Australian retailers approach the EOFY with the usual anticipation—and an unusual amount of caution. Why this mixed bag of emotions? During the pandemic, over 200,000 Australians shopped online for the first time ever, and online sales grew by over 55%. These numbers are good enough reason for retailers to expect huge traffic and transaction volumes throughout this year’s EOFY sales rush—enough to hammer and potentially cripple their IT infrastructure.
That means outages, shaky transactions, and frustrated customers, any of which could cripple business profitability during this lucrative period. Here are five things Australian businesses can do to shore up back-end infrastructure—and prepare for the big event.
1. Establish best-in-class network monitoring
You can’t address what you can’t see. You can’t spot trouble on your network until it’s happened—and by then it’s too late. Therefore, ensure your IT team is equipped with some form of network monitoring capability that gives them full visibility of your entire infrastructure, whether it’s on-prem infrastructure or the cloud.
It’s worth looking past monitoring solutions that track simplistic single-value metrics and alerts—like latencies or traffic volume—and adopt in-depth but scalable monitoring solutions for larger and growing organizations or small business IT solutions that offer deep levels of visibility and alerts into every integrated application, system, and environment within your network. This should provide your business with critical traffic data you need for decision-making, which you’ll be glad to have if and when things start going south.
2. Determine your acceptable baselines
With monitoring in place, your IT team should have everything they need to establish some baselines. Network baselining measures what normal traffic looks like on an average day, so abnormal traffic can be more easily identified when it occurs. Real-time data on things like traffic volumes, device utilisation, memory usage, and network loads are being collected, but overall trends should be noted over time—giving you a frame of reference for what a ‘stable’ network looks like for your business. Of course, robust monitoring solutions do all this for you, but nevertheless, it’s important to ensure you know the information is there to be used.
It’s essentially a red line that, when crossed, means trouble is afoot. Without properly collecting and tracking baselines, your IT team will never know when your network begins to deviate away from stability and into the danger zone—until it cascades into a major outage and leaves your team scrambling and confused.
3. Formulate a plan but expect it to fail
I know this sounds exasperating, but hear me out. Monitoring gives you data, and baselines turn data into actionable information, which your IT team can study to gain insight into what your existing infrastructure can do. Taken together, this should give you everything you need to formulate a solid response plan when EOFY traffic streams in and begins hammering your network.
Except even the best plans can fail. Having been in network management for years, I speak from experience when I say no IT professional can properly predict and plan for all manner of traffic spikes and issues, even with the best tools at their disposal. More important to have is a flexible mindset that’s technically and mentally prepared to put out fires as they occur, mentally note why the fire started in the first place, so changes can be implemented post-event, and remaining cool and calm throughout.
4. Dive into optimising the buying experience
Another upside to having unparallel visibility over your network is the ability to leverage user experience monitoring insights to test and optimise the online experience for customers. I’d recommend a two-step process with synthetic transactions and tracing. Synthetic transactions involve building scripts to simulate user activity on an application or system—you’re essentially testing if your online infrastructure works as intended.
This gives you an idea of the stability of your infrastructure. The only downside? It doesn’t fully simulate real-world user interactions, such as more unpredictable events like EOFY sales. That’s where tracing comes in. Using monitoring data, tracing allows you to gather critical information on how or where users are actually interacting with your systems—giving you a better idea on where to divert precious IT resources.
Tracing data, however, is a lot harder to analyse compared to traditional monitoring data due to its breadth and interconnectivity with multiple systems and applications. Determine if this is something your business would benefit from and begin hiring the appropriate analysts and experts needed to both implement the initial tooling you’ll need and to interpret the data on an ongoing basis.
5. Use monitoring data as the bedrock for practical changes
Network and monitoring data, no matter how insightful or useful, can never truly paint the picture of how the customer experience will pan out. Only when you correlate your monitoring data with customer support or sales would you get a clearer understanding of how your online customers think or feel.
Monitoring isn’t the holy grail to customer satisfaction. There’s still a need to take a temperature check of your other areas of business and use the information you’ve collected to further optimise your EOFY sales experience. You can then utilise your monitoring to see if these changes have resulted in positive outcomes—ensuring your EOFY sales ends in success and not disaster.
Leon Adato is head geek at SolarWinds.