Black Friday and Cyber Monday are the two biggest online shopping days of the year. They are also the most highly anticipated days for our hosting team, as we generally see triple and sometimes quadruple the traffic levels compared to any other given day. This year we saw by far the highest website traffic levels ever, including over 6M page views and over 15k online purchases. Below are 5 steps we followed to make the holiday season successful for our e-commerce clients:
1. Consult with the Client
First thing to do is consult with your e-commerce clients and ask lots of questions like:
It may seem obvious, but the more specific information you can obtain from the client the better prepared you will at handling peak traffic spikes.
2. Work Hand in Hand with your Development Team
The hosting and development teams must work together in order to determine a website's potential bottle neck(s) or single points of failure. This is a crucial step, because without having full visibility into the platform, neither team would be fully equipped to deal with any potential problems. Lastly, both teams should test the environment, when possible, using load testing tools such as LoadImpact to discover and fix any potential problem areas.
3. Discuss Scalability Plans with your Service Providers
It doesn't matter how well a plan looks if it can't be executed by the deadline. It's important to discuss these plans with your service providers and hosting partners, like Rackspace, to ensure what you have planned is actually possible within a given time frame. There are always nuances that the service providers don't divulge in their marketing & sales materials and it's better to find out well before the eve of your deadline.
4. Plan & Scale Up Early
Plan early and beat the rush. We've learned to start planning for holiday traffic around the end of the 3rd Quarter. This not only gives you plenty of time to prepare and execute your plans, it also gives the client time to arrange for any additional budget that might be required. Beat the rush and start scaling up well before Black Friday. We start scaling up the week before in order to ensure our service providers can provide for the additional capacity we require.
5. Build in Redundancy and Plan for Failure
System failure is inevitable in the cloud. Knowing this, redundancy should be one of the focal points of your scalability plan. This means to plan for not only server redundancy, but also for service provider redundancy. If you rely on a single provider for everything (hosting, monitoring, alerting, etc), then you have a single point of failure that is highly vulnerable. Like investing, diversification is crucial to building a successful redundancy strategy.