavior (Fogg, 2020). Fogg's Tiny Habits method uses the explicit format: "After I [anchor moment], I will [tiny behavior]." The anchor moment is an event-based trigger, and the tiny behavior is the agent's action. Fogg's key insight is that the anchor must be something you already do with near-perfect consistency. If the anchor is unreliable, the entire chain collapses.
The power of habit stacking is that it transforms your existing behavioral sequence into a trigger infrastructure. Every reliable thing you already do — brushing your teeth, pouring your coffee, locking your front door, opening your laptop, finishing a meal — is a potential event-based trigger waiting to be loaded with an intentional action. You do not need to build triggers from scratch. You already have dozens of them firing every day. You just need to attach agents to them.
Transition moments: where event-based triggers are most powerful
Not all events are equally effective as triggers. The research on behavior change consistently points to one category of event that is disproportionately powerful: transition moments.
A transition moment is the boundary between one activity or context and another. Walking through a doorway. Arriving at a new location. Finishing a phone call. Closing one application and opening another. Sitting down after standing. These are liminal moments — thresholds where you cross from one state of activity to another.
Transition moments are powerful trigger points for three reasons.
First, they are perceptually distinct. Your brain naturally registers context changes. The phenomenon of "doorway forgetting" — where you walk into a room and forget why you came — actually demonstrates how strongly your brain encodes transitions. Research by Gabriel Radvansky and colleagues has shown that passing through a doorway creates an "event boundary" in memory, causing the brain to compartmentalize and partially reset the contents of working memory (Radvansky and Copeland, 2006). The same mechanism that causes doorway forgetting can be harnessed in your favor: if transitions are moments your brain naturally notices and processes as discrete events, they are ideal trigger points for retrieving stored intentions.
Second, transition moments are decision points. When you finish one activity and have not yet started the next, there is a brief window of cognitive openness — a moment where your executive function is momentarily unoccupied and available for redirection. If you have no agent loaded for that transition, your default behavior fills the gap, which usually means checking your phone, scrolling email, or drifting toward whatever is most salient. If you have an event-based trigger loaded, the transition becomes a moment of deliberate action instead of default behavior.
Third, transition moments are observable and discrete. "When I finish the meeting" is a clear, unambiguous event — you can point to the exact second it occurs. This observability is what makes transition moments reliable triggers. They pass the camera test: a neutral observer could see the event happen and confirm that it occurred. Compare this with "when I feel ready to start working," which is a subjective state that has no observable boundary and therefore triggers inconsistently.
The AI parallel: event-driven architecture
If you work with software or are simply curious about how systems are built, the parallel between event-based triggers in behavior design and event-driven architecture in software engineering is exact — and illuminating.
In traditional request-response computing, System A asks System B for something and waits for an answer. This is like time-based triggering: you check in at scheduled intervals to see if anything has changed. It is polling. It works, but it is wasteful — most of the time, nothing has changed, and the check was unnecessary.
Event-driven architecture inverts this. Instead of asking "has anything happened?" on a schedule, you declare: "When this specific event occurs, execute this function." The system listens for events and responds to them as they happen. No polling. No wasted checks. The event is the trigger, and the response is automatic.
Webhooks are a concrete implementation of this pattern. When you configure a webhook, you are telling one system: "When event X happens in your domain, send an HTTP request to this URL." A payment processor fires a webhook when a charge succeeds. A version control system fires a webhook when code is pushed. A form service fires a webhook when a submission arrives. The receiving system does not need to check constantly. It simply waits, and when the event occurs, it acts.
Publish-subscribe (pub/sub) systems extend this further. In a pub/sub architecture, event producers publish events to channels or topics. Event consumers subscribe to the channels they care about. The producers do not need to know who is listening. The consumers do not need to know who is producing. They are decoupled — connected only by the event itself.
The design principles that make event-driven architecture effective in software are the same principles that make event-based triggers effective for personal agents:
Loose coupling. In a pub/sub system, the event producer does not need to know about the event consumer. Similarly, your morning coffee routine does not need to "know" about your priority-review habit. The coffee is just coffee. But you have subscribed an agent to the event of pouring it, and that subscription is what creates the behavioral link.
Responsiveness. Event-driven systems respond to events as they occur, not on a fixed schedule. Your event-based triggers respond to transitions in your day as they actually happen, not when an alarm says they should happen.
Scalability. In software, you can subscribe multiple consumers to the same event channel. In your personal system, you can stack multiple micro-behaviors onto the same trigger event. "When I open my laptop" can trigger a priority review, a posture check, and a water intake reminder — multiple agents subscribed to a single event.
You are, in effect, building a personal event-driven architecture. Your daily events are the message bus. Your agents are the subscribers. The trigger-action format is your webhook configuration. The architecture that runs the internet's most responsive systems is the same architecture you use to run your most responsive personal behaviors.
