Are you set for the biggest e-commerce day of the year? Adobe predicts record earnings for U.S. retailers in the shortest holiday shopping season of the past decade – $1.1B for Thanksgiving, $1.6B for Black Friday, with Cyber Monday topping it off at $2.3B.
If you’re relying on network sniffers or browser instrumentation alone to monitor your customer experience, you may be setting yourself up for some serious blind spots. While traditional monitoring approaches are no longer adequate on their own, abandoning them in favor of the newer (easier?) routes is not entirely advisable either.
Here are four tips on designing an updated APM strategy to keep a close watch on your e-commerce revenue streams this holiday season:
- Capture functional issues and establish context. Performance is not all that matters when it comes to customer-facing websites. Functional issues, such as errors or design gaffes, are far more common than performance issues, and a major factor in abandonment and conversion rates. Consider a customer unable to complete their checkout because a website’s shipping address field was not designed to parse a lead zero in the zip code. Implementing a solution that captures browser events, such as mouse clicks and keyboard input data, and has the ability to replay a user’s session activity - and show you exactly what your customer saw - will help you proactively identify and troubleshoot these kinds of problems.
- Look for detailed insight into page load times. To be fully inclusive of ads, maps, shopping carts, web analytics, social media modules, CDN and DNS response times, etc., page load time must be monitored from within the browser itself. Luckily, we have the HTML 5 navigation timing feature available in newer Internet Explorer, Firefox, and Chrome browsers. It includes full page load times broken out by DNS lookup, redirect, SSL handshake, processing, and cache access timing. Look for an APM solution that integrates this feature.
- Isolate problems to specific individual page elements. Currently the information captured from within the browser is limited to full page loads and does not provide timing information on individual page hits, such as loading of graphics or images, CSS stylesheets, or back-end calls to web servers or REST APIs. Network sniffers with web page analysis capability can time HTTP request and response for individual page objects, enabling troubleshooters to isolate issues related to specific page elements. Make sure your monitoring solution includes this feature.
A multi-pronged approach across sniffer, browser and server-based monitoring is necessary for maximum visibility into modern website performance.
Happy shopping trails!