Letters From Past Self: An Anniversary Journal App Feature Design

The best feature in any paper journal is one no one designed. You open last year's notebook and find an entry from a Tuesday you forgot existed. The words catch you off guard. That accident is the entire feature: same-date-different-year recall.

Digital check-in apps can do the same thing on purpose. Soulwise calls it "letters from past self." This is a short note about how it works, why it is harder than it looks to design well, and where the line between gift and surveillance sits.

What the feature does

The mechanic is small. On the anniversary of a past check-in - typically one year ago, six months ago, or three months ago - Soulwise resurfaces the original response inside the app's home screen. Users see:

A year ago today, you wrote: "Two things on your mind, none of them yours to fix today."

The date and the original chips that produced the response are shown next to it. The user can dismiss, save it as a highlight, or write a short reply.

That is the whole interaction. The complexity is everywhere except the feature itself.

Why it is hard to design well

Four risks worth naming.

The painful-anniversary problem. Some dates are heavy. A breakup, a loss, a hospital admission. The app cannot reliably know which ones. Pushing a memory from one of those days onto a home screen is a failure mode, not a feature.

The surveillance line. A feature that surfaces past entries reads as a gift if the user already trusts the app, and as a creepy reminder of how much the app knows if the user does not. The trust has to be earned before the feature appears.

The cold-start problem. New users have no past to surface. If the feature lights up on day three, it feels manufactured. If it never lights up, users do not know it exists.

The selection problem. Most check-ins are not interesting in retrospect. The day you logged "stretched" with a neutral mood is not a Tuesday worth re-living. The app has to pick the entries that actually have value to re-surface.

How Soulwise handles each

In-app only, never a push. Push notifications are governed by a strict rule: no menstrual content, no astrology content, no daily-check-in content of any kind. Letters from past self appear only inside the app, only when the user opens it on the anniversary date. The lock screen is left alone.

30-day warm-up. The feature does not activate until the user has been using Soulwise for 30 days and has completed at least 10 check-ins. Before that threshold, the past is too thin to mine.

Per-date skip setting. Users can mark a date as "skip" in a quiet settings screen. The skip applies forever; the app will never surface a memory from that calendar day. There is no follow-up question and no analytics event on what they marked.

Selection by signal density. The feature prefers check-ins where multiple chips were selected, where the mood was non-neutral, and where the user spent more than the median time on the response card. Routine days do not get re-surfaced. Days that took a moment to log do.

Plain dating. "A year ago today" works. "365 days ago" is creepy. Plain language about time keeps the feature on the right side of warmth.

Why borrow from paper journals

Three reasons.

The feature works in paper. People have been opening old notebooks and finding old selves since notebooks existed. The pattern is robust because it is grounded in how memory and time interact, not in any specific app design.

It does not require new user input. The check-in already exists. The feature is just a different way to surface it. No new logging, no new screens, no new permissions.

It strengthens the daily ritual. Knowing that today's check-in might come back on this date next year changes what users are willing to write. They are not writing for the algorithm; they are writing for themselves a year from now. That is a healthier reading frame.

What this is not

A note on what we deliberately did not build.

No social sharing of past entries. The letter is for the user. It is not a card to send to a friend or post on a timeline.

No AI rewriting of past entries. The original words stand. The app does not "improve" or "summarise" what the user wrote a year ago.

No streaks or gamification. Anniversaries are not a metric to optimise. Users do not see a counter for "memories surfaced."

No predictive framing. The feature does not say "based on last year, you'll feel X today." That is the model imagining it can predict your day from a year-ago data point. It cannot.

When the feature ships

Letters from past self is part of the Soulwise daily ritual scope but does not appear in the first month for any user, by design. The full architecture and timing are documented in the product spec; the Soulwise hub page is the public surface where the daily ritual itself can be tried.

The shorter version: the best journal features are accidents, and the job of the app is to schedule the accident without ruining it.