A Tornado Gives You Thirteen Minutes. Your Continuity Plan Was Decided Years Ago.
July 23, 2026 · Alex Weeks · Cloud & Infrastructure, Title Industry
Our headquarters is in Oklahoma City, so I’ll skip the abstract framing: the average tornado warning gives you about thirteen minutes. That’s enough time to get your people into the shelter, and it is enough time for absolutely nothing else. Nobody is grabbing the backup drives. Nobody is starting one last sync. Coastal operators get days of forecast cone to move closings up and improvise; here, the entire disaster-response phase of the plan is “everyone goes to the safe room.” Which means that for a title company in tornado country, business continuity isn’t a response plan at all. It’s an architecture decision — and by the time the sirens go, you made it years ago whether you know it or not.
The Building Is the Least of It
When owners think about storm risk, they think about the physical office — and the office is the most recoverable thing on the list. You can work from kitchen tables for a month. What you can’t improvise around is where the operation actually lives: the server in the closet running your production system, the title plant that exists on one machine in one building, the paper files for every open order, and the escrow and closing pipeline that doesn’t pause because you did. Anyone who’s lived here through a May like 1999 or 2013 knows what an EF4 does to a commercial strip — it doesn’t damage the contents of a building, it removes the concept of the building. A title company that loses its plant and its production data hasn’t lost a month. It’s lost the asset the business is built on. That’s the difference between an interruption and an extinction event, and thirteen minutes of warning has no influence over which one you get.
Two numbers from the IT world are worth stealing, in plain English. Recovery time: how long until you can work again? Recovery point: how much data do you lose — everything since last night’s backup? Since Friday? Ever tested either one, or are they both guesses? For an operation whose plant and production system live on-premises, with a backup drive that gets rotated to somebody’s house when somebody remembers, the honest answers are usually “weeks” and “we’re not sure.” And notice that the backup drive at an employee’s house is a hurricane-era control anyway — a tornado that tracks through your part of town has a real chance of visiting both addresses in the same four minutes. The April post on cloud title plants made the operational case for getting the plant out of the building; this is the continuity case, and it doesn’t care about your ROI spreadsheet. Cloud infrastructure doesn’t make disaster planning unnecessary — it makes the worst categories of loss structurally impossible. The plant isn’t in the building. The production data isn’t under the debris field. Your recovery point is minutes, and your recovery time is however long it takes your team to find a connection.
The County Takes the Same Hit
The part of the plan almost everyone skips: your continuity is only half the equation. The same storm that misses your office by a mile can put the county courthouse inside the damage path — and plenty of Oklahoma, Kansas, and Nebraska counties are still running recording operations that live substantially on paper and on servers in that one building. When a recorder’s office goes down, recording queues stop, and when it reopens it’s working a backlog with reduced staff — the recording bottleneck I wrote about in June, except with a debris field. That means gap risk on every file in flight: the search is done, the closing happened or didn’t, and the recording window just stretched from days to weeks. Your plan has to cover the operational questions — which closings pause, how you proactively tell lenders why, and what your underwriter’s position is on gap coverage during a declared disaster. Have that conversation this month, not from a parking lot the week after.
The same logic applies to your vendors. Your production system, your plant provider, your document repository — where does each of them physically live? A regional vendor running everything out of one data center in the same metro as your office hasn’t diversified your risk; it’s doubled down on it. Ask now. The answers are only embarrassing before the storm.
Run the Drill Without the Weather
Since the warning gives you nothing to work with, the drill is purely a tabletop exercise, and it costs one page and an hour. Assume it’s a Tuesday in May. Assume the office and everything in it is gone at 4:40 p.m., and everyone walked out of the shelter fine. Now walk it forward honestly: can your people work tomorrow? From where? Is the plant alive? The open orders? The escrow accounting? Who calls the underwriter, and what exactly do they say? Every question that produces an awkward silence is an item on the list, ranked by how expensive the silence was. Some items are process and cost nothing. The big ones are architecture — and architecture is precisely the thing thirteen minutes cannot buy you.
