Build Ransomware Resilience Through Backup Strategy and Recovery Planning

Photo of author

By Markus Winkelhock

Resilience is about the restore, not the block

You will probably not stop every attack. Someone will click a link, a contractor's password will leak, or a VPN appliance will go unpatched for a month. Resilience means that when this happens, you're back to work in days rather than weeks, and you never have to consider paying.

That changes where the money goes. Less on detecting every threat, more on making sure the damage is bounded and reversible.

Put two numbers on paper first

Before you buy anything, answer two questions.

How much data can you afford to lose? This is your recovery point objective. If the answer is "a day of orders," you need daily backups at minimum. If it's "an hour," you need something closer to continuous replication for that system.

How long can you be down? This is your recovery time objective. Be honest. Restoring 5 TB over a typical business internet connection of 500 Mbps takes about 22 hours of pure transfer time, before anything goes wrong. That's arithmetic, not a vendor claim, and many teams haven't done it.

A rule of thumb for small and midsize organizations: back up core systems like finance, email, and customer records at least daily, and databases that change constantly every 1 to 4 hours. File shares that change slowly can go daily or even weekly. Rank your systems into three tiers and set a different target for each. Treating everything as critical is how budgets blow up.

To get a rough figure for what downtime costs, add your daily revenue (or daily output) to the payroll you'd pay for idle staff. If that sum is $20,000 a day, a $6,000-a-year backup upgrade pays for itself if it saves you a single day.

Backups the attacker can't touch

Modern ransomware crews go after your backups first. They hunt for backup servers, delete snapshots, and wipe cloud storage using stolen admin credentials. A backup that sits on the same network, under the same admin login, is a convenience copy, not protection.

The long-standing 3-2-1 rule still holds: three copies of your data, on two different types of storage, with one copy offsite. Add a fourth idea to it: at least one copy must be immutable, meaning it can't be altered or deleted for a set period even by an administrator.

Here's where I'll take a side. For most organizations under a few hundred employees, I'd choose immutable cloud object storage over tape rotation or an "air-gapped" drive someone carries home on Fridays. Object lock features on the major cloud storage providers let you set a retention window, say 30 days, during which nothing can delete the data. Tape is fine in theory, but in practice it fails through human habit: the person who swaps the tapes goes on vacation, and nobody notices for a month.

Cost is modest. Bulk object storage generally runs somewhere between $5 and $25 per terabyte per month depending on provider and tier. Prices change, so check current rates. For 10 TB, that's roughly $50 to $250 a month. Compared with a day of downtime, it's small.

Set retention at 30 to 90 days. Attackers often sit inside a network for days or weeks before they encrypt anything, so you want restore points that predate the break-in.

Lock down identity

Most ransomware incidents involve stolen or weak credentials somewhere along the way. Fixing identity gives you more protection per dollar than almost anything else.

  • Multi-factor authentication on everything that touches the internet. Email, VPN, remote desktop gateways, cloud admin consoles. Start with admin accounts, then extend to everyone.
  • Prefer app-based or hardware-key MFA to text messages. Text codes are better than nothing, but they can be intercepted or socially engineered. Hardware keys run about $25 to $55 each, and issuing two to every admin is a cheap, sensible purchase.
  • Separate admin accounts from daily accounts. Your IT lead shouldn't read email while logged in as a domain administrator.
  • Cut the number of admins. If twelve people hold domain admin rights, ask why. A typical small organization needs two or three.

Backup system credentials deserve special treatment. Use a separate account, separate MFA, and don't join the backup server to your main domain if you can avoid it.

Close the doors you leave open

Attackers favor a short list of entry points: exposed remote desktop, unpatched VPNs and firewalls, and phishing emails.

Remote desktop open to the internet should simply not exist. Put it behind a VPN with MFA, or remove it.

For patching, a workable rule of thumb is internet-facing systems within 7 days for critical fixes, and everything else within 30. CISA, the US cybersecurity agency, publishes a Known Exploited Vulnerabilities catalog. It lists flaws that attackers are using right now, and it's a good priority list when you can't patch everything at once.

Email filtering matters too, but don't expect it to be perfect. Assume something will get through.

Limit how far an infection can spread

Once a single laptop is compromised, the question is how much else it can reach. On a flat network, the answer is everything.

You don't need a complex architecture to improve this. Three practical steps:

  1. Put servers, workstations, and guest devices on separate network segments, with firewall rules between them.
  2. Turn off file-sharing permissions that give "Everyone" access. Staff should reach only the folders they need.
  3. Deploy endpoint detection and response software on servers first, then laptops. It can isolate an infected machine within minutes if someone is watching the alerts, or if a service watches them for you.

Monitoring service pricing varies widely. Expect somewhere around $5 to $15 per device per month for managed endpoint protection, as a ballpark rather than a quote.

Rehearse the restore

An untested backup is a hope, not a plan.

Pick a quarter-hour of each quarter, or better, one afternoon each quarter, and restore a real system to a clean environment. Time it. Write down what went wrong. You'll find things: a missing encryption key, a database that restores but won't start, a dependency nobody documented.

Once a year, run a tabletop exercise. Gather IT, leadership, legal, and whoever handles communications. Walk through "it's Saturday at 3 a.m. and every file server shows a ransom note." Who decides to shut down the network? Who calls the insurer? Where's the phone list, and is it stored somewhere that survives the outage? Print it. Seriously.

Keep a clean "break glass" kit offline: contact numbers, recovery credentials, network diagrams, and your insurer's incident hotline.

What I wouldn't bother with

I wouldn't set up a cryptocurrency wallet in advance to be ready to pay. Payment doesn't guarantee a working decryptor, it may carry legal risk depending on who the attacker is, and it marks you as someone who pays. Spend that effort on restoration.

I'd also skip buying a fourth security tool before the basics are done. If admin accounts lack MFA and backups aren't immutable, another dashboard won't help.

Cyber insurance is worth having, but it covers costs after the fact. It doesn't restore your files. Insurers increasingly require MFA and tested backups anyway, so doing the work above often lowers your premium.

A sensible first 90 days

Days 1 to 30: Set your recovery targets. Turn on MFA for admins and email. Remove internet-exposed remote desktop. Start an immutable backup copy of your top-tier systems.

Days 31 to 60: Extend MFA to all staff. Separate admin accounts. Patch internet-facing devices. Run your first restore test.

Days 61 to 90: Segment the network, deploy endpoint protection on servers, write the incident plan, and hold a tabletop session.

None of this is exotic. It's dull, repeatable work, and that's why it succeeds. The organizations that recover quickly aren't the ones with the cleverest tools. They're the ones that can say, with proof, "we restored that last month, and it took six hours."

Leave a Comment