Return Path — give the learner a useful next step

Return Path is an original Northstar product prototype about respectful re-entry. A dormant learner may have left a worthwhile goal unfinished. A reminder helps only when its value, channel and timing fit that person's context.

Primary reviewer: a product manager deciding who should receive a return experience, what it offers, and who should be left alone. Beneficiary: a fictional learner resuming a saved goal without repeating setup.

Mo owns product and program direction. AI assisted implementation and software verification. The material choices below are provisional product judgments; no learner research or commercial results are claimed.

The decision

Offer a concrete continuation of an unfinished goal, then apply consent, preference, quiet hours and frequency limits before permitting a local simulated invitation. Preserve accepted decisions across overlapping runs. A more permissive policy revision must not erase earlier protections.

The principal rejected alternative was a broad dormant-user blast with a generic “we miss you” message. It has little connection to the learner's purpose and treats dormant status as permission to interrupt. Return Path uses an unfinished goal as the reason to return and makes suppression a legitimate product outcome.

Smallest observable outcome

A reviewer chooses C01, reads Aster's saved walking-plan context, previews the exact final-stop destination and a six-outcome local timeline, then confirms it. Aster is eligible at the start and one week later; the intervening attempt is suppressed. Birch's email attempts are suppressed because there is no email consent. Repeating the run leaves every new Aster attempt suppressed rather than bypassing the prior decision.

Success in this prototype means understandable, reproducible decisions and recovery. It does not mean retention improved or a message reached anyone.

Boundaries

Six original fictional profiles, three cohorts, three channel choices and three fixed starts. No real contact details, campaigns, notifications, accounts, tracking, automated persuasion, APIs, live AI or analytics. Return destinations are previews of fictional saved steps; the full learning workspace is outside scope.

Sand, ink and rose support a quiet workspace: context at left, value-led message at right, with the timeline following the user's deliberate confirmation. Case study · Sample walkthrough.