Personal and Family Websites

What people around here actually build

Not everything on the web is a business. A fair share of the sites we are asked about have nothing to sell, and they matter to the people who own them in a way a plumbing company's site does not.

• Family history and genealogy sites. Somebody has spent fifteen years in county records, cemetery walks and courthouse basements across Kaufman, Henderson and Van Zandt counties, and the result currently exists as a software file on one laptop.
• Memorial and tribute pages. The obituary, the photographs, the service, and what people wrote afterward, gathered somewhere that will not scroll away.
• Family reunion sites. Where and when, who is bringing what, directions to the pavilion, and eleven years of photographs.
• Small clubs and societies. A quilting group, a bass club, a garden club, a historical society.
• Small churches. Service times, address, a map, and who to call.

These share one requirement that no business site has: they need to still be there in twenty years. A business site gets rebuilt every five or six years by whoever runs the business then. A family history site has one job, which is to outlive the person who made it. That single requirement changes almost every technical decision, and it points away from most of what the industry sells.

Why not just use Facebook

The honest answer rather than the sales answer: Facebook is genuinely better at one half of the job. Reaching relatives is its whole purpose, it is where your cousins already are, and a post about the reunion gets seen by people who would never visit a website. Nobody should pretend otherwise.

What it is bad at is keeping. A feed is chronological and algorithmic, so a story posted in 2019 is functionally unfindable in 2026 even to somebody who knows it exists. There is no table of contents and no index. There is no stable web address a cousin can bookmark, a newspaper can cite, or a search engine can send somebody to years later. The organizing principle is recency, and family history is the opposite of recent. There is also advertising sold against your grandmother's obituary, chosen by an auction, with no say from the family. And all of it is reachable only while the platform exists and while somebody's account exists.

The sensible answer for most families is both, with the roles kept straight. The website is the archive: permanent addresses, a table of contents, the full material, arranged so somebody can find one branch of the family without scrolling. Facebook is the megaphone that points at it.

What happens when the admin dies

This is the question nobody asks until it is urgent, and it has a documented answer that surprises most people.

Facebook introduced the legacy contact on 12 February 2015. A legacy contact for a memorialized personal profile can write a pinned post, respond to new friend requests, and update the profile and cover photographs. With permission granted while the person was alive, they can download an archive. Facebook's wording on the limits is unambiguous: the legacy contact will not be able to log in as the person who passed away or see that person's private messages. Read that carefully, because it is the whole problem. Anything that required that login still requires it, and nobody has it.

Facebook Pages, as distinct from personal profiles, are a separate question, and we will not state an outcome as fact because we could not verify one; the relevant help documentation is not publicly retrievable. What can be said responsibly is that a Page administered by a single person whose account is later memorialized is at minimum not straightforwardly transferable. Treat it as a risk rather than a rule — and if a family, church or club page has exactly one administrator, the fix costs nothing: add a second one today.

Platforms delete things, and the web decays

The case for owning the archive is not opinion. It is one of the few arguments in this industry with a documented evidence base, and every item below is checkable.

Platforms shut down deliberately. Yahoo closed GeoCities in the United States on 26 October 2009; at least 38 million pages, most written by ordinary users, had been published there. Volunteers salvaged a few hundred gigabytes. Everything they did not reach is gone. Google announced the shutdown of Google+ on 8 October 2018 and consumer service ended on 2 April 2019, earlier than first planned, after a second security flaw. What was not exported first was deleted.

And the web at large decays quietly. Pew Research Center published a study on 17 May 2024 finding that a quarter of all webpages that existed at some point between 2013 and 2023 are no longer accessible, and that 38 percent of pages from 2013 were gone by October 2023.

None of this makes a family website immortal. It means the failure modes are different, and controllable. A website fails when the domain lapses or the hosting bill stops being paid, both of which are visible and both of which somebody can be made responsible for. A platform fails when a company several thousand miles away makes a business decision, and nobody in your family gets a vote.

