Why we built Loristry the way we did

Few people need convincing that their parents' stories are worth keeping. Ask whether someone would want their mother's childhood written down and the answer is an immediate yes, usually followed by a slightly guilty look. The trouble comes later, in the finishing.

So when we started building Loristry, we spent our early research on why these projects stall and which of those reasons a tool can do something about. Three answers kept coming back, and each one turned into a decision we've had to defend since.

A question a week reads like an assignment

The best-known model in this category mails the storyteller a question every week and asks them to write back. It's an elegant design, with a gentle cadence and a hardcover waiting at the end of the year. For someone who enjoys writing, it works, and we say so plainly in our comparison with StoryWorth.

But watch what happens to everyone else. An email arrives asking your father to describe his first job in a few paragraphs. He hasn't written anything longer than a text message in decades. The email sits. A second one arrives the following Tuesday, and now there are two. By week five the thing on his desk has stopped being a gift and become an overdue assignment. Families quit these projects when writing turns a conversation into work, and homework loses to almost everything.

In the reviews and forum threads we studied while building this, the same pattern repeats: people tell their stories willingly the moment somebody sits down and asks. That's the whole gap. The writing is the extra step, and it comes with the bill.

So Loristry asks out loud. It interviews the person by voice — in an ordinary browser, or over a regular phone call for the grandmother whose relationship with technology ends at the telephone — and it asks the follow-up question the moment a name or a half-told detail goes by. No one in the family has to run the interview or worry about what to ask next.

One narrator only gets you one version

The second thing we noticed is that these tools nearly all assume a single storyteller. One person answers, one book comes out. It's a clean model, and it quietly loses most of the story.

Your mother can't tell you what it was like to be her daughter. She'll skip the year she thinks is boring and the year she thinks is nobody's business, and she'll flatten the parts she's told a hundred times into the version she's settled on. Meanwhile your sister remembers the same summer differently, and a grandchild can ask a question that would sound strange coming from you (and get a real answer).

A single narrator also puts the entire project on one pair of shoulders. When that person gets busy, the biography stops.

So we built Loristry to interview everyone. Your mother's siblings, her oldest friend, the grandchildren: anyone invited is interviewed the same way and adds their own memories and photos. When two people describe the same summer, we connect the tellings, and what comes back reads as one story told by everyone who was there.

Generic tools flatten the person

The third problem is subtler, and it's the one we're still working hardest on. Run someone's life through a general-purpose tool and you tend to get competent, tidy prose that could be about anyone: the dates survive and the person goes missing.

What makes a biography worth reading is the stuff that resists tidying: the phrase your grandfather always used, or the way he answers a question about the war by talking about a dog. A system optimizing for clean paragraphs edits exactly that out.

So we aimed at the opposite. Loristry keeps the recording, follows the digression instead of steering back to the outline, and organizes the result into chapters and a timeline rather than smoothing it into a single voice. It's a harder target and we don't always hit it. But a readable book about a stranger who shares your father's name isn't what anyone paid for.

What we gave up

These decisions cost something.

A conversation is harder to schedule than an email. A question a week asks almost nothing of the storyteller's calendar; an interview asks for twenty minutes of real attention, and some people would rather have the email. Inviting the whole family means coordinating a family, which is its own project. And a record that keeps growing has no natural finish line. For some people, the fixed twelve-month assignment ending in one good book is the better shape, and they should buy that instead.

We've written up the field, including where we're the wrong pick, in our guide to StoryWorth alternatives. I'd rather you choose the tool that fits the person you have in mind than assume ours always is.

What we weren't willing to trade was the part that decides whether any of this finishes: somebody has to draw the story out, and it shouldn't have to be you.

— Scott