Website Backups: A Small Business Guide

Why backups fail small businesses

Almost every small business owner we speak to believes they have backups. Most of them do, in some form. What they do not have is a tested, current, independent copy they can actually put back on a bad day, and the difference only becomes visible during the bad day itself.

There are several ways a backup quietly stops being a backup, and they are worth naming because they are all avoidable.

What actually needs backing up

A modern website is not one thing. It is several moving parts that must be captured at roughly the same moment, or the restored copy will not match itself. Here is what each part holds and why skipping it hurts.

The database. On WordPress, or almost any content-managed site, the database is where your pages, posts, products, orders, customer accounts, form entries, menus, widgets and plugin settings live. It is usually small, often only a few megabytes, and it is the single most valuable file you own. Losing it means losing the actual business content while keeping the empty shell.

The uploads folder. Every image, PDF, brochure and downloadable file you have ever added sits here. On a mature site this is normally the largest part of the backup by size and the slowest to move. It is also the part cheap backup tools most often skip in order to keep archives small.

Themes and child themes. If any custom CSS, template edits or design work was done directly in theme files, that work only exists here. A fresh install of the same theme will not bring it back.

Plugins and extensions. The plugin code itself can usually be re-downloaded, but licensed or custom-built plugins often cannot. Any plugin you paid for, or that a developer wrote for you, is unique and irreplaceable.

Configuration files. On WordPress that means wp-config.php and .htaccess. These carry database connection details, security keys, redirect rules and performance directives. They are tiny, easily missed, and painful to reconstruct from memory.

Things that live outside the site entirely. Your DNS zone, your email routing, your SSL setup, your payment gateway keys and your domain registrar login are all part of a working website and none of them are inside a site backup. Keep a written record of where each one lives and who controls it.

Full, incremental and differential

Three words get used loosely by backup vendors. Knowing what they actually mean helps you read a pricing page honestly.

In practice most managed backup services run a full baseline plus continuous incrementals and hide the mechanics from you. That is fine, and usually the right choice for a small business, as long as the service verifies the chain and can produce a single downloadable archive on request. If you cannot download a complete copy on demand, you are renting access to your data rather than holding it.

“If you cannot download a complete copy of your own website on demand, you do not have a backup. You have a subscription.”

The 3-2-1 rule, applied

The 3-2-1 rule comes from general IT practice and translates cleanly to small business websites. Three copies of the data. On two different types of storage. With at least one copy held off-site, away from the system running the original.

1. Copy one: the live site. This counts as one of the three. It is the working original, not a backup, but it is one instance of the data.
2. Copy two: the host or plugin backup. Usually stored on or near the hosting infrastructure. Fast to restore, convenient for everyday accidents like a bad plugin update.
3. Copy three: independent, off-site storage you control. A cloud drive, object storage, or even an external hard drive on a monthly schedule. Crucially, the account must be yours, not your developer’s and not your host’s.
4. Two storage types. Server disk plus cloud object storage satisfies this. Two folders on the same server does not.
5. One truly off-site. Different provider, different login, different failure mode. If your hosting company has a bad week, this copy is untouched.

How often should you back up?

Frequency is not a matter of preference. It follows from one question: how much work are you willing to redo? That number is your recovery point objective, and it decides the schedule.

There is a second number worth writing down beside it: your recovery time objective, meaning how long you can be offline before the damage stops being an inconvenience. A restaurant taking bookings on a Friday has a very different tolerance from an accountancy practice with a static brochure site. Knowing both numbers turns backup shopping from guesswork into a specification.

Where to store copies

Storage destination matters more than most owners expect, because it decides whether the copy survives the event that destroyed the original.

Backup methods compared

Four broad approaches cover almost every small business site. Most owners end up using two of them at once, which is exactly right.

The combination we recommend to most small businesses is simple: keep whatever your host provides, add one independent tool pushing to cloud storage you own, and take a manual copy before any significant change. That gives you speed for small mistakes, independence for large ones, and a known-good point immediately before the riskiest moments.

Why your host’s backup is not enough

This deserves its own section because it is the assumption that causes the most damage. Host backups are real and often good. They are also subject to conditions that tend to appear in the small print rather than the sales page.

1. Retention is usually short. Entry-level plans commonly keep only a handful of days. Damage discovered later than the window is unrecoverable from that source.
2. Restores can be all-or-nothing. Many host restores replace the entire account. If you need to recover one deleted page from Tuesday without undoing Wednesday’s orders, that is often not offered.
3. Availability follows account status. A suspended, expired or disputed account can mean no control panel, which means no restore button, at precisely the moment you need it.
4. Shared fate. A serious incident at the provider affects the live site and the provider-held copies together. This is uncommon, and it is exactly the scenario off-site copies exist for.
5. Terms often disclaim responsibility. Read your own hosting terms. Many state plainly that backups are provided as a courtesy and that the customer remains responsible for maintaining their own copies.

