Working synthesis

Steering vs. Becoming

A map of reflection tools for life transitions.
Lo/Be Lab State of the field 2026
The gap

Over the last two decades, decision science has been remarkably successful at one thing: changing behavior at scale. Nudge units sit inside governments; choice architecture shapes retirement savings, organ donation, and checkout flows. The field's center of gravity, visible in the dominant frameworks (COM-B, EAST, MINDSPACE, the Fogg model), is the engineering of an environment so that a person is more likely to take a predetermined action.

Almost none of this has been turned on the decisions that matter most to the people making them: the high-stakes, low-feedback, identity-laden transitions of choosing a field, leaving one, deciding who to become. When decision science has touched life planning, it has usually treated it as one more behavior to steer, for example a life-planning app that nudges users toward holistic planning (see the applied case material published by behavioral-science consultancies such as The Decision Lab).

That is the gap this lab works in: applying the rigor of decision science to life transitions, while refusing its default intent.

Two intents, one literature

It helps to separate two things the literature usually blurs: the method (study a decision context rigorously, diagnose, prototype, test) and the intent (what you want to end up in the person's hands). On intent, there are two poles, and they already exist in the literature.

Steering, the nudge
Engineer the choice environment so the person is more likely to pick the option the designer judges better. Effective, scalable, and by design it does its work whether or not the person's own competence changes (Thaler & Sunstein, 2008).
Becoming, the boost
Build the person's competence to decide for themselves, so the benefit persists after the designer leaves. The target is the chooser, not the choice (Hertwig & Grüne-Yanoff, 2017).

These are not good versus bad; both can be rigorous and well-intentioned. They differ in where judgment is meant to end up. A nudge keeps judgment with the designer; a boost returns it to the person.

A telling detail: the leading applied-behavioral-science practices hold a boost-compatible theory in their own canon, Self-Determination Theory, with its account of autonomy, competence, and relatedness (Deci & Ryan, 1985), yet subordinate it beneath a stack of steering frameworks, where the day-to-day delivery vocabulary is EAST, MINDSPACE, and the like. The becoming end of the spectrum is under-built.

A map

Place reflection and decision tools on one axis, steering to becoming, and a second axis for time horizon, single session to longitudinal.

Steering, single-session
Classic nudges, default settings, one-shot framing.
Steering, longitudinal
Habit apps, streaks, sustained behavioral programs.
Becoming, single-session
One-off reflective exercises, coaching prompts.
Becoming, longitudinal
The under-occupied quadrant: tools that build a person's reflective and decision-making capacity, sustained across time and at the scale of a whole cohort. This is where Lo/Be Lab's projects sit.

Mapping the lab's own work onto the becoming and longitudinal quadrant:

  • DartWorld: sustained reflection across a term, not a session (RQ 1).
  • Synapse: replaces the ranked list, a steering move that says "here is number one", with a spatial map the person navigates themselves, a becoming move.
  • AI as mirror, not advisor: the boost stance rendered in software, surface the evidence and return the judgment.
  • Narrative by Design and Career Design Lab: competence-building reflective practice, run as repeated research cycles with real cohorts.
Why the becoming end is harder, and why that is the research

Steering has a built-in success metric: did the behavior change? Becoming does not. Building someone's capacity to decide is harder to instrument than counting clicks. This is exactly why it needs design-based research: multiple versions, mixed quantitative traces and qualitative coding, measured across terms, with failures reported. The measurement problem is not a reason to avoid the becoming end; it is the research program. See the methods.

What this implies for practice
Diagnose before you design
Borrow the discipline of behavioral diagnosis, but ask a becoming question: not "what barrier blocks the target behavior?" but "what would let this person see and decide for themselves?"
Prefer boosts where the decision is the person's to live with
For high-stakes, identity-laden, infrequent decisions, a nudge that fades leaves nothing behind; a boost compounds.
Keep judgment with the person, especially with AI
As models get more capable of recommending, the autonomy-preserving choice, mirror not advisor, becomes a design stance you have to take deliberately, because the default will be to advise.
Where this leaves the lab

Lo/Be Lab's wager is that the becoming and longitudinal quadrant is both the most neglected and the most consequential place to apply decision science, and that life transitions are its natural home. The Architecture of Becoming is the attempt to give that quadrant a design theory. The projects on this site are that theory under test.

Cite this

Looper, S. (2026). Steering vs. becoming: a map of reflection tools for life transitions (Working synthesis). Lo/Be Lab. lo-be-lab.com/publications/steering-vs-becoming.html

Open to pushback, particularly on the two-axis map and on the claim that the becoming and longitudinal quadrant is genuinely under-occupied. Write to the lab.

References
  • Deci, E. L., & Ryan, R. M. (1985). Intrinsic Motivation and Self-Determination in Human Behavior. New York: Plenum. (Self-Determination Theory: autonomy, competence, relatedness.)
  • Hertwig, R., & Grüne-Yanoff, T. (2017). Nudging and boosting: Steering or empowering good decisions. Perspectives on Psychological Science, 12(6), 973–986. (The boost tradition, the becoming pole.)
  • Kahneman, D. (2011). Thinking, Fast and Slow. New York: Farrar, Straus and Giroux. (Bounded rationality: why decisions need scaffolding at all.)
  • Michie, S., van Stralen, M. M., & West, R. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science, 6, 42. (COM-B and the diagnose-before-intervention discipline.)
  • Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving Decisions About Health, Wealth, and Happiness. New Haven: Yale University Press. (Choice architecture, the steering pole.)
  • The Decision Lab. Applied behavioral-science guides and case material. (The contrasting steering practice.)