When someone dies, most of us know roughly what to do for the first week. We come over. We bring food. We sit in the room and say the things people say, which are never quite the right things and are usually enough anyway.
Then the week ends. Everyone goes back to work, the house gets quiet, and the family is left holding the part almost nobody warns them about.
If you’ve been through it, you already know where this is going.
Grief does measurable things to a person. The best documented of them is attention. Concentration goes, processing slows, you read the same paragraph four times and walk into a room having forgotten the errand. Researchers still argue about how deep it runs and how long it lasts, and for a minority it becomes prolonged grief disorder, which the ICD-11 recognizes and which the DSM added in 2022, after listing a version of it in 2013 only as a condition needing further study. The early fog, though, isn’t in dispute. Joan Didion, who wrote the book people still hand each other after a funeral, called grief a place none of us know until we reach it. C.S. Lewis, writing after his wife died, described a winding valley where every bend shows you a landscape you weren’t expecting.
Now here’s what America asks of that same person, starting about a week later.
Order the death certificates, a dozen of them, because every institution wants an original. Notify Social Security. File the life insurance claim. Find the will. Get through probate, the court process that decides whether the will is valid and who’s legally allowed to act for the person who died, which means that until a judge says otherwise you can be a spouse of 40 years and the bank still won’t discuss the account. Close the accounts. Cancel the phone. Cancel the subscriptions nobody remembers signing up for. File a final tax return for someone who isn’t there to explain their own paperwork.
Most of the numbers below come from Empathy’s own annual research.
The average American bereavement leave is 5.6 days. Settling the affairs takes 15 months, or 18 if you’re the executor.
No federal law requires bereavement leave at all, though a handful of states now mandate some version of it, and SHRM puts the share of employers offering it voluntarily at roughly 90 percent. Nearly everyone offers it, and almost nobody offers enough.
The rest: around $12,600 in costs, a median of 12 hours a week of administrative work, from the 2023 report, and 92 percent of executors reporting their job took the hit, from 2024. EstateExec, which has no stake in any of this, puts the total effort near 570 hours per estate. And it doesn’t stop when the leave ends. Loss kept affecting employees for 10 months after they went back, and 19 months if they were the executor.
Amir was eight, the middle of three brothers.
Gura has said he watched what it did to his mother, and that what he remembers most is wanting desperately to take some of it off her. She never spoke about what happened. In his own words, she held the pain in silence.
The questions came as he got older, the ordinary things a boy wants to know about death, and he learned fast that nobody wanted that conversation. So he stopped asking. He kept thinking about it and writing about it for years and showed it to no one.
A house that couldn’t say it out loud, and a son who wrote it down and told nobody.
Years later Gura was running product at eBay, which had bought his first company, The Gifts Project, in 2011. A man on his team named Mark lost his wife.
Gura went to the house. He has called that conversation surreal, and it comes down to one sentence. Mark asked when he was expected back at work, one day after burying his wife.
Mark had never heard the word probate, Gura has said, until he was in it.
That’s the company. Everything after this is engineering.
Almost everyone knows the five stages, denial through acceptance. Two things about them matter here.
Elisabeth Kübler-Ross drew them from interviews with terminally ill patients facing their own deaths, not from families who had lost someone. And when the Yale Bereavement Study went looking for them in actual bereaved people, the reaction people reported most often, from the very first month, was acceptance. The dominant negative feeling across the whole two years was yearning. Missing someone, rather than denying or despairing.
Yale is a mixed witness on the stages themselves, since the five indicators did peak roughly in the order stage theory predicts. The harder verdicts came later. A 2010 replication found only limited support, and in 2017 Margaret Stroebe, Henk Schut and Kathrin Boerner argued that stage theory lacks sound empirical foundation, conceptual clarity and explanatory power and should be dropped from clinical practice outright, partly because grieving people who don’t move through the stages conclude they’re doing it wrong.
What replaced it for most of the field is a model Stroebe and Schut published in 1999. Bereavement, they proposed, pulls in two directions at once. There’s the loss itself, the missing and the memories. And there’s everything the death changed about daily life, which in their own list means the legal matters, the finances, the new roles nobody asked for and the future that has to be rethought. Healthy grieving means swinging between the two, which they call oscillation, and taking breaks from both, because nobody can carry either one for long.
Their name for the second category is restoration-oriented stress. Researchers have been saying for 25 years that the administrative aftermath isn’t separate from the grief. It’s half of it.
Gura’s version, which he arrived at from his mother’s kitchen and from Mark’s living room: grief is made hard by logistics, and logistics are made hard by grief.
Gura and Yonatan Bergman, who’d worked together since The Gifts Project and again at WeWork, founded Empathy in 2020. Gura had come out of Reichman University’s Zell program and served as a first lieutenant in the Israeli Air Force. Early backers included Shai Wininger of Lemonade, Micha Kaufman of Fiverr and Gura’s own brother Eyal.
