Picture a will signed eight years ago. It names a guardian who has since moved to Lisbon, an executor the family stopped speaking to after a bad Thanksgiving, and a house that was sold in 2021. Legally, it is still a perfectly valid will. Practically, it is a description of a family that no longer exists.
The question people usually ask — how often should I update my will? — is quietly the wrong question, because the honest answer isn't a number of years. Estate plans don't decay on a schedule, the way a roof or a passport does. They break at specific moments: a wedding, a divorce, a birth, a move, a sale. The moments are predictable. What's unpredictable is whether anyone notices, because the document sits in a drawer where nothing beeps when your life outgrows it.
A finished plan is an unwatched plan
There's a well-documented quirk in how memory handles unfinished business. The Zeigarnik effect describes how open tasks keep intruding on our attention — the email you haven't answered nags at you in the shower. Finishing a task releases that tension. That release is precisely the problem with estate planning: the day you sign, the mental file closes. The nagging stops. The will drops out of active monitoring entirely, and it does not reopen on its own.
Research on prospective memory — remembering to do things in the future — sharpens the picture. Psychologists distinguish event-based intentions (when X happens, do Y) from time-based ones (in five years, do Y). Event-based intentions succeed far more reliably, because the event itself acts as the reminder. Time-based intentions require self-initiated monitoring: you have to spontaneously check a mental clock that nothing prompts you to check. "Review the estate plan every few years" is a time-based intention with no clock, no cue, and no deadline. It is engineered to fail.
Status quo bias finishes the job. Once a plan exists, keeping it feels safer than touching it — reopening the document means reopening the mortality conversation, the family conversation, maybe the lawyer's invoice. So the signed-in-2014 version persists, not because anyone decided it was still right, but because no moment ever arrived that forced the question.
The events that quietly break a will
Some life changes merely make a plan suboptimal. Others break it outright. The breaking kind cluster into a handful of categories.
Marriage and divorce. In many states, divorce automatically revokes the provisions of a will that favor an ex-spouse — the law essentially pencils them out. But that patch applies unevenly: it varies by state, it typically reaches the will and not everything around it, and it does nothing for the ex-spouse still listed as your healthcare agent or emergency contact. Marriage runs the other direction: if you marry after signing a will that doesn't mention your spouse, many states grant that spouse a statutory share of your estate regardless of what the document says. Either way, the plan on paper and the plan in force have silently diverged.
A child arrives. Most states have pretermitted heir rules that give a child born after the will a share of the estate, so a new baby usually isn't disinherited by accident. But a statute cannot nominate a guardian, stagger an inheritance past age eighteen, or distinguish between the child you named and the sibling who came later. The law hands over money; it cannot hand over judgment.
Someone named dies, or drifts. Executors get sick. Guardians divorce, relocate, burn out. A beneficiary passes away before you do, and the gift either lapses or gets redirected by an anti-lapse statute's best guess. Every name in an estate plan is a bet on a relationship staying roughly as it was, and relationships don't sign anything.
You move states. A will validly executed in one state is generally recognized in another — the document doesn't void at the border. But the terrain under it changes. Community property and common-law states treat marital assets differently. Some states impose residency or bonding requirements on out-of-state executors. Formalities like self-proving affidavits, which let a will sail through probate without tracking down witnesses, differ in form. The will survives the move; its smoothness may not.
Assets change shape. Here the failure has a name lawyers know well: ademption. If your will leaves "the house on Maple Street" to your daughter and you sold that house two years before you died, the gift simply fails — she doesn't automatically get the new house or the sale proceeds. Specific bequests are photographs of a moment; sell the thing photographed and the frame is empty. Starting a business, receiving an inheritance, or moving significant money between account types all reshape an estate the old document was never sized for.
Children grow up. The guardianship clause you agonized over expires in relevance the day your youngest turns eighteen. The live question shifts from who raises them to how and when they receive — and a will written for toddlers rarely answers it.
What the law patches — and how crudely
Legislatures know people forget, which is why the default machinery exists at all: divorce-revocation statutes, pretermitted heir shares, anti-lapse rules. It's worth seeing these for what they are — evidence that failing to update is so common the law had to build airbags for it.
But airbags are not steering. The statutes guess at what an average person would have wanted, and they stop at jurisdictional lines. The most famous gap is federal: retirement plans governed by ERISA generally pay whoever is named on the beneficiary form, and in Egelhoff v. Egelhoff the U.S. Supreme Court held that federal law can override a state statute that tried to strike an ex-spouse from such a designation after divorce. The form on file won, years after the marriage ended. Every default fix is like this — partial, blunt, and indifferent to what you actually meant.
Tie the review to the event, not the calendar
The research on implementation intentions, developed by psychologist Peter Gollwitzer, offers the practical repair. Vague goals — be better about reviewing the estate plan — reliably underperform if-then plans that bind an action to a concrete trigger: when I change my address, I check the plan. When I file taxes the year of a family change, I check the plan. When anyone named in the documents has their own major life event, I check the plan. The trigger does the remembering, which is exactly what a time-based resolution can't do.
The other half of the repair is lowering the cost of the check itself. Nobody rereads a thirty-page will on a Tuesday. But almost anyone will scan a one-page summary: who is named for each role, what is specifically given to whom, which accounts carry their own beneficiary designations, where the originals live. Reviewing a summary takes five minutes and answers the only question that matters — does this still describe my life? Keep an annual backstop if you like (tax season is a natural anchor, since your finances are already spread on the table), but treat it as the safety net, not the system. The events are the system.
This is, in the end, what Heirloom is built around: not the one-time act of making a plan, but the unglamorous work of keeping it true. It holds that living one-page picture — the people you've named, the assets you've listed, the documents and where they are — so that when life shifts, the check is a glance instead of an archaeology project, and the plan your family eventually relies on describes the family they actually are. If you'd rather not maintain that summary by hand, you can start yours at Heirloom.