Avoiding instability of major e-commerce platforms over the holidays (Part 2)
Holiday online shopping is growing massively year over year. At this point, the only thing getting in the way of e-commerce businesses and huge spikes in revenue is the availability of e-commerce platforms during critical shopping times.
As we saw in last week’s blog post, during last year’s holiday season, several leading e-commerce business faced major outages, leaving customers unable to purchase their goods online. These included Target and Neiman Marcus, among many others. Despite the uniqueness of the incidents that ultimately caused the outages, and the predictability of massive spikes in seasonal traffic, the consequences of that traffic on service delivery infrastructures are unpredictable. Despite investments in best-of-breed monitoring technologies and improved processes to ensure fluid shopping experiences during the holiday season, we hear about these major e-commerce outages every year.
This year, Forrester has estimated that consumers will spend a record-breaking $112 billion online during the holiday season of November and December — a jump of 13% from the $99 billion spent in the same period in 2015. In November of this year alone, consumers spent over $27 billion online.
On Thanksgiving Day of this year, Adobe data reported that total e-commerce sales in the U.S. would surpass $2 billion. According to Fortune, Target and Kohl’s, two leading e-commerce business, each reported record online sales this Thanksgiving. However, things didn’t go as well for Macy’s, a direct competitor to Target and Kohl’s.
On Black Friday of last week, Macys.com experienced a major outage, leaving shoppers unable to access the site. In the morning, shoppers were informed that there was a temporary shopping jam, yet the outage persisted until the middle of the afternoon. This outage likely contributed to Target and Kohl’s Thanksgiving success.
Expectations on Online Shopping are Shifting
Despite the increasing complexity of IT infrastructures that is making it more difficult to identify incidents, consumers now have higher expectations for online shopping experiences than ever before. According to a 2016 Holiday Retail Insights Report by Soasta, the optimal load time for online shopping conversions could be 2.4 seconds or less. This means that if a page or a transaction takes longer than 2.4 seconds to load, an online shopper will likely leave that site and go to a competitor.
This report also indicates that a one-second slowdown could result in an 8% increase in bounce rate. Further, a five-second slowdown in load time could result in a 42% increase in bounce rate!
IT Lacks the Analytics to Identify Incidents
Whether it’s a major service outage or a minor slowdown in performance, the impact to customer experience and their interaction with a service is astonishing, especially in the world of e-commerce, where immediate gratification is crucial.
Executives at top e-commerce businesses have taken the right step by supporting the adoption of best-of-breed monitoring tools to gain visibility across their productions stacks. However, the mistake is making the assumption that their existing operations and support teams are capable of analyzing the millions of events that are produced each day across their production stack. These teams need to leverage algorithms to provide the analytics in real-time, to understand any abnormality before service is impacted.
For more information on how Moogsoft addresses this space, check out Moogsoft’s AIOps solution.
About the author
Sahil Khanna is a Sr. Product Marketing Manager at Moogsoft, where he focuses on the emergence of Algorithmic IT Operations. In his free time, Sahil enjoys banging on drums and participating in high-stakes bets.