If your organization has a seasonal spike in business around the end of the year, you're likely to find demand for compute resources spiking, too. And if your profitability depends on seasonal revenue, you'll need to architect scalable applications with an underlying infrastructure that can adapt to dynamically changing conditions — and you'll need to deploy it quickly because your holiday ads will be running before you know it.
Is your organization ready to meet the demands of all those holiday shoppers who will visit your site? Take the RightScale Holiday Readiness Quiz and see how you stack up before you launch your holiday advertising campaigns.
As a digital marketer, your dream come true is a viral campaign that will drive increased awareness of your brand. But that dream can turn into a nightmare if your success leads to load spikes that cause your website or application to fail. Such a success disaster can lead to publicity, too, but not the kind you're looking for.
High-Profile Success Disasters
Avoiding success disasters is a lesson some organizations learn the hard way. Last year, for instance , customers shopping on BestBuy.com had to wait for nearly an hour to complete the checkout process. The company had to reduce website traffic flow due to an application issue. Visitors to BestBuy.com saw the message, “We were expecting snow but we got a blizzard. Our site is incredibly busy! Please be patient while we shovel you a path."
Last year too, Netflix users who wanted to enjoy a movie on Christmas Eve ended up having to look elsewhere for their entertainment after problems at Amazon Web Services (AWS) caused a day-long service failure. “Terrible timing!” Netflix tweeted at the time, although its customers used far more colorful language in commenting on the situation.
Meanwhile, in Australia, retailers promoted a 24-hour online sales campaign called Click Frenzy that was designed to be a response to the U.S. Cyber Monday. The Click Frenzy site failed almost simultaneously with the start of the event and was unavailable for most of the day. There was a rapid backlash from the Australian public that led to the trending of the #clickfail hashtag on Twitter and the creation of various memes mocking the event.
With almost a third of some companies' annual sales at stake during the holiday season, organizations need a strategy for avoiding a success disaster. Ideally such a plan would also include high-availability (HA) tactics to insure against the inevitable failure of cloud resources and services. And as a bonus it could also provide technical teams with resources to rapidly load test sites and apps to ensure they're fully prepared for the load they may receive.
How Cloud Computing Can Help
|To learn how RightScale helped several organizations leverage the cloud for ecommerce success, download our white paper Avoid Success Disasters by Leveraging Cloud Infrastructure for Your Holiday Campaigns. In it we share four tips that will help you prepare for an increase in website or application traffic during the holidays.|
The cloud can be the prescription for all these ills: It provides scalable resources for both production applications and developer usage; it enables organizations to architect solutions for HA; and it allows for rapid resource allocation in times of high demand and appropriate resource deallocation as demand declines. Without cloud resources, the cost of permanently allocating enough hardware to meet traffic spikes can be prohibitive.
In the traditional hardware model, you can take two approaches to address the issue of the unpredictability of site traffic and system load:
The Cost of Overprovisioning — Inefficient Investment
The first approach is to overprovision: Deploy enough resources to handle any spikes in traffic that may occur. Although this enables you to improve an application's availability in high-traffic situations, it is not an effective use of resources, because a portion (and perhaps the majority) of these resources sit idle before and after each traffic surge.
The Cost of Underprovisioning — Success Disaster
The second approach in the traditional hardware model is to provision for the typical usage pattern of the application and suffer the consequences of lost traffic during peak demand. This is more cost friendly in times of normal usage, but during traffic spikes lost traffic means lost revenue opportunities and damage to the brand. When demand exceeds capacity, traffic is lost, and the application may be unavailable — a success disaster.
The Cloud Alternative
Neither of these approaches is ideal, which is why the scalable cloud model is an excellent fit for a dynamic and unpredictable environment. With the scalable cloud model, you can dynamically provision additional resources only when they are needed and then decommission them when they are not. You incur charges only for the time you use the resources. At no time is there a large capacity excess in which servers sit idle. Even more importantly, at no time is there insufficient capacity to accommodate the demand for the application.
The cloud is a perfect fit for holiday campaigns and their accompanying traffic spikes because of its unique characteristics: elasticity, instant on, high performance, redundant architecture, and pay-per-use cost model. Rackspace customer Spencer's is an example of a retailer that successfully met seasonal demand when it used a hybrid cloud setup for its Spirit Halloween site.
RightScale has the solutions and the expertise to make the 2013 holiday season a huge success for your business. We have worked with hundreds of clients to ensure their success in the cloud. Call us for a free consultation with our cloud experts to talk about the specifics of your project.