How to design effective event-based triggers
Knowing that event-based triggers are powerful is not enough. You need to design them well. Here are the principles that separate reliable event-based triggers from ones that fail silently.
The trigger must be observable. You must be able to perceive the event with your senses — see it, hear it, feel it. "When I open my laptop" is observable. "When I am ready to focus" is not. Observable events fire reliably because your perceptual system detects them automatically. Internal states fire unreliably because they require self-monitoring, which defeats the purpose of event-based triggering.
The trigger must be specific. "When something happens at work" is not a trigger. "When I close the door of the conference room after a meeting" is a trigger. Specificity determines whether the event activates your stored intention. Einstein and McDaniel's research shows that distinctive, well-specified cues produce stronger prospective memory retrieval than vague or generic ones.
The trigger must be reliable. If the event does not occur consistently, it cannot serve as a trigger. "When my coworker asks about the project" fails if your coworker does not ask every day. The best event-based triggers are events that happen with near-certainty in your daily sequence — the non-negotiable transitions that occur regardless of what else varies.
The action must be immediate. The power of event-based triggers diminishes with delay. "When I open my laptop, I will review my priorities" works because the action follows the event immediately. "When I open my laptop, I will review my priorities sometime this morning" does not work because the delay severs the event-action link and re-introduces deliberation.
Start with one trigger. The temptation is to design event-based triggers for every transition in your day. Resist it. Trigger overload is a real failure mode — covered in L-0436. Start with one trigger attached to one transition. Run it for a week. Once it becomes automatic, add a second. Build your event-driven architecture one subscription at a time.
The hierarchy of trigger reliability
You now have the full picture of trigger types across L-0424, L-0425, and this lesson. Here is how they rank in reliability:
Environmental triggers (L-0424) are the most reliable because they require zero memory — the environment itself contains the cue. A sticky note on your monitor, a book on your pillow, running shoes by the front door.
Event-based triggers (this lesson) are the second most reliable. They require one act of memory encoding — forming the if-then association — but after that, the event itself handles retrieval. You do not need to monitor or check. The event shows up, and the intention comes with it.
Time-based triggers (L-0425) are less reliable for most behaviors because they require ongoing self-monitoring or external reminders, and they fire according to the clock rather than the actual context of your day.
Emotional triggers (L-0427, next) are the least reliable because emotional states are subjective, variable, and difficult to detect with precision. They have their uses, but they are advanced territory.
For most of the agents you design in Phase 22, event-based triggers will be your primary tool. They balance reliability with flexibility — they are nearly as dependable as physical environmental cues, but they adapt to the natural rhythm of your day in a way that static environmental cues cannot.
Your daily sequence is an event stream
Step back and look at your day as a whole. It is not a block of undifferentiated time. It is a sequence of events — a stream of discrete transitions, each one flowing into the next. You wake up. You get out of bed. You brush your teeth. You make coffee. You open your laptop. You check messages. You attend a meeting. You finish the meeting. You eat lunch. You return to your desk. You close your laptop. You commute home. You walk through your front door.
Each of those events is a potential trigger point. Each one is a moment where you can intercept your default behavior and replace it with an intentional one. Right now, most of those transitions are unloaded — they happen, and whatever default behavior is most salient fills the gap. An event-based trigger loads those transitions with purpose.
You do not need to redesign your entire day. You need to identify the two or three transitions where your default behavior is least aligned with your intentions, and load those specific moments with event-based agents. The rest of your event stream can run on autopilot. But at the moments that matter, you have an agent subscribed, waiting for the event, ready to fire.
That is what event-based trigger design is: subscribing your best intentions to the moments where they are most needed, so the events themselves carry out the work that willpower alone cannot sustain.
Sources
- Einstein, G. O., and McDaniel, M. A. (1990). Normal aging and prospective memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16(4), 717-726.
- Einstein, G. O., and McDaniel, M. A. (2005). Prospective memory: Multiple retrieval processes. Current Directions in Psychological Science, 14(6), 286-290.
- Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493-503.
- Gollwitzer, P. M., and Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69-119.
- Thurn, C. M., Wolff, W., and Gollwitzer, P. M. (2024). The when and how of planning: Meta-analysis of the scope and components of implementation intentions in 642 tests. European Review of Social Psychology.
- Clear, J. (2018). Atomic Habits: An Easy and Proven Way to Build Good Habits and Break Bad Ones. Avery.
- Fogg, B. J. (2020). Tiny Habits: The Small Changes That Change Everything. Houghton Mifflin Harcourt.
- Radvansky, G. A., and Copeland, D. E. (2006). Walking through doorways causes forgetting: Situation models and experienced space. Memory and Cognition, 34(5), 1150-1156.