<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1642713012720816&amp;ev=PageView&amp;noscript=1">

Latest News from 100TB

Get it from the source

6 Steps You Should Take Now To Prepare For Seasonal Ecommerce Spikes

Published Sep 14, 2017 by Sarah Chambers

Black Friday is coming. Is your site ready for the chaos? Last year, 40% of UK eCommerce sites experienced downtime during seasonal peaks. Being unable to handle the incoming flood of excited holiday shoppers means putting a lot of revenue at risk. Last year, the Black Friday to Cyber Monday online shopping frenzy brought in over $12.8 billion dollars.

If your site goes down, customers won’t wait around for long. They’re also far less likely to return in the future as a result of this. 79% of customers say they are unlikely to purchase from sites with unsatisfactory performance (KissMetrics). Optimal site performance keeps customers coming back time and time again.

Your marketing team has sent out the promotions, the warehouse team is stocked up. Customer service is ready and waiting. The ability to scale your technology to the challenge should not be the weakest link in your chain.

Ready to prep for the busiest time of the year?


Here are 6 steps your tech team can take now to prepare for traffic spikes:

 

Get mobile ready

Black Friday was traditionally a boon for brick-and-mortar stores where customers would line up outside for hours to take advantage of doorcrasher deals. But consumers are starting to blur the line between in-store and online shopping. 83% of consumers use their smartphone while they are shopping to look up prices, reviews and competitors.

This means that over half of your site traffic is likely coming from mobile devices. It’s worth checking how your site performs on different devices and over a mobile data connection to optimize conversion from these visitors.

  • Using a tool like BrowserStack, confirm that your site is responsive and loads properly on mobile browsers.
  • Implement digital wallets like PayPal or Apple Pay to make it easy for customers to check out.
  • Keep designs optimized and first load time quick to make sure customers browsing on a mobile connection have a good experience.

Plan to scale horizontally

Unfortunately, there’s no option to “turn it up to 11” when it comes to vertically scaling servers. Operating at near-full capacity comes with a set of risks, especially if your business is seasonal. While it might be more cost efficient in the short term, if volume spikes suddenly, it will be a real hassle to scale.

Containerization is a great way to accommodate future scaling needs, which is important if you might need to deploy more resources quickly. By hosting microservices within a container, you can scale horizontally simply by deploying a new container image. They deploy and spin up almost instantly, without requiring additional installation packages to be run. Containers also work effectively with load balancers to ensure no server is overwhelmed.

Containerization allows developers to operate with just-in-time scaling, knowing that every deployment will execute quickly and accurately.

Setting up containers takes time and forward planning. However, the alternative is a few frantic late nights when things get busy. You don’t want to be in the position where you are  updating architecture to accommodate volume during Cyber Monday.

 

Embrace regional redundancy

The further away a customer physically is from your server, the slower response time they will see when browsing your site. Waiting to load images from a server halfway across the world will put off even the most dedicated shoppers.

You have a few options to improve delivery for your global customers. Probably the least efficient option is to set up multiple servers on different continents. The biggest problem with this method is the overhead costs in maintenance and deployment.

A better way is to use a content delivery network, or CDN. A CDN provider offers Points of Presence (PoP) all over the world where they locally cache resources for your site. Customers will access your site through the closest PoP to them. For example, a customer in London will connect to the UK PoP instead of bouncing to a server located in the US. This speeds up delivery substantially.

You can also avoid any content delivery problems by choosing the right infrastructure provider, who has the capacity to deliver quickly to a worldwide audience. A provider with global data centers and a high performance network like 100TB can eliminate congested routes and deliver your service as quickly as possible to multiple locations.


Reduce server load

When the going gets tough, the tough get efficient. It’s much easier to manage sudden increases in customer volume if each visitor is handled as efficiently as possible.

HTTP caching allows the browser to store local copies of resources for faster retrieval. Instead of downloading everything from the server on each page refresh, the browser can serve up local copies of stored resources like images and stylesheets. Cache headers can be attached to resources to tell the server when they need to refresh their stored local copies to ensure nothing is shown out of date.

Ecommerce stores are heavy on images, which can put a heavy load on servers. Compressing content as much as possible, while still maintaining quality, will save bandwidth for everyone. Consider delivering content using gzip compression, and following image optimization best practices.

 

Create a Performance Dashboard

Setting up tracking systems before the storm hits can help your team keep an eye on things, as well as evaluate past performance.

Geckoboard, a DIY dashboard software for business metrics, has an in-depth blog post on eight metrics ecommerce companies should track over the holidays. Here are three must-track metrics for developers to have up on their wall at all times..

Uptime - if we’re trying to avoid downtime, the first thing to do is know if we actually have any downtime. While you could hire someone to just refresh the page time and time again and record the results, a service like Pingboard will do this automatically for you. They can also provide degraded service analysis, like search downtime, regional load issues or slow page loading.

Page Load Time - An increase in page load time can be your first sign that something is about to go horribly wrong. Customers hate slow loading too.  Even a 1-second delay in page response can result in a 7% reduction in conversions, according to KissMetrics.

Volume - Keep an eye on the flow of traffic so you don’t get caught off guard when volume creeps up. Having a historical view of traffic spikes can help you plan for future promotions and holiday surges.

Setting up a dashboard before the busy period sticks will keep you laser focused on the metrics that matter.

 

Make a plan for the bad times

If the worst does happen and downtime occurs, it’s important to get up and running as quickly as possible. That means having a plan in place so every stakeholder knows what part they play in recovery.

Make a list of your third party support contracts and contacts. If you need to purchase elevated support for some services to make sure you’re front of the line, consider doing so. Ensure your DDoS protection service is up to date. Assign a specific Point of Contact for each service who’s responsible for managing any conversations with the vendor when things blow up, and set up on-call policies along with the numbers of the first people to contact.

Write all of this down in an escalation procedure that’s easily accessible to everyone. You might even want to practice going through simulated downtime to make sure you’re not missing information.

 

Don’t let your downtime become your competition’s advantage

Customers won’t wait around for your site performance to improve. They’ll return to the Google search results and choose the next site on the list.

Planning ahead for scale takes time and effort. If you’re growing at a steady pace now, it’s easy to just keep growing servers and adding more resources. But a strategy that works for 10% growth won’t keep up to a 10x growth spike. You might be tempted to play it by ear and just wait and see how the holiday season goes. But by the time you start experiencing scaling issues, it will be too late to put these strategies into place. Instead of being proactive, you’ll end up patching holes with duct tape. It’s not a fun way to spend your busiest weeks.

Get ahead of the volume and start planning today to scale gracefully. This is the quiet before the storm.  

 

Ready to deploy your infrastructure? Let 100TB’s servers give your business  power.

How Artificial Intelligence Delivers Marketing Results of the Future

Previous Article

Machines, Robots and Imagination Run Riot: TedxLugano Speakers Tell The Future

Next Article