What Is a Gratitude Jar — and Why Does Ours Pick a Memory at Random?

By Matt Hughes · 2026-07-27 · 7 min read

A gratitude jar is meant to be opened once a year. I built one that hands you a memory back at random, any time — and that turned out to be the whole point.

A gratitude jar is one of the oldest tricks in the book: you write down a good moment on a slip of paper, drop it in a jar, and at some point in the future — usually New Year's Eve — you tip the whole thing out and read a year of your life back to yourself. I've kept a version of this on paper before. I've also built a digital one into Kansha. And the thing I didn't expect, until I'd actually shipped it, is that the writing was never the important part. The important part is being handed one back when you're not expecting it.

What is a gratitude jar, actually?

Strip away the mason-jar aesthetic and it's a very simple mechanism: a place to store the good bits, separate from the rest of your journal, so they don't get buried under everything else you wrote that week. Most versions of the physical practice work the same way — you write a line, fold it, drop it in, and leave it alone until a set date. The "leave it alone" part is doing more work than it looks like. You're deliberately delaying the reward. You're banking something good for a version of yourself who, on 31 December, might really need reminding that the year wasn't just the bad weeks that stick in your memory.

When I built the Jar feature into Kansha, I kept that core idea — you tap a heart or a "jar" icon on any entry or streaming note and it gets set aside — but I changed the ending. Instead of one big reveal on one day a year, you can tap the jar itself at any point and it hands you back one moment, chosen at random, from everything you've kept. Sometimes it's from last week. Sometimes it's from eight months ago and you'd completely forgotten it happened.

Why do people wait a year to read them back?

The traditional version delays gratification on purpose, and there's a reasonable logic to it: if you read your notes back every day, you'd habituate to them fast — the tenth time you read "grateful for a good coffee with Sam" it barely registers. Waiting a year means the notes have gone properly cold. You've forgotten most of the detail, so reading them back feels like discovering something rather than confirming something you already knew.

That's a real effect, and it's the same one that makes old photos hit harder than the photo did on the day you took it. But a full year is also a long time to keep faith with a jar full of paper that's doing precisely nothing for you in between. I wanted the payoff sooner than December, without it going stale after the second read.

What does the research actually say about rereading old good moments?

This is the bit I wanted to get right before I built anything, because "just reminisce more" is exactly the kind of vague advice I try not to put in this app. The most useful study I found is Speer and Delgado's 2017 paper in Emotion: they had people either recall a positive memory or do a neutral task before a stressor, and the group who'd just relived something good showed a dampened cortisol response and reported less negative affect afterwards. Recalling a specific good memory isn't just pleasant in the moment — it changes how your body handles what comes right after it.

There's also a long line of work under the label "savoring" (Bryant and Veroff's research is the reference point most positive psychologists cite), which distinguishes between experiencing something good and deliberately extending or reliving it afterwards. Their findings are consistent: people who make a habit of reminiscing on purpose — not just letting good moments happen and move on — report higher life satisfaction than people who experience similar amounts of good but never revisit it. The jar is, structurally, a savoring device. It forces the second half of the process that most of us skip.

None of this is a claim that a random note from a jar will fix a bad day. It's a much smaller, more honest claim: deliberately re-encountering a specific good memory measurably shifts your physiological and emotional state in the following minutes. That's a real, if modest, effect — and it's one you can only get if something is actually stored somewhere to be re-encountered.

Why does the random pick matter more than the writing?

Here's the part that surprised me once real people started using it. I'd assumed the value was almost entirely in the writing — the moment of noticing something good and getting it down. The jar was going to be a nice-to-have, a bit of delight sprinkled on top. What I actually saw, watching how people used it, is that the tap-to-reveal is where the emotional hit lands, and it lands harder specifically because you don't control which memory you get.

If you scroll your own journal in date order, you're the one deciding what to look at, which means you're half-anticipating it before you read it. A random pull removes that. You tap not knowing whether you're about to get something from last Tuesday or from a March afternoon you'd genuinely forgotten. That small unpredictability is doing something similar to what behavioural research on variable rewards has shown for decades in other contexts — an outcome you can't predict holds your attention differently to one you can. I'm not claiming the jar is a slot machine dressed up in pastel colours. I am saying that removing your own control over which memory surfaces seems to be exactly what stops the tenth reread from feeling as flat as the tenth reread of a diary you're paging through in order. It also solves the year-long patience problem from the traditional version. You don't have to wait for New Year's Eve for the payoff, and because it's random rather than sequential, reading it back in week three doesn't spoil anything — there's no "ending" to spoil. The jar can be tapped a hundred times and still occasionally surprise you, because the surfacing logic doesn't repeat until you've genuinely cycled through everything you've kept.

What building this taught me that I didn't expect

A few things became obvious only after watching it in use rather than designing it on a whiteboard. The mood colour on each note — I coded each kept entry with the soft colour tied to how you were feeling that day — turned out to matter more than I predicted. People jar plenty of notes from flat, ordinary, even mildly rough days, not just the obviously good ones. When you tap the jar and get back a moment that was written on a grey-mood day but still had something worth keeping in it, that seems to land differently — a reminder that the good stuff isn't reserved for the days that already felt good.

The other thing: almost nobody jars something the day they write it. It's nearly always a note you come back to later — during the "show all" list, or scrolling old entries — and decide, in hindsight, was worth keeping separately. That's the opposite of what I expected. I thought the jar would be a same-day impulse. It's actually a curatorial one, which tells me the jar is less "capture the good moment" and more "a second, slower pass where you decide what actually mattered."

Does it actually work, or is it just a nice feature?

I want to be straight about the limits here, because it would be easy to oversell this. The jar surfaces what you've already written — it doesn't manufacture gratitude out of nothing. If you've never tapped the little jar icon on an entry, your jar is empty, and an empty jar does nothing for anyone. It's also not a substitute for the daily habit of noticing and writing something down in the first place; it's downstream of that habit, a second use for entries you've already made. And I'd be lying if I said every single random pull lands — some days you get a note about a decent sandwich and it doesn't move you at all. The effect is real but it's not guaranteed on every tap, any more than every photo in an old album hits you the same way.

What I can say honestly, from watching this feature get used for months now, is that it's the single part of Kansha people screenshot and send to me unprompted — usually with some version of "I'd completely forgotten this happened." That's a smaller, more specific win than "this app changed my life," but it's the one I actually trust, because it's the same thing Speer and Delgado were measuring in a lab: a moment of genuine, involuntary reencounter with something good, landing on a day you weren't expecting it.