“Our site is experiencing higher traffic than usual, please check back soon.” We’ve all seen it, the dreaded “sorry for the inconvenience” alert while attempting to purchase from your favorite website, book a flight or buy tickets to that concert that’s about to sell out. In some cases, time is of the essence and in our on-demand consumer culture, when we don’t get what we want, when we want it, then it’s on to the next one.
Never are these issues more disastrous for B2C companies than during the holiday season, when consumers rush by the millions to e-commerce sites in hopes of finding the gifts they want at a dramatically reduced price. During this time, specifically between Thanksgiving and New Years, online retailers have the potential to make up to 40% of their yearly revenue. Yes, you read correctly, nearly half of annual profits in about five weeks. Not too shabby. But what happens when your customer’s shopping cart stalls, or worse, your website goes down on the biggest shopping day of the year? When you consider that the average online shopper is 40% more likely to leave a site for a competitor if it takes longer than 2 seconds to load, lags in performance, simply aren’t an option.
As of June 10, we are 169 days out from Black Friday 2015, and 172 days from Cyber Monday. Are you ready?
How can e-commerce companies prepare themselves for the upcoming shopping frenzy
1. Go hybrid (cloud, that is) with confidence
While your private cloud or internal infrastructure may provide enough capacity to satisfy customer demand during the rest of the year, it may not be enough during the holidays.
One study found that online retailers can see a 200% increase in traffic during the Thanksgiving holiday weekend alone, a number that could even jump to 800% if your marketing team is stepping up their game. One of the most cost effective ways to meet this need is by temporarily leveraging a public cloud platform, scaling out workloads to meet increased demand without overspending on expensive infrastructure resources that sit unused for the better portion of the year. However, it doesn’t come without it’s own set of difficulties.
As discussed in previous blog posts, leveraging the public cloud isn’t as simple as bursting workloads to the cloud. Because e-commerce applications are made up of many interconnected parts, from the web and app servers to the database servers, moving one component to the cloud and leaving the rest on-premise could have disastrous effects in terms of latency and application response time.
Public clouds hide supply and give only limited information on demand – making it difficult to manage effectively as you scale workloads to meet fluctuating demand. The only way to get the most out of your investment in the hybrid cloud is to know exactly which workloads to burst, where to burst them and when to do so. With a holistic view of the environment, including workloads within the public cloud, VMTurbo’s actions allow users to do just that, effectively utilizing the public cloud while meeting the ever-increasing needs of their customers.
2. Combat latency across the stack
Latency can occur at every layer of the stack from the application layer, across the network and down to the storage, at any time, for a number of reasons. It’s impossible, and far beyond human scale, to solve for all sources of latency at once and maintain perpetual health. And when business-critical applications, such as online shopping carts and financial transactions platforms, don’t have constant access to the resources they need, customers are sure to notice the decreased performance.
VMTurbo understands the application demand for compute, storage and network resources and is able to make decisions and manages the tradeoffs in virtualized and hybrid cloud environments. Tradeoffs like how to make sure frequently communicating application tiers (e.g. database and app server) are placed close together to avoid network latency without sacrificing sufficient access to compute and storage resources.
These decisions avoid the resource congestions that cause latency in the first place while making sure you get the most out of your investment in the underlying infrastructure.
3. Assure quality of service
You can’t forget that statistic mentioned earlier. Two seconds is about all you have to satisfy online shoppers. That’s seven seconds less than a goldfish’s attention span…If customers aren’t getting the level of services they want from a vendor, they will very quickly make a decision to leave, ultimately impacting both revenue and reputation.
But when there are so many variables at play, how could you possibly control for the end-user’s experience? VMTurbo enables its users to define application QoS targets, including transactions per second and average response time, and meet them with real-time decisions to scale up or out as demand fluctuates.
Whether it’s Black Friday or Cyber Monday, or just an average weekday, QoS Adherence automatically adjusts to meet the fluctuating demands of virtual workloads.
For more on how companies can assure application performance and combat latency during high-traffic periods such as Black Friday, download our latest whitepaper.