Files are encrypted. An invoice was paid to the wrong account. Staff cannot log in, or can log in and find mail they never sent. The moment you realise something is wrong is the worst possible time to improvise, and most small businesses improvise because nobody wrote the plan down.
Here is the order that matters.

Step 1 — Stop the Spread, and Do Not Wipe Anything
Disconnect affected machines from the network. Pull the cable, switch off Wi-Fi. That stops ransomware from reaching the next server or the backup.
Do not power the machine off and do not reinstall anything. Doing so destroys the evidence an incident responder needs to work out how the attacker got in and what they took — and if the answer is "we wiped it", you may never know whether client data left the building. That matters later, when you have to decide who to notify.
Do not start deleting suspicious emails or "cleaning up" either. Those messages are evidence.
Step 2 — Call for Help Before You Touch Anything Else
Two calls, early: your IT provider and your cyber insurance carrier.
Most policies include access to an incident response firm, and most require you to notify them promptly. Waiting until you understand the situation is a natural instinct and a common mistake — notice provisions exist precisely because the carrier wants to be involved before decisions get expensive.
Start a written timeline as you go: what you saw, when, who did what. It is tedious in the moment and invaluable three weeks later when you are dealing with the insurer, your lawyer and possibly a regulator.
Step 3 — Change Credentials From a Clean Device
Use a device that was not part of the incident. Resetting passwords from an infected machine hands the attacker each new password as you type it.
Work in order: email accounts first, then any admin accounts, then finance and banking, then everyone else. Turn on multi-factor authentication wherever it is missing — including on accounts that "never mattered", because those are the ones attackers keep.
Then revoke active sessions. Changing a password does not necessarily end a session an attacker already opened, so sign out everywhere and re-issue tokens. In Microsoft 365 this is a single action per account and it is the step most often forgotten.
Step 4 — Verify Your Backups Before You Restore Them
Attackers go for backups first, because a business with a restorable backup is far less likely to pay. So before you celebrate finding a backup, confirm three things: that it exists, that it is recent enough to be useful, and that it is not encrypted or corrupted by the same event.
Restore into a clean environment. Do not reattach the backup store to a network that is still compromised, or you will simply encrypt the backup too.
And never restore over data you have not yet examined — sometimes the "lost" files are recoverable, and overwriting them removes the option.
Step 5 — Work Out What Was Actually Taken
Scope the incident honestly: which systems, which accounts, what data, and whose data. This is what decides your obligations.
Regulated information changes the rules. Patient information brings health-privacy requirements, card payments bring payment-card requirements, and personal information brings state breach-notification law. Your client contracts may impose notice deadlines that are shorter and stricter than any statute.
Assume the attacker read what they had access to. "They only got the file server" is how under-reporting happens.
Step 6 — Let Counsel and Your Carrier Decide Who Gets Told
Notifications are a legal process, not a customer-service gesture. Work with your lawyer and your carrier on who must be told, in what order, and with what wording.
Do not send your own notice to clients or patients first, however well-intentioned. A badly worded notification can create liability that the breach itself did not, and in some cases early disclosure breaks a confidentiality obligation.
Step 7 — Rebuild Rather Than Clean
Once you know how they got in, rebuild. Reinstall from known-good images, patch fully, restore data, then verify. Cleaning an infected machine usually leaves persistence behind — a rogue inbox rule, an extra MFA device, a new admin account, a scheduled task that wakes up in six weeks.
Check those four specifically, because they are how attackers come back after a cleanup that looked successful.
Step 8 — Close the Door Behind You
When the dust settles, write down the root cause in plain language and fix it. In practice it is nearly always one of a short list: MFA that was not enforced, a device or appliance that had not been patched, a successful phishing email, remote access left open to the internet, or backups that were online and therefore reachable.
Then rehearse. A response plan nobody has read is not a plan. An hour spent walking through "what would we do" is worth more than a document in a drawer.
The Part Nobody Wants to Hear
Responding well costs money. Responding badly costs more — in downtime, in lost data, and in the kind of attention that follows a preventable breach. Preparation is the cheaper half of the equation, and it is the half you can still control.
At Virtue Technology Solutions we help Southern California SMBs with cybersecurity, backup and recovery and the compliance obligations that follow an incident. The best time to build a response plan is before you need one.
Want a response plan in place before anything happens? Contact Virtue Technology Solutions for a no-pressure consultation.