Put high-stakes decisions early in a flow, before a pile of minor ones

What it is

Since decision quality declines systematically with the number of decisions already made in a sequence, there's a directly actionable design conclusion: place decisions with serious consequences early in a flow, and push trivial, low-risk decisions later — rather than ordering by arbitrary sequence or the intuitive "easy first" heuristic. "Important" here means the severity and reversibility of the consequences if the decision is made poorly, which is a different axis from how hard or complex the decision is to operate: difficulty determines how easy this step is to get through, stakes determine how costly it is if it's gotten wrong, and ordering should follow the latter.

Why it happens

Decision fatigue is a position-dependent depletion of resources — the later in the sequence, the less deliberative capacity remains. Placing a high-stakes decision at the end of a sequence means the user faces it exactly when resources are tightest and most prone to sliding into sloppy judgment or default acceptance; the two compound so that the decision that most needed care is the one most likely to get shortchanged. Reversing the order — handling high-stakes decisions while capacity is still ample — uses position as a near-free variable to offset fatigue, without needing to change the decision's own complexity at all.

Where it stops holding

This recommendation holds only when the flow's ordering is actually under the designer's control. If a high-stakes decision's position is locked by an external constraint — it can only be made after information produced by some prior step becomes available — reordering isn't an option, and mitigation has to come from elsewhere (raising that step's visual prominence, adding an independent confirmation). Also, "important" should be judged by the actual consequences to the user, not by how complex or polished the step looks to the designer: a decision that's simple to operate but nearly impossible to undo if wrong (a payment method, an irreversible deletion scope) should be moved earlier ahead of a decision that's complex to operate but has inconsequential outcomes.

Applying it

• Map every decision point in the flow and order it by the severity and reversibility of a wrong choice — not by operational difficulty — placing the heaviest and least-reversible consequences earliest in the sequence.
• For decisions that must stay late in the flow because of a genuine business constraint but remain important, raise that step's visual prominence or add a mandatory confirmation on its own, using a local fix to offset the fatigue penalty of its position.
• Don't conflate "important decision" with "complex decision" when ordering — a complex but low-risk decision can safely sit late in the flow, since even a sloppy pass through it carries a bounded cost.
• How to verify it: run a comparison test on the same class of high-risk decision placed at different positions in the flow (early versus late in the sequence), comparing error rate, post-hoc correction rate, or support-ticket rate. Worse metrics the later the decision sits confirms the benefit of moving it earlier.

Related

• Same group: A9.12.1 decision fatigue shows up either as sloppy choosing or as avoidance and default-keeping · A9.12.2 the physiological mechanism behind decision fatigue is disputed, but the behavioral pattern is not · A9.12.4 cutting the number of decisions beats polishing any single decision's interface
• Nearby: A5.03 Sustained attention and vigilance decrement
• Search terms: decision fatigue · decision sequencing · high-stakes decisions first