Every day should be Black Friday. By that we mean that retailers should be in a continuous state of peak readiness if they want to deliver the best possible experience for shoppers. In the run up to the holiday season, retailers pull out all the stops to ensure that their systems are thoroughly optimized to handle big and prolonged spikes in traffic and purchasing activity.

We’ve identified three primary reasons retailers experience site problems, which are more likely to occur during peak times, and how to address them during the most important time of the year.

Pipeline runtimes 
Retailers should strive for a pipeline average runtime target of about 500 milliseconds. Anything under this threshold is considered “safe” while anything approaching or over one second is a red flag indicating the site is not scaling properly. How do you know what’s causing slowness? A pipeline profiler shows a processing breakdown by site and pinpoints problem areas and allows you to monitor performance after a code deployment, which is common around peak selling times. And to dispel the myth: pipeline profilers do not degrade performance — there is no reason not to use one.

Retailers should strive for a 50%-70% caching rate on product detail pages and about 90% on the homepage. The more data that’s cached, of course, the faster it’s available to users as it avoids making a trip to the server. In fact, a web service call lasting more than two seconds is considered poor and will likely frustrate your shoppers and drive them away. So how can you determine why something isn’t being cached? One way is to limit the amount of custom or unique iterations for a landing page from a large email campaign. While a more personalized web experience may seem ideal, a shopper will definitely appreciate site speed more when clicking through a promotional email, especially on a mobile device.

Performance problems are often not necessarily the result of poorly written storefront code, but the result of ineffective planning in regard to third-party integrations. It’s therefore critical when preparing for peak traffic to ensure that both the integrations with, and the services provided by, third-party vendors can effectively handle the additional stress. Retail brands should always have a resiliency plan for storefront integrations, particularly with outgoing API calls and job integrations that import and export data. All third- party integrations should have explicit and appropriate timeout values defined. As a best practice, it is recommend to always plan for the worst case scenario; will your site continue to function if any of your integrations fail?

This and much more is covered in our 2017 Guide to Holiday Readiness. Download it today!