Did you, along with billions of others around the world, snag some deals yesterday at the start of Amazon’s Prime Day 2022? It's no secret that July 12-13th marks Amazon’s two-day online shopping event. There are no hard stats yet on this year's numbers, but according to Influencer Marketing Hub, the world’s largest online retailer generated record sales of about $11.2 billion during their 2021 Prime Day, a 7.6% increase from 2020. With over 250 million products sold, there was no need to worry about a shortage of supplies—network bandwidth on the other hand was a different story.
With peak traffic events happening year round, eCommerce and enterprise DevOps teams, now more than ever, must anticipate and prepare for periods of potential high traffic. Online events like Prime Day (not to mention the simultaneous competitor activity worldwide) create the perfect environment for network issues, with consumers simultaneously flooding web pages – creating traffic spikes and potential bottlenecks. And while an increase in consumers can be a great thing for business, a slow network could lead to loss of revenues and a damaged reputation. In fact, 88% of web visitors are less likely to return to a site after one bad experience.
Increased digital complexity speeds up sites, but also creates challenges
Ensuring your website is available, reachable and highly performant has become more complicated over time. Retailers depend on a huge number of third-party services to achieve certain objectives, such as optimizing speed and reliability. Bringing on board one or multiple content delivery networks (CDN) or domain name servers (DNS), for instance, is common practice to achieve dynamic site acceleration and prevent failover.
Most eCommerce sites also depend on a plethora of APIs for things like site search, social proof, personalization, shipping, and price comparison engines. An API that isn’t responding properly or performing very slowly can significantly slow down your site’s performance and ruin customer experience.
Use end-to-end observability across the peak event
The rise in complexity has pushed the need for a more proactive, end-to-end approach to observability, especially when preparing for high-traffic events. Using digital experience observability tools before, during, and throughout periods of high traffic can help organizations stay ahead of network issues and exceed customer expectations.
We’ve prepared an eBook, 5 Ways to Prepare Your Webste for High Traffic, to help DevOps teams gear up for peak periods ahead of time and know what to do during and after the event. The holiday season is no longer a once yearly event. Planning and preparing for peak periods is a continual project. Why not start today?
Recommendations for eCommerce and enterprise DevOps teams
Preparing for times of high traffic may be an obvious need for eCommerce businesses but creating a stable environment for traffic-inducing events also benefits larger enterprises. Corporations who double as online retailers, like Walmart or Apple, are bound to experience times of peak traffic, especially during events such as Black Friday and Tax-Free Weekend. According to Unbounce, 70% of consumers admitted page speed impacts their willingness to buy from an online retailer. Similarly for enterprises that are not retailers but rely on high traffic, such as media and entertainment companies, site speed and reliability makes a difference. During site optimization testing, the BBC, for instance, found they lost 10% of users for each additional second their site took to load.
When experiencing spikes in traffic, the means of preparation between eCommerce and enterprise vary. eCommerce will have a stronger focus on site optimization and efficient use of third-party tags, hosts and requests, while enterprise will benefit from proper setup and configuration of third-party services such as CDN and DNS.
Heavier traffic than normal during peak periods can cost you if you don't properly plan ahead. On the flip side, if you ensure your website is equipped for above-average traffic, you can reap the rewards.
1. Test the performance of third-party tags, hosts, and requests on your web page since not all perform equally well worldwide and some can cause significant performance bottlenecks. Temporarily disable any non-essential third-party adds-on e.g. chatbots or commenting systems.
2. Optimize and streamline your page for performance ahead of time to anticipate the rush.
3. Make sure you're planning for mobile (retailer mobile apps are largely a connection of API requests) and optimize your page and design for speed and ease of navigation.
Enterprises should focus on creating a multiple-CDN strategy to better network and application performance. Testing and utilizing multiple CDNs allow site visitors the same quality of experience, whether they were first to visit or last.
1. At times when your site is likely to be busier than normal (e.g. at the moment of a new software release), think ahead, know your breaking point and set up failover strategies in advance.
2. Consider a multi-vendor strategy, for instance, adopting a multi-CDN approach will allow you to optimize performance and boost failover support when too many customers try to download your new release at the same time.
3. Before key company announcements, warm the CDN cache ahead of time in order to speed up content delivery for those first users who would otherwise receive a slow ‘cache miss’ response. This ensures that every visitor on your site has the same experience and will benefit from a similarly fast page load time.
Run a Site Audit with WebPageTest
- Understand your website performance/site speed today by running a WebPageTest. Is it quick? usable? resilient?
- Find areas to improve your site performance/speed and get optimization suggestions (HTML, image rendering, third party tags) before your performance hurts during a high traffic time.
- Test out how much faster your site can load before implementing the changes. WebPageTest analyzes test results for all users and offers opportunities to improve a site either by running No-Code Experiments or applying suggestions on your own.
Stay Ahead with Digital Experience Observability
Although developments in architecture improve the end-user experience, they do come with a cost. Every additional layer in the delivery chain adds complexity, introduces visibility gaps, and reduces DevOps teams’ ability to understand how infrastructure health is affecting the end-user experience. This means that whenever there is a disruption in the underlying digital infrastructure, IT teams are left scrambling to identify the root cause of the problem.
Digital experience observability gives us visibility into those gaps and allows for faster troubleshooting and a decreased mean time to resolution. Implementing digital experience observability can help maintain optimal website performance, availability and reachability at the times you need it most.