What a family or memorial site actually needs

Almost none of what a business site needs. Recognizing that is what keeps these projects small and durable.

Durability over features. Plain HTML pages and image files, on a domain the family owns, with the whole site backed up as an ordinary folder somebody can copy to a hard drive. Nothing to update, nothing to patch, nothing to expire except the domain. A static site has essentially no attack surface, costs very little to host, and will still open in a browser in fifteen years.

A domain registered for many years, paid from an account somebody else can reach. The most common way a family site dies is not hacking. It is a lapsed domain on an expired credit card belonging to somebody who has died. Register for five or ten years at a time, and make sure at least two people can get into the registrar account.

A named successor. Who has the login, where the backup lives, written on paper in the same place as the will. Nobody does this, and it is the single most useful sentence on this page.

A privacy decision made deliberately. Living people's birth dates, mothers' maiden names and home addresses are precisely the answers to bank security questions. Standard genealogy practice is to publish full detail only for the deceased and names only for the living, or to put living relatives behind a password.

Photo scans at archival resolution, stored separately from the website. The site shows a screen-sized version; the high-resolution master lives on two drives in two buildings. A website is a publishing channel, not a backup.

Portability for the tree. Genealogy data has an open interchange format, GEDCOM, maintained as an open-source project by FamilySearch; the current release is 7.0.18, published 17 February 2026, though the older 5.5.1 from November 2019 remains what most software actually exchanges. Tree data can be exported and rehosted; the photographs, document scans and written stories usually cannot. Put the tree on the platform if you like, and the irreplaceable material on the site you own.

The mistakes that end family sites

Every one of these has happened to somebody, and most are free to avoid.

• The domain registered to one person's personal email and personal card. When that person dies or changes banks, the site vanishes at renewal, and after the redemption period anybody in the world can register the name.
• Publishing living relatives' full birth dates and mothers' maiden names. Free material for identity theft, and it causes real friction at reunions.
• Building it on a content management system with a gallery plugin and a slider plugin. Now the memorial site has a patching schedule, and in three years it will be broken by an upgrade nobody performed or quietly serving spam the family learns about from a search result.

When you do not need a website for this

We would rather talk somebody out of a project than build something that becomes a chore.

A one-off event. A reunion, a wedding, a ninetieth birthday. A shared photo album and a group message thread do the job better, because everyone is already in them. A website with three visits is not a memorial, it is homework.

A memorial in the first weeks. When somebody has just died, the memorialized Facebook profile is honestly the better place: people are already there, comments arrive, and the grieving happens in company. A static page cannot do that. Build the permanent site later, calmly, and move the good writing into it before it scrolls away.

A small club where everyone is already in one message group. Thirty members who talk daily do not need a members area, an events calendar and a newsletter signup. They need one page saying when and where they meet, who to contact, and whether visitors are welcome, so a stranger who moves to Mabank can find them.

A church that already keeps its Facebook page current. If the real need is for service times to be findable, one page with times, address, a map and a phone number, plus a Google Business Profile, covers it. A ten-page site with a weekly sermon archive requires somebody to upload sermons every week. In most small congregations that person is enthusiastic for about four months, and then the site says the newest sermon is from last spring — worse than no archive at all.

How we build these

Small, plain, and made to be handed over. We build family, memorial, club and church sites as static pages: real HTML, real image files, no database and no plugins. They load quickly on a phone anywhere around Cedar Creek Lake, cost very little to host, and the whole site is a folder that can be copied to a hard drive and given to whoever takes it over.

As a matter of course we register the domain for a long term in the family's or the organization's name, with more than one person able to reach the account; write a handover sheet on paper saying where everything is and who to contact; deliver a complete backup to you rather than only holding it ourselves; run a privacy pass before publishing; and keep addresses stable, so a link printed in a family newsletter in 2027 still works in 2040. We have been building websites from Kaufman since 2003, for families and organizations around Mabank, Gun Barrel City, Kemp, Seven Points, Payne Springs, Athens and Canton.