How much does downtime cost a small business? A way to estimate it
A practical way to estimate what an hour of website downtime costs your business, so decisions about hosting, monitoring and response are made with a number.
The short answer
Downtime costs a small business lost sales and enquiries, staff time spent fielding questions and chasing the fix, the recovery work itself, and a harder-to-measure loss of trust from everyone who saw the error. Without an estimate of the figure a business either underinvests in preventing downtime or, occasionally, buy infrastructure far beyond what the number justifies. The estimate is simple enough to do in an hour: work out what one hour of downtime costs at a busy time and at a quiet time, multiply by the realistic annual downtime for your kind of setup, and compare that with the cost of monitoring, a fast response arrangement and better hosting. The result justifies monitoring and a response arrangement many times over, and rarely justifies enterprise infrastructure.
Building the estimate
| Component | How to estimate it | Notes |
|---|---|---|
| Lost revenue | Average revenue the site produces per hour at that time of day, from orders or from leads times close rate times value | Use real analytics and sales data, by hour |
| Lost leads that do not come back | Share of would-be enquiries that go to a competitor | Higher for urgent services; lower for considered purchases |
| Staff time | Hours spent by everyone dealing with it, at their cost | Phone calls, emails, coordination, apologies |
| Recovery cost | The partner’s or supplier’s time to diagnose and fix | Higher at night and at weekends without an arrangement |
| Follow-up cost | Making good with affected customers, explaining, discounts | Often the largest item for stores |
| Reputation | Hard to quantify; note it and decide whether it changes the conclusion | Media moments and launches multiply it |
| Advertising waste | Paid traffic sent to a page that was down | Campaigns make downtime expensive per minute |
The calculation
- Pick two hours: your busiest and a quiet one. Estimate each component for each.
- Weight them by how much of the year is busy versus quiet, to get a blended hourly cost.
- Estimate realistic annual downtime for your current setup, honestly, including the time before anyone noticed.
- Multiply to get the annual cost of downtime as things stand.
- Price the prevention: monitoring with alerts, a response arrangement with defined times, a platform with rollback and previews.
- Compare, and also compare against the enterprise option, which usually loses.
- Decide and write the decision down with the numbers.
A worked shape
A service business whose site produces a few qualified leads a day, each worth a meaningful sum with a reasonable close rate, finds that a busy-hour outage costs a few hundred euros in expected lost business plus staff time, and a night-time outage costs almost nothing. Its unmonitored shared hosting has gone down perhaps five times in two years, each for several hours before anyone noticed. The arithmetic says monitoring and a response arrangement, at a modest monthly cost, and a move to a platform with rollback when the site is next rebuilt. It does not say enterprise hosting, and it does not say do nothing.
What this means for you
Put a number on an hour of downtime, busy and quiet, multiply by realistic annual hours for your setup, and compare it with the cost of monitoring, a response arrangement and a platform that recovers in minutes. The result usually says: monitor, arrange the response, and choose the architecture that rarely goes down, without paying for infrastructure built for a different scale.
Frequently asked questions
Our website is mostly a brochure. Does downtime really cost us anything?
Less than a store, but not nothing: the enquiry that went to a competitor, the customer who checked your opening hours and could not, the job applicant who gave up, the impression left on anyone who saw the error. Estimate it honestly and it is usually modest per hour but real, which justifies monitoring and a response arrangement, and does not justify much more.
What if downtime happens at night?
Then it costs what your site would have produced at night, which for many businesses is little, unless you serve other time zones or run a store. That is why the estimate is per hour and per time of day, and why the response arrangement can distinguish between business hours and the middle of the night for a brochure site while treating them the same for a store.
How do we use the number?
Compare it with what prevention costs. If an hour costs a few hundred euros and realistic annual downtime on your setup is several hours, monitoring that catches problems in a minute and a response arrangement that fixes them in an hour pay for themselves quickly. If the number is small, spend proportionally. The point is to decide with a figure rather than a feeling.