The product was pointed at the United States, and Gura’s stated reason is the part I keep turning over. The market is enormous, obviously. He also says American bureaucracy after a death is unusually punishing, which is a thing you only say out loud if you’ve looked at somebody else’s country and found it wanting.
Ten institutions, several more than a century old, invested in the same company and called themselves the Empathy Alliance. The venture arms of New York Life, MetLife, Aflac, Allianz, MassMutual, Securian, TIAA and Sumitomo, plus Citi’s impact fund and Munich Re. A bank and a reinsurer sitting down alongside eight life carriers. Gura called it unprecedented and I can’t find a reason to argue.
The list of partners who actually hand it to families is wider. Empathy now reaches more than 50 million policyholders across the US, Canada and the UK, works with eight of the ten largest life insurers in America, covers roughly one in five life insurance claims in the US, and sits inside the benefits packages of more than a thousand employers. Transamerica went live in July. TD, BMO and iA Financial in Canada. Zurich in the UK.
The product grew to match. Loss Support for the aftermath, LifeVault for the planning people do beforehand so their families aren’t guessing, and this year Leave Support and Caregiving Support, because the crisis usually starts months before anybody dies. Through all of it a Care Team of qualified humans, reachable around the clock.
Fifty million people are carrying something they never bought, from a company most of them have never heard of, and they won’t find out it’s there until the week they need it.
Empathy runs mostly through insurers and employers, who pay so the family doesn’t. Gura has written that he always meant it to work directly with families too, because there’s no other way to meet people where they are, but the carrier route is the one that dominates.
Gura’s stated rule is that the company should never increase the share of a family’s wallet spent during this transition, only decrease it. He’s also said plainly that the loss industry is built to serve the rich and take advantage of everyone else, and that the exploitation runs on a lack of transparency at the moment people are least able to push back.
He’s described what changes on the carrier’s side, too. The first conversation an insurer has with a beneficiary arrives with a death certificate attached, and the industry is moving from sorry for your loss to we’re going to stick around.
You could build a very good business on that vulnerability. Empathy structured itself so it can’t.
In June the company published research run with Censuswide across the US, UK and Canada, surveying 6,002 people who’d been through loss. Only 31 percent said they got the support they needed, one in five had nobody at all to turn to, and so people improvised: nearly a quarter of AI users said they’d turned to a chatbot while grieving, and 97 percent of those reported at least one positive effect. Bergman’s explanation is that people turn to any help they can find, because the existing systems simply don’t show up for the complexity of loss.
The same respondents drew hard lines. Eighty-six percent said AI tools must be transparent about what’s real and what’s generated. Eighty percent said consent should be required before using someone’s voice, likeness or data. Seventy-one percent said a counselor or therapist should be in the loop. And most rejected or were uncomfortable with applications that try to recreate the person who died.
Empathy’s formula is AI plus human care, and for years the AI stayed backstage carrying administrative weight so the Care Team could be present for the rest. Recently some of it moved forward. LifeVault Conversations gives people a private, AI-guided space to rehearse the conversations loss demands before they have to have them for real.
Whether that’s the right line is an open question.
This isn’t a charity and it isn’t magic.
Empathy operates in the US, Canada and the UK, and for most people it arrives through a life insurance policy or an employer’s benefits package rather than something you go buy, so if your carrier isn’t a partner you may not have access at all. It doesn’t replace an estate attorney on a complicated estate and isn’t meant to. Strip the branding off and it’s two things at once, administrative labor and a trained ear, paid for by somebody other than the family. A few others are building toward the same thing, but nobody was delivering it at this scale before.
The support-gap research is Empathy’s own, commissioned by a company with a commercial interest in the answer. It was fielded independently across 6,002 people, which is a real sample, and I’d still read it as a company’s research rather than as neutral science. Same for the reach figures, which are the company’s count of itself. The Alliance was ten institutions when they announced it and nobody has published a revised number since.
And there’s an Israeli irony here. The engineering happened in Israel, roughly half the staff is still there, and the whole thing exists to solve somebody else’s bureaucracy.
Gura has said the Israeli market is too small to justify full deployment. The people who built it largely can’t use it.
One personal note, and it isn’t an argument. I’ve sat shiva more times than I want to count, and I never once had to ask what day it was or who was bringing dinner. Somebody worked that out a long time ago.
In the essay Gura wrote a year into this, he reduced the whole company to three words. Empathy, not sympathy.
Families in Ohio and Manchester and Toronto will go through the worst stretch of their lives this year, and someone will walk them through it, and almost none of them will ever know it was built in Tel Aviv.
Built in Israel. For Everyone.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.