Web site crashes can be devastating to a small company, and unfortunately, there’s a limit as to what a small business owner can do to prevent a crash from occurring. However, there are at least a few practical steps your firm can take to decrease the likelihood of a site failure—and to mitigate the damage if one occurs.
First, you should start with an understanding of the potential root causes.
Web sites crash for many reasons. Sometimes, it’s the result of an infrastructure failure, such as power loss, software failures or network problems. These problems can themselves be caused by a variety of factors, such as natural disasters, spikes in traffic, programming or operational errors, and of course, foul play.
Other failures occur due to the site reaching capacity, where the demand on a server simply exceeds its technical ability to perform. We saw this happen several times this past holiday season, particularly in the context of mobile access to e-commerce sites, where the sheer volume of traffic to a site’s server was far in excess of what that server and hosting configuration were designed to handle.
So how do you fortify your site and protect your company?
From a technical perspective, small firms should always prepare for success when building their Web sites. Sites should be built not only to handle the traffic you expect to experience on launch, and not only the traffic that you expect a few months after launch, but also the exponential growth that you hope to experience, say, the next holiday season. This must be kept in mind while the site, especially an e-commerce site, is being built.
From a legal perspective, small businesses should try to obtain contractual obligations from their infrastructure partners to limit the chance of an outage and, if one does occur, to hasten repairs. Of course, to the extent a website is hosted on commoditized, low-cost providers, it may be more difficult to negotiate for meaningful contractual provisions.
What kinds of provisions should a small business seek? First of all, service-level commitments (uptime, throughput, latency, etc.) could be included in the service contract. It is important to charge the service provider with monitoring these metrics so that the provider will be altered to any brewing problem before the site fails completely. A contract might provide that if the metrics aren’t satisfied, the Web site owner might be entitled to credits–thus encouraging the service provider to nip problems in the bud and, if one arises, fix it as fast as possible. Of course, the devil is in the details, and you should pay attention to how the metrics are defined, what the traffic assumptions are, and which viewers might be excluded from the traffic measurement process.
Moreover, small businesses will likely find that a provider that offers disaster recovery and business continuity plans will be in a better position to restore certain outages.
Here are some additional things to consider in connection with website outages:
1. Scalability and Cloud Computing. Flexible web hosting plans employing cloud computing software allow customers to scale up and scale down their services according to company’s needs. For instance, an e-commerce company might pay for more server space during the holiday sales season and return to normal levels afterward. Many companies are employing “hybrid cloud computing” models to address, among other things, latency problems and other performance failures.
2.Testing and backup. Site owners should test to measure the impact of increases in traffic and note whether their servers are consistently running at or near peak capacity. If they are nearing capacity, they should be upgraded.
3.Shared server vs. dedicated hosting. No site owner wants another web business with which it shares server space to overload the server and crash its site too. A dedicated hosting plan can be prohibitively expensive, though.
No matter what, owners should understand what their hosting agreement provides in terms of support: What level of resource is available to the customer? What is the contingency plan for crashes? What “reserve resources” will be available on an on-demand basis? What are the procedures for support during off-hours?
4.Mobile traffic. As more and more people access the Web and shop online from portable devices, retailers and small businesses must have a “mobile-friendly” website that is configured to handle spikes in mobile traffic and avoid outages during peak shopping days.
Jeffrey Neuburger is a partner with Proskauer Rose LLP in the New York office, co-head of the Technology, Media & Communications Group, a member of the Privacy & Data Security Group and editor of the firm’s New Media and Technology Law blog.
This article was written by Jeff Neuburger from The Washington Post and was legally licensed through the NewsCred publisher network.