There’s very little more frustrating online than trying to access a site that isn’t available because it’s crashed. Unless, of course you own the site that’s crashed, in which case your frustration knows no bounds. It gets even worse if your business relies on your website to generate a profit – losing leads is bad enough, losing money is ever so much worse. The thing is that most websites will crash at least once in their lifetime. It’s almost a rite of passage. Even the biggies like Google, Twitter, Facebook, and Amazon have downtime.
Even though it’s inevitable, you still want answers to some important questions, like what caused the crash, and what can be done to minimise the damage?
Websites crash for several reasons. The most common culprits are:
- A spike in traffic. Some sites get traffic spikes after the launch of a very successful advertising campaign, when a much-anticipated product becomes available, when sales season starts, when a viral video goes truly viral, or even when the website launches, as was the case with the Healthcare.gov site that launched at the beginning of October.
- Not enough testing. According to an article by Lance Ulanoff, editor in chief of Mashable, the company that built Healthcare.gov claimed that they didn’t have enough time to test the site properly before it went live. Don’t let this happen to you. Ensure that your developers are aware of your deadline and that they have included time for testing (and fixing any problems that arise) in the schedule.You can do your part to help developers by not insisting on making change after change after change, and especially by not making any last minute changes or additions. Every change has to be tested, and if you throw in a couple at the 11th hour, well, they may not be tested as thoroughly as they should be.
Load testing is especially important. Laura Strassman says that there are three types of load testing. Load testing 1.0 tests applications inside the firewall. Load testing 1.5 tests these applications in the cloud. Load testing 2.0 tests the delivery chain from the browser to the data centre and includes infrastructure testing and user experience testing. Make sure that your developer has this well covered.
- You don’t have a content delivery network (CDN). According to Mikal E. Belicove, a CDN is a cloud-based scaling service that allows you to deliver online content to millions of website users at a time. When choosing a CDN you need to consider your traffic needs, for example, does your traffic spike seasonally, does it spike whenever you run promotions, is your audience local or international? The latter is important because it’s generally recommended that your CDN has a strong presence in the country where most of your audience is located. This indicates (but doesn’t guarantee) stability.
- You’ve been hacked. Ensure that your security systems are top notch and constantly up-to-date. You won’t ever be 100% safe, but you can make the job as difficult as possible.
- Web outages. As mentioned, it happens to everyone, even website hosts and Google. You can’t do anything to prevent it; you just have to ride it out.
- Plugins. If you use a platform like WordPress to build your website you’ll often be asked to add various plugins to make it run more smoothly. Sometimes this backfires and the plugin isn’t compatible with your site, or it throws a spanner in the carefully balanced works and everything goes haywire. Only ever add plugins that you need, and always do so with the utmost care.
If your website has gone down, the very first thing you need to do is test to see if it really is down or if your device or connection is having a temperamental day. Tecworks recommends that you try to access the site from several different devices, with different connections, and ask some friends in different locations to do the same. If it really is down, then you need to tell someone. The problem could have to do with your website, or it could be related to hosting. Contact your host first, to see if the problem is that side (if they’re on the ball, they may already be working on the problem). If it’s not a hosting problem then get your developer on the phone ASAP.
The next step is to acknowledge the problem and let customers/visitors know. Get onto your social media pages and post an announcement online. Always apologise for the inconvenience and assure the public that you’re working on the problem. Don’t make promises about when you’ll be back online unless you’re 100% sure that you can deliver. Once you are back online, use social media to announce your return.
On Inc.com, Jessica Stillman recommends that you temporarily stop all of your online advertising campaigns. They’re designed to drive people to your website, which is not only a waste when your site is down, but which also risks frustrating users.
Back up data regularly so that if your site does crash, you have the most recent version saved.
One of the most frustrating things about having your website crash is that the problem is out of your hands. You usually have to wait for other people to sort out the situation before you can get back to business. This is one reason why it’s a good idea to maintain good working relationships with your web developer and hosting company. You never know when you’re going to need them to bail you out in an emergency.
- License: Creative Commons image source
Jemima Winslow has been on both sides of the fence; she’s desperately tried to access websites that have crashed and she’s tried desperately to get a crashed site back in working order. So she knows full well the importance of stable, yet flexible web design.