The criterion is whether the influence serves the user's own goal

What it is

The line between persuasion and manipulation is judged by three parallel criteria — whether the influence serves the user's own goal, whether information is truthful and complete, and whether the user can easily reverse the choice — and all three must hold for it to count as persuasion. Goal alignment is the one that asks for whom: does this influence move users toward a goal they have declared or would recognize as their own, or toward the operator's goal at the expense of theirs? The dark patterns tradition is built along this axis: designs that make you do things you didn't mean to do. The criterion does not turn on intensity or legality — the same technique (a progress bar, a pre-ticked box, a countdown) can be entirely legitimate or entirely over the line; the difference is whose goal it serves.

Why it happens

Goal-misaligned influence is usually invisible within a single interaction: the interface wraps each request in its own local justification ("just remembering your choice", "one quick step"), every concession is small and individually reasonable, and the cost of the drift is paid elsewhere, later, in aggregate — so local evaluation never registers it. The audit has to move up a level and ask whose goal is actually being advanced: persuasion changes how the user evaluates means toward their own end; manipulation rewrites the end itself without the rewrite being visible. The asymmetry is also structural: operators run metric systems that continuously optimize per-step conversion, while users have no counterpart mechanism guarding their goals — misalignment is standing pressure, not occasional malice. The felt moment of being manipulated comes in retrospect: when users can no longer narrate the outcome as their own decision, the attribution conflict turns into betrayal, which is why misaligned designs can win short-term conversion while spending long-term trust.

Studying it

• Paradigm: dark patterns research is dominated by corpus audits — large crawls of real interfaces (shopping sites in particular), hand-labeled against a typology (sneaking, urgency, forced action), with misalignment with user intent as a core definitional dimension; complemented by controlled experiments manipulating pre-ticked defaults and reminder frequency, measuring sign-up conversion and post-hoc reversal, crossed with whether participants were informed beforehand.
• Variables: direction and strength of misalignment (mild vs. aggressive), informed vs. uninformed; outcomes are conversion rate, reversal/cancellation rate, and self-reported trust and felt manipulation.
• Methodological cautions: audits establish that such designs exist, not that users didn't want it — user goals must be measured, not assigned by the annotator; post-hoc reversal rates underestimate misalignment because users often cannot articulate their original goal afterwards either; practitioner studies trace misalignment to metric pressure propagating through organizations rather than to individual designers, which puts the intervention point in the incentive structure, not personal ethics.

Where it stops holding

User goals are not singular: when goals conflict internally (save money vs. have it now), "which goal to align with" is itself a design decision — the criterion yields no binary answer and only demands the trade-off be recorded rather than swallowed by defaults. Declared goals can be socially desirable answers (health on the survey, convenience in behavior), so "recognizable" must be cross-checked against behavioral evidence, not self-report alone. Nor does alignment license paternalism: serving users' long-term interests against their momentary goals is a different kind of overreach — the criterion separates operator from user; it does not adjudicate the user's own long-versus-short conflict. Evidence boundary: in corpus audits "misalignment" is judged by annotators, a gap away from actual users' endorsement of their goals, so audit findings are not direct proof of user harm.

Applying it

• Write a user-goal alignment note for every conversion design: what the user wants at this moment (with evidence — the search query, the task flow they are in), how the design serves it, and whether a reading user would claim "yes, this is what I want". If the note cannot be written, that is the red light.
• Run a purpose test on each persuasive technique: a progress bar serving "the user wants to finish" is legitimate; serving "retain a user who wants to leave" crosses the line; whether a pre-ticked box is legitimate depends on how the "user wants all of it" premise relates to actual user goals, not on the checkbox itself.
• Add an alignment check at the metrics layer: track post-hoc reversal rates and complaint keywords alongside every funnel step — conversion rising while reversals stay flat is what healthy persuasion looks like.
• To validate: hand the alignment note to real users and ask them to paraphrase it — can they say "this design helps me do what I want to do"; if the goals they name diverge from the goals the designers claimed, rework.

Related

• Same group: P2.08.2 The criterion is whether information is truthful and complete · P2.08.3 The criterion is whether the user can easily reverse the choice
• Nearby: P2.05 Loss aversion and framing · P2.06 Commitment and consistency · P2.11 Progress feedback and completion drive · P3.06 Notification-driven re-engagement
• Search terms: dark patterns · goal alignment · goal misalignment