Retention: how far back to keep

Depth matters as much as frequency, because the worst problems are the quiet ones. A defaced homepage is noticed in an hour. A malicious script injected into a template, a corrupted product import, or a staff member deleting the wrong category can go unnoticed for a fortnight. If every copy you hold is newer than the damage, every copy contains the damage.

A retention pattern that works well for small businesses without becoming expensive is a tiered one, sometimes called grandfather-father-son.

1. Keep every daily backup for the last 14 to 30 days. This covers ordinary accidents and most quietly discovered problems.
2. Keep one weekly backup for the last 8 to 12 weeks. This gives you a fallback point well behind any slow-burning issue.
3. Keep one monthly backup for the last 12 months. Cheap, because it is only twelve files, and occasionally the thing that saves a business.
4. Keep one annual archive indefinitely. Useful for record keeping, disputes, and reconstructing what the site said on a given date.

Testing a restore

This is the step almost nobody does, and it is the step that turns a copy into a backup. A restore test answers three questions: does the archive actually open, does the restored site actually work, and how long did the whole thing take.

1. Create a staging site. Most decent hosts offer one-click staging. If yours does not, a cheap separate hosting account or a local environment works equally well. Never test a restore over your live site.
2. Restore the most recent backup into it. Use only the archive, not the live site, and resist the urge to fix anything by hand as you go. The point is to find out what breaks.
3. Walk the critical paths. Load the homepage. Log into the admin. Open a product or service page. Submit the contact form. Complete a test checkout if you sell. Check that images load rather than showing placeholders.
4. Check the edges. Look for missing menus, broken redirects, lost custom fields and plugin settings that reset to default. These are the usual casualties of an incomplete archive.
5. Time it and write it down. Record how long the restore took end to end. That number is your true recovery time, and it is the figure to plan around rather than the optimistic one on a vendor page.
6. Repeat quarterly, and always after changing hosts, backup tools or storage destinations.

“A backup you have never restored is a theory. Test it once and it becomes a plan.”

The recovery playbook

When something goes wrong, the instinct is to start clicking. Resist it. The first ten minutes decide whether you have one problem or two, and the most common way a small incident becomes a large one is a panicked fix applied on top of it.

Scenario one: an update broke the site. Usually a white screen or a layout collapse after a plugin, theme or platform update. This is the easiest case. Restore the most recent backup taken before the update, then reapply updates one at a time on staging until you find the culprit. If you cannot reach the admin at all, deactivating plugins by renaming the plugins folder over SFTP will normally get you back in.

Scenario two: content was deleted. A page, a batch of products or a category is gone. A full restore would undo everything else since, so prefer a partial one: restore the backup to staging, export just the missing content, and import it into the live site. This is where having a downloadable archive rather than a vendor-only restore button really pays.

Scenario three: the site was hacked. Restoring a clean backup is necessary but not sufficient, because whatever let the attacker in is usually still there. Work in this order: take the site into maintenance mode, change every password including hosting, admin, database and FTP, restore from a backup taken before the earliest sign of compromise, update everything to current versions, then scan before going live again. If payment or personal data may have been exposed, treat notification obligations as a separate and urgent question for a qualified adviser.

Scenario four: the host is down or gone. This is what the off-site copy exists for. Provision hosting elsewhere, restore the independent archive, and repoint DNS. The bottleneck is almost never the restore, it is DNS propagation and the fact that nobody remembers where the domain is registered. Keep registrar and DNS access documented separately from the site itself.

Scenario five: the developer relationship ended. Not a technical failure, but the same recovery problem. If the hosting, domain, backups and storage accounts were all in someone else’s name, you are locked out of your own business asset. Fix this before it happens by making sure every account is registered to an email address the business controls.

Backups and security

A backup archive is a complete copy of your database. That means it contains user accounts, hashed passwords, customer details, order history and any personal data your site collects. Treated carelessly, your backup becomes the easiest way into your business.

1. Never store archives in a publicly reachable folder. If a URL can reach it, so can an automated scanner.
2. Encrypt archives at rest where your tool supports it, and record the key somewhere other than the site.
3. Use a dedicated storage account rather than the personal cloud drive you also share holiday photos from.
4. Limit who can restore. Restore access is effectively full access to the business.
5. Keep at least one copy immutable or versioned so that ransomware or a rogue login cannot delete your history along with your live site.
6. Turn on failure notifications. A backup system that cannot tell you it has stopped working is worse than none, because it produces false confidence.

What this costs

Backup is one of the few areas where the sensible option is genuinely cheap, so the honest answer is that cost is rarely the real obstacle. Free backup plugins pushing to a cloud drive you already pay for will cover a brochure site adequately. Object storage for a few gigabytes costs a very small amount per month. Managed third-party backup services with real-time capture and one-click restores sit in the range where a single prevented incident pays for several years of subscription.

We are deliberately not quoting figures here because plugin pricing, storage pricing and hosting plans change constantly and vary by region and currency. Check current pricing directly with the provider before budgeting. What does not change is the comparison worth making: put the monthly cost beside the cost of rebuilding your site from scratch, including the lost enquiries during the rebuild. For almost every business, that comparison ends the discussion.

Your backup checklist

Frequently asked questions