person with a credit card  in hand, about to complete an online ecommerce transaction
PHOTO: rupixen

The spread of COVID-19 has changed our social and economic landscape. The indiscriminate nature of the virus perpetuates fear, affecting the goods and services we buy. Sadly, this pandemic will result in economic winners and losers — an outcome no economist or analyst could have predicted.

We’ve all seen the sobering headlines of stores closing across the nation. Hopefully, the recovery will also be swift, as we learn to adapt to our new reality. For some sectors, however, the crisis has resulted in unforeseen product demand, putting an immediate strain on their IT infrastructure and capacity to deliver.

Virtual Scalability to Prevent Server Overload

IT preparedness and proactive measures are helping some businesses scale during this period of unexpected demand. And while this is a painful time, there are lessons to be learned, and this is one. 

As the coronavirus makes its way through the country, some of our clients are experiencing exponential growth. Healthcare services, precious metals, household goods and wellness products are verticals facing unprecedented spikes in demand. As consumers race to stockpile goods, their websites and infrastructures are under intense pressure.

One client saw a one-day traffic spike of 500%, and another 300% growth in transactions. These are millions of new visitors, cramming their way through a commerce flow urgently trying to checkout or book an appointment.

A pillar of modern architecture is a cloud-based delivery network that relies on content delivery networks (CDNs), caches and APIs to fetch, synchronize and deliver content to users. Over the years, we’ve seen traffic spikes related to Black Friday and other predictable, seasonal variables. But we’ve never seen unanticipated growth like this — nor have our clients. 

Moving to the cloud and hosting content at the edge is proving to be a good strategy. Because CDNs use a distributed system, the cloud is built to self-diagnose when nodes are overloaded and to automatically reroute traffic to available nodes. Self-healing, redundant infrastructure is uniquely able to maintain uptime and deliver virtually unlimited scale.

Related Article: Nipping the AWS Outage Outrage in the Bud

Staying Connected Is Half the Challenge

We all know that speed matters and that milliseconds are critical to the user experience. Maintaining performance, while absorbing traffic surges, is a compounding challenge.

Achieving sub-second speeds is a science that encompasses the entire stack, from ensuring APIs scale and UIs are delivered instantly to eliminating bottlenecks and preventing regression. A core group of techniques and technologies can assure performance, even under the heaviest of loads. The following are a few modern techniques to preserve speed:

  • AMP-as-a-Framework: The front-end framework sets guardrails to guarantee speed. AMP circumvents servers, uses caches for distribution, pages are pre-fetched, and a built-in validator ensures nothing blocks the critical rendering path. AMP replaces heavy and sluggish responsive web design with lightweight portable content units, purpose-built for cloud-based distribution.
  • Server-Side Rendering (SSR): The ability to render and optimize interfaces on the server, instead of on the browser, sends a fully rendered, faster page to the client.
  • Strategic Caching: Caching content close to users, and delivering stale content under select conditions, provides responsiveness, and a protective layer against traffic surges.
  • Headless, API Deployment: After caching static content, and moving it to the edge, APIs serve as data connections that allow the UI to fetch dynamic content and to complete transactions. Deploying APIs for content that cannot be cached — carts, pricing and inventory — reduces total network traffic and enables back-end servers to focus on delivering mission-critical, real-time content.

As traffic unexpectedly spiked, page speed remained unphased
As traffic unexpectedly spiked, page speed remained unphased
The adage “an ounce of prevention is worth a pound of cure” rings true. Without proper IT preparedness, sites crash and opportunities evaporate when you least expect it.

Related Article: Google AMP Is 1-Year-Old Today: Can We Make the Web Even Faster Now?

Now Is Not the Time to Go Dark

With America at home — and online shopping the primary gateway to connect shoppers to retailers — now is not the time to fail. The combination of a serverless delivery network coupled with a modern front-end is proving resilient under tremendous loads. And for the businesses that invested wisely, they are reaping the returns, albeit during a difficult time in our collective history.

The lessons from this moment should be factored into any digital roadmap. Products go viral, marketing campaigns work, and once-in-a-lifetime events happen. IT preparedness keeps websites from crashing and coverts traffic — expected or unexpected — into revenue.

Scalability and performance converted traffic into revenue
Scalability and performance converted traffic into revenue

The coming months will test us all. But we will emerge stronger, smarter and better prepared for the future. We have to. For commerce-driven businesses, we hope this provides a view into how a resilient infrastructure can accommodate for the most unexpected of moments.