Every June, the cone of uncertainty appears on the evening news and you can watch an entire state change its behaviour inside a day. Plywood moves. Gas lines form. Group chats fill with screenshots of spaghetti models. And in hundreds of small businesses, somebody opens a folder called "Emergency Plan" for the first time since the last near miss and discovers it lists a phone number for an employee who left two years ago.
That folder is not a business continuity plan. A business continuity plan for a small business is a set of decisions made in calm weather, because the decisions themselves are not hard — they are just impossible to make well while you are trying to get your family somewhere safe. If you are in Florida and you have not settled payroll, data access, communications and your actual insurance position before a storm has a name, you have not prepared. You have rehearsed panic.
Businesses don't fail from the wind. They fail in week three.
The failures I have watched up close over the years in this state rarely trace back to the storm itself. Roofs get patched. Inventory gets replaced. What kills a small company is the slow squeeze afterward: three weeks with no revenue and full fixed costs, a receivables book where every customer is dealing with their own damage, a key employee who relocated and did not come back, a landlord whose building will not be occupiable for months, a lease that keeps running anyway.
That is the real Florida business risk, and it is a cash and dependency problem more than a physical one. So the plan you need is not mostly about generators and shutters. It is about how long you can operate with no income, who can move money if you cannot, and which of your suppliers, customers and staff sit inside the same weather system you do. Concentration is comfortable until the day it isn't. If your best vendor, your bank branch, your bookkeeper and your top three accounts are all within the same county, you have built a business that cannot fail gracefully.
Settle payroll before the storm has a name
Payroll is the first thing that breaks and the last thing you can afford to improvise. Three questions, and they should have written answers by the start of the season.
First: can payroll actually be run with no office, no power and nobody at their desk? That means a second authorised approver with real credentials, not a promise that someone will "call the bookkeeper" — and ideally someone who would not be evacuating in the same direction as you.
Second: will you pay people for days you are closed? Decide now, tell them now, and put it in writing. In the aftermath, this is the single question that determines whether your team treats you as an employer worth returning to. Deciding it under pressure, employee by employee, produces inconsistency you will be explaining for a year.
Third: where does that money live? An undrawn line of credit is a promise. Cash in an operating account is a fact. I am biased toward holding a genuine payroll buffer in reserve through the season even when it looks like lazy capital, and I would rather an owner get there by deferring a growth purchase than by hoping. That trade-off is the same one I wrote about in where a Florida business should put its next dollar — resilience is a legitimate use of the next dollar, not a failure to deploy it. Keep some physical cash too. When card networks and ATMs are down, small purchases become a problem you did not model.
Your data is probably fine. Your access to it is the risk.
Almost everything is in the cloud now, so most owners assume disaster recovery is handled. It usually isn't, because the failure mode has moved. The data survives; the ability to reach it does not.
Test a restore before the season. Not a backup — a restore. Pull one real file and one real system back from your backup and time how long it takes and who can do it. A backup nobody has ever restored is a belief, not a control.
Then look at single points of human failure. If one person holds the administrator account for your email, your accounting system, your point of sale and your domain registrar, your company is one unreachable person away from paralysis. Break glass access, documented emergency access in a password manager, a second admin — pick a mechanism, but pick one. And keep a short offline copy of the things you will need while holding a phone with two bars: policy numbers, your agent's mobile, bank contacts, your EIN, your lender, your landlord, your top vendors, your team's personal numbers. Printed, laminated, in the go bag, and also as a photo on two phones.
Communications: one channel you control and one list you own
Hurricane preparedness for a business is largely a communications exercise. Two audiences, and they need different things.
Your team needs a roll call protocol: one method, one designated time, one person responsible for the tally. Say it out loud before the season so nobody is guessing whether to text, call or post. Assume your primary channel fails and name the fallback.
Your customers need to know whether you are open, and they will look for the answer in places you may not be updating. Your website banner, your hours on the map listing, your voicemail, your social profile — these are the places people and increasingly automated answer engines go first. If your published hours say open and your door says closed, you have burned trust you spent years earning. This is the unglamorous side of the work I described in being found by AI as a local operator: the systems that broadcast your status are only as good as the last time a human touched them. Assign one person the job of updating all of them, and have them practise the update in July when the stakes are zero.
One more thing worth deciding early: an out-of-state answering point. Someone outside the cone who can take calls, confirm you exist and pass messages along is a cheap, disproportionately effective piece of continuity.
Read your policy the way an underwriter would
This is where I see the most expensive surprises, and it is entirely avoidable in the spring. Most owners know their deductible number. Fewer know that hurricane and named-storm losses are commonly treated differently from an ordinary claim, and that the applicable deductible may be calculated as a share of insured value rather than as the flat figure they have in mind. Fewer still know whether flood is covered at all — it is typically a separate matter — or what specifically has to happen before business interruption coverage responds.
That last point is the one that catches people. Interruption coverage generally keys off physical damage to your own property. A storm that misses your building but closes your road, kills your power, empties your city for ten days and flattens your three largest customers can produce a devastating revenue loss and a very thin claim. If that scenario would end your business, you need to know it in May, not in September. Ask your agent in writing to walk you through the named-storm deductible, the flood position, what triggers interruption coverage, whether extra expense and dependent-property exposures are addressed, and what documentation a claim will demand.
Then produce that documentation in advance. Walk the premises with a phone on video. Photograph inventory, equipment, serial numbers, the roof, the condition of everything. Pre-loss evidence is what turns a disputed claim into a paid one, and it takes an afternoon.
The review nobody runs
Every storm season generates lessons and almost nobody writes them down. The discipline here is not complicated, it is just unrewarding: within two weeks of a storm or a near miss, while the irritation is still fresh, sit down and list what broke. Which number was wrong. Who you could not reach. What you could not find. What you bought at four times the price because you waited. Then fix three of them and schedule the rest.
I have come to treat this the way I treat any post-mortem on a bad quarter — as the cheapest information available, and worthless if it stays in someone's head. It is the same instinct behind the operating habits in what private equity discipline teaches an owner-operator: write down what you expected, compare it to what happened, and let the gap change your behaviour. Do it annually in the spring, with the payroll question, the restore test, the contact list and the insurance conversation as the standing agenda.
Preparation as an operating habit
The honest value of a continuity plan is not the document. It is that building it forces you to have four or five uncomfortable arguments — about cash, about concentration, about who really has the keys — at a moment when you can still lose those arguments cheaply. Every one of those conversations makes the business better on a completely ordinary Tuesday in February.
We do not get to choose whether the season happens. We choose whether it finds us with decisions already made. In Florida that is not pessimism; it is just what competent operating looks like here.