Monitor, Analyze, and Test Website Load Speed and Performance

When performing a website speed test, there are two common technologies used to measure webpage performance:

1. The first is real user monitoring, commonly abbreviated as RUM, which grants visibility into the performance of webpages and applications by collecting webpage load test data from actual desktop and mobile browsers and web connections around the globe. Using real end-user data to measure webpage performance means you don’t need to define in advance which use cases or performance parameters are most important or valuable.
   Once a RUM tool is activated, it continuously collects performance data from real user interactions with the monitored website, regardless of the specific pages visited—no matter where the end user goes or the extent to which they engage with the site, there will be viable performance data collected. RUM also allows for targeted actions and isn’t impacted by performance inhibitors, making it an ideal option for running page load speed tests on large websites and highly complex web applications, so you can achieve deeper insights and website speed analysis.
   
2. The second option for website performance tests is synthetic monitoring, which simulates browser traffic along common site paths end users typically follow. Synthetic testing is useful for how it allows site admins to verify critical website properties, functions, and end-user interfaces are operating as intended—all without the need for real site or application traffic. Since it allows for highly controlled testing, synthetic monitoring is also highly repeatable and typically provides reliable information—making it useful for conducting page load speed tests, checking site elements for optimal performance, and gathering data for website speed analysis—though it’s limited because it inherently doesn’t collect data on actual end-user traffic and site experience. It’s also an excellent option for ad-hoc website performance tests, which help both to build a structured understanding of overall site or application experience and to detect performance issues.

Webpage size and site traffic volume can impact the usefulness of the two monitoring technologies. However, because RUM and synthetic monitoring each offer a set of unique advantages and insights, using a combination of the two technologies is ideal for running website performance tests and website speed analysis. Using RUM to monitor core site and application traffic helps establish the common end-user pathways, which can be supplemented with synthetic page load speed tests, allowing admins to quickly identify and troubleshoot specific problems to build a more holistic understanding of application or site performance.

To maximize the accuracy and relevancy of the results of website speed tests, admins should pay attention to several key speed test elements, including:

• Number of tests performed: One webpage load test is insufficient for a thorough understanding of site and application performance. The results of a website speed test can vary widely based on the time the test is performed, as site traffic is an extremely dynamic metric. Running page load speed tests at different times is critical for creating a better picture of site performance, as traffic naturally fluctuates.
• Test location: The geographic location of accessing a site or application will have an impact on load speed and performance. Therefore, knowing whether typical site or application traffic is local, regional, or global is key to determining the most informative location(s) from which to conduct website speed tests.
• Test objects: Website performance tests should include a range of site posts and pages to establish a more precise overall performance baseline, which can then be used to highlight underperforming elements.

Many website performance testing tools will also gather data on additional metrics to help improve site speed and performance. Content size, type, and requests by content type can help to reveal the page elements requiring the most resources for browsers to load. Large images, for example, can increase load time for sites and pages, which is well-documented as being directly linked to increased customer or end-user abandonment. Requests by domain is another useful metric, as it can be used to determine how external scripts and services are impacting site performance. Some website speed testing solutions will analyze individual requests to identify the specific elements responsible for creating delays.

To improve the results of a website performance test, here are common ways to improve loading speed:

• Minimizing TTFB: While many site owners will be inclined to focus primarily on front-end performance issues, time to first byte (TTFB) is actually a server-side metric. For a browser to load a site, it must send an HTTP request to the site’s web server, which must then perform the Domain Name System (DNS) lookup, process the data, and respond to the request. Most TTFB-associated problems can be traced back to issues with dynamic content creation, improper web server configurations, network bandwidth, and site traffic. Google recommends server-side response time be kept to under 200ms.
• Optimizing images: Images can easily make up a significant portion of the weight of a webpage, with larger images in particular being responsible for increased load times. Many images—such as web posts, logo files, and other graphic elements—can be reduced in size without sacrificing quality or negatively affecting the end-user experience.
• Minimizing HTTP requests: After the initial HTTP request to the web server, a browser will examine the incoming HTML and send additional requests for elements like images and scripts. More elements mean more HTTP requests, which in turn means slower site loads. Using external style sheets, CSS sprites, and image maps can help streamline the number of incoming HTTP requests and improve server response time.
• Working with a Content Delivery Network (CDN): CDNs are distributed server networks designed to provide site content with greater speed. By replicating content across the network, CDNs allow web browsers to access the nearest version of the data, instead of funneling all requests to one centralized server. This helps prevent server bottlenecks and improves the utilization of network resources. Performing page load speed tests from multiple geographic locations is also advantageous because it helps reveal how and the degree to which a CDN can improve site load times.
• Caching content: Caching refers to the temporary storage and subsequent delivery of site pages to improve load times and minimize bandwidth usage. Unless changes have been made to a site, the previously cached version can be delivered to client browsers with greater speed without increasing web server load.
• Switch to cloud-based DNS infrastructure: An efficient DNS is a critical factor for any company or organization operating via the internet. As computing environments and customer bases become more distributed, adopting cloud-based technologies often results in better performance and reliability—which holds true for DNSs, as well. Every browser interaction with a website starts with a set of DNS queries—and if the DNS has failed, each underperforming query will create a cascading effect on the performance of all site content, data, and applications. If pages are unavailable or loading extremely slowly, end users are more likely to abandon the page, leading to further loss of revenue and brand reputation. Cloud-based DNS infrastructure provides additional resiliency and scalability by distributing digital content and resources between corporate data centers, CDNs, and the cloud. Using a CDN in combination with a managed DNS solution can lead