Downtime Metrics
Before you can work on improving your website's performance, you have to get an accurate view of the current state. This should be done by tracking certain metrics over an extended period of time, such as monthly. Then you can analyze the data to determine where your weaknesses are and how they can be addressed.
The first key metric when it comes to online performance is uptime, which is simply a measure of how often your website is alive and responding. To calculate it, divide the number of minutes of uptime by the total number of minutes in a month (43800) to arrive at the uptime percentage. Companies should strive for 100% uptime, though anything close to 99.99% is considered high-performing.
But just having a stable website is not enough to support an online business. Slow page loading times can do as much damage as a full outage because customers will not have patience when there are so many shopping options available. For this reason, you should also track your time to first byte and total page load time metrics.
Time to first byte measures how long it takes your webserver to deliver the first piece of a response to the user's browser. Every element of your network and infrastructure can have an impact on this metric. We’re talking about local internet speed and even seemingly insignificant things like plugins, if you happen to use a content management system (CMS).
Calculating Downtime Cost
Let's assume you've built your online enterprise into a full-fledged business that operates around the clock. How can you calculate the exact financial pain that website downtime will cost you? Across all industries, it's estimated that every minute of an IT outage results in $5,600 of revenue lost, which totals $336,000 per hour and over $8 million per day.
For certain types of companies, the money lost can be even more dramatic. As an example. major global retailers like Amazon rely on having their website available and responsive at all times. It's possible that a single minute of downtime could cost them upwards of $200,000.
To perform an exact calculation of hourly downtime cost, you first need to tabulate your company's monthly revenue. Then subtract your average uptime percentage from one and multiply it by the monthly revenue figure. The total will be your estimated lost earnings on a monthly basis, assuming your uptime metrics do not improve or worsen.
Then consider banks, credit cards, and other financial institutions that use the internet as their primary source of customer interaction. An hour of downtime might not result in lost sales, but it will create endless hours of work for developers and database administrators to clean up the mess and ensure that no records were affected by the outage.
We've already discussed lost revenue and recovery costs, but the full equation for calculating the impact of downtime involves two other categories as well: lost productivity and intangible costs. These are harder to assign an exact dollar but they should not be ignored.
Companies plan their daily operations and larger projects under the assumption that IT systems will be functioning at their normal levels. But when an outage occurs, team members from across the organization must scramble to react and diagnose the problem. Their combined hourly salary should be multiplied by the length of the outage to find the lost productivity cost.
Lastly, there are intangible costs which can have lasting effects even after an outage has been resolved and all websites are restored to normal. To examine one real-world example, these days it’s not unusual for a business to have automated appointment reminder software plugged into a digital calendar. It’s a simple, reliable way to have clients automatically reminded of appointments - except when the website goes down and reminders go out late or not at all.
The result is frustrated customers who (fairly or not) have come to rely on your system. When it breaks down, your company's reputation may suffer damage or faithful customers may switch to a competitor during the downtime.
Causes of Outages
The only way to start improving your website's performance is to understand where the issues are coming from. Downtime can usually be categorized into two main areas: outages caused by internal factors or outages caused by an outside factor like malware deposited by a cybercriminal. Online businesses need to be prepared for both.
Downtime caused by internal problems usually originates with a change to code or infrastructure that had unexpected dependencies and impacts. For example, a change of IP address at the domain level requires updates to all backend servers, including databases. If the needed changes are not synchronized, then your website could go down for a period of time.
Even if your internal systems are well designed, there is always the threat of a hacker trying to attack your website for monetary gain. It's critical for companies to employ a suite of cybersecurity tools and products to protect themselves from common scams like phishing and ransomware. In a worst case scenario, hackers gain access to your internal servers and steal data while also keeping your website offline.
Value of Testing
As discussed, there is pressure in the online retail world to move quickly and keep up with the competition. But one mistake that many companies make is to rush out new code and features to their website without thorough testing. This can result in a wide range of issues that often lead to downtime.
To follow best practices, your quality assurance team should use a rigorous process for regression testing, which checks all new code and changes against a set of scenarios to ensure no negative consequences. In addition, testing should occur on a separate staging environment and should not be promoted to your live production systems until all checks are passed.
The Bottom Line
Online shoppers in the modern world are looking for more than just low prices. They only want to spend their time on websites that are secure, trustworthy, and quick to load. Otherwise, they will simply find an alternative. For retailers, this underscores the importance of uptime and website performance. Failing in either category can result in a drop in sales that will doom the business overall.
Maintaining a retail website is hard work, even in the cloud era when hosting providers offer many services. At the end of the day, you are responsible for ensuring that your customers have a positive experience using your website on an ongoing basis.