There is a sentence I hear constantly in organizations. It sounds reasonable, it sounds responsible, and it is one of the most effective barriers to lasting impact I know: "We need to be able to measure this."
The problem is not that the sentence is wrong. The problem is what it quietly does: it systematically shifts investment away from what actually shapes behavior and toward what can be defended in a quarterly report. And these two things, as I will show, almost never coincide.
Anyone working with behavioral design knows the pattern. The onboarding experience gets broken into individual steps, each step gets rated, the weakest gets improved. The performance review gets evaluated with a questionnaire, satisfaction scores get compared. The team event gets praised or criticized after the fact. All measurable. All defensible. And yet the behavior you actually wanted to change stays stubbornly unchanged.
Touchpoints, moments, experiences. Three things that are not the same.
To understand why, we need a distinction that is almost never made in practice.
A touchpoint is any sensory element of an experience that a person perceives: the tone of a welcome email, the acoustics of a meeting room, the wording of a feedback question, the color of the conference room. Touchpoints are the raw material from which experiences are built.
A moment is a single interaction that prompts an immediate reaction. The manager who says after the presentation: "Strong work, the team counts on you." The colleague who holds the door. The unexpected note during lunch. Moments are good. Positive moments are very good. They lift your mood, sometimes for the entire day. Behavioral economist Daniel Kahneman has shown that the peak and the end of an experience are remembered disproportionately, a finding known as the Peak-End Rule.
But moments, even exceptional ones, are like vitamin B12 shots. They give a boost. They do not shape behavior.
An experience is a series of moments that has repeated itself often enough that a person can predict it. Only that prediction shapes behavior.
That is the decisive difference. An experience is not a collection of moments. It is an internalized expectation. When someone knows, not hopes, not wishes, but knows, that the next team meeting will unfold like the last five, that their contribution will be acknowledged, that mistakes can be raised without consequence, then that prediction shapes their behavior. They prepare differently. They invest differently. They stay.
And that prediction is precisely what most measurement systems structurally render invisible.
Why the system favors the wrong thing
Marcus Buckingham at the Buckingham Institute has shown something uncomfortable. When you look at experience scores on a scale of one to five and ask which scores actually influence behavior, meaning customer loyalty, employee engagement, productivity, you do not find a linear curve. The world does not work such that a move from two to three has the same effect as a move from four to five.
The curve is exponential. The fives, the extreme experiences, produce a disproportionate share of what organizations actually want. Moving the ones to threes produces little of it, even when it feels like progress and looks good in the dashboard.
Hundreds of studies linking experience scores to behavioral outcomes reveal a curvilinear relationship: it is not the improvement from worst to average, but the design of extreme experiences, the fives, that produces the greatest leverage on behavior and results. Organizations that collapse fours and fives into a single "percent favorable" category lose exactly this information.
And yet most organizations invest in precisely the opposite direction. They look at the red numbers. They work on the weaknesses. They try to lift the average. This is not a thinking error. It is a system error. Because the system rewards exactly this logic: weaknesses can be named, fixed, and documented. Extreme experiences can rarely be demonstrated within a reporting cycle.
The problem with time horizons
Here lies the actual core of the problem, and it is more counterintuitive than it first appears.
External regulation, meaning rewards, controls, incentives, operates on short, clear time horizons. I do X, I get Y. That is transactional, close, and measurable. Intrinsic motivation, the state in which people perform from genuine drive, operates on time horizons that often have no concrete end date. I am growing as a leader. I am becoming the person I want to be. I am developing a practice.
These are not transactions. These are trajectories.
And this is precisely why short time horizons allow for calculation while long time horizons require identification. You cannot calculate whether it is worth spending five years on something whose returns only show up in year six. You have to want it as part of who you are or who you intend to become. Intrinsic motivation is not only future-oriented. It is the only form of regulation that functions under fundamental uncertainty about the future.
This means: anyone who wants to shape behavior durably must design experiences that enable exactly this identification. Not moments of recognition that lift mood. Experiences of belonging that shape identity.
What Trajectory Design means
This is the starting point for what I call Trajectory Design: the deliberate shaping of experience arcs over time, rather than optimizing individual touchpoints or moments.
Trajectory Design does not ask: how do we make this meeting better? It asks: what prediction should a person hold after twelve weeks in this organization? What should they know, not think, but know, because they have experienced it consistently, about what this collaboration means? And which series of touchpoints reliably produces that prediction?
This is not more abstract thinking. It is more precise thinking. A single moment can succeed or fail for many reasons. A consistent sequence of touchpoints, each carrying the same signal, shapes reliably. And it shapes more efficiently, because it does not require fresh investment each time. It runs.
The hockey-stick effect in experience design: invest long enough in consistent experiences and you do not get linear returns. You get identification. And identification scales.
That is the real cost of measurability. Not that you make bad decisions. But that you are systematically unable to make the decisions that would produce the most, because their returns only become visible long after the next reporting cycle has closed.
What this means in practice
Anyone in an organization who is responsible for experiences, whether that is onboarding, leadership development, team culture, or customer journeys, sits in a structural trap. The system demands short-term justification. The impact arrives long-term. And the language in which impact becomes visible is a different language than the language of the dashboard.
The way out is not to ignore measurability. It lies in running two things in parallel: what can be measured and keeps the organization calm, and what cannot be measured and shapes the organization over time. And knowing clearly which is which.
Trajectory Design is not a critique of short-term optimization. It is an extension. It asks: if we design this touchpoint this way today, what prediction does that create three months from now? If we hold this sequence of moments consistently, who will this person want to be in a year?
These are questions that cannot be answered in a quarterly report. But they are the only questions whose answers actually change the behavior you set out to change.