For most of the businesses that started writing online in the mid 2000s, the blog was not really theirs. It lived on someone else's servers, behind someone else's login, under an address that someone else could change. That arrangement felt like a gift. Setup took an afternoon, the bills were small, and a whole category of technical worry belonged to a vendor. The catch was invisible until the day the vendor had a problem.
TypePad was one of the platforms that made publishing that easy. Six Apart launched it in 2003 as a hosted service, and for a stretch of years it was where a great many company blogs and personal sites actually lived. You wrote, you clicked publish, and the pages appeared. The machinery underneath was not your concern. Right up until it was.
A bad day, and what it revealed
On December 15 and 16 of 2005, TypePad went through a rough one. A storage upgrade, scheduled as routine maintenance, failed. The service was disrupted for roughly eighteen hours. When things came back, some content that had been posted in the two days before the work could not be recovered from backups. Writing that people had published, that they believed was safely stored, was simply gone.
Eighteen hours is not a catastrophe. Sites recovered, people kept writing, and the platform carried on for many more years. What made the episode worth remembering was not its size. It was the shape of the lesson underneath it. Everyone affected had made the same reasonable bet: that the ground their words stood on was solid because a competent company was maintaining it. On those two days, for reasons entirely outside their control, that ground moved.
This is the quiet mechanics of platform risk. When you publish on rented infrastructure, you inherit every decision the landlord makes and every accident that befalls them. Their scheduled maintenance is your downtime. Their backup gap is your lost archive. Their outage is your silence, on a day you may have picked precisely because you had something to say.
The bad day is not the real danger
An outage ends. The more serious version of the same risk is the one that does not. On September 30 of 2025, after about twenty two years, TypePad shut down for good. A platform that had hosted an enormous amount of writing, some of it two decades old, closed its doors. The eighteen hour scare of 2005 was a preview. The permanent shutdown was the feature.
Think about what a closure like that actually does. Every address on that platform stops resolving. Every link anyone ever made to that writing, from other sites, from search results, from old emails, points at nothing. Years of published work, and the search reputation it had slowly earned, evaporates on a date chosen by a company that is winding down. If your business had built its whole presence there, you did not lose a weekend. You lost your home.
Platforms close for ordinary reasons. They get acquired and folded into something else. They stop being profitable. A strategy shifts and a product that once mattered becomes a line item to cut. None of this requires villainy. It only requires that the platform's interests and yours stop pointing the same direction, which, given enough time, they usually do. The internet is littered with the shells of services that were once obviously permanent.
What owning the ground looks like
The remedy is not to swear off hosted tools. Hosted tools are wonderful, and rebuilding a mail server or a content platform from scratch is a poor use of a small company's time. The remedy is to make sure the parts that carry your identity and your archive are parts you control, so that any single vendor can fail or vanish without taking your business with it.
Own the domain
Your domain name is the one piece of this that is genuinely yours. Register it in your own account, keep the registration current, and point it at whatever platform you use today. When you own the address, moving hosts is a change of scenery rather than a change of identity. Every link and every search result that mentions your domain keeps working, because the domain outlives whatever sits behind it. A business that publishes on a platform's own subdomain has handed away the one asset that would have let it move.
Keep an archive you can actually take with you
Before you commit to any platform, find the export. Not the promise of an export, the working feature that hands you your posts, your pages, and your images in a format you can read somewhere else. Run it now, while everything is calm, and keep the result. A backup you have never tested is a rumor. If a platform makes it hard to leave with your own words, treat that friction as the warning it is, because the day you most need to leave is the day they are least able to help you.
Do not build your whole presence on ground you cannot export from
Some platforms are storefronts you rent for a season, and that is fine. The mistake is letting a rented storefront become the only place your business exists. Anything you would grieve to lose belongs somewhere you can pack up and carry out. The rule of thumb is simple. If a service disappeared tomorrow, could you stand your work back up somewhere else by the end of the week? If the honest answer is no, you have found your exposure.
This is the whole reason a company blog reads better living on the company's own site than on a platform's subdomain. The writing earns search authority for an address you keep, the archive stays in your hands, and no vendor's shutdown date is quietly circled on your calendar. It is more setup at the start. It is far less loss at the end. AtRadiant Marketing we treat that ownership as the first decision of any content program, not a cleanup task for later, and the rest of theinsights library keeps circling back to the same idea from other angles.
The 2005 outage cost some writers a couple of days of work. The 2025 shutdown cost anyone who never planned for it far more. Both were the same event at different speeds, and both were survivable for exactly the sites that had kept the deed to their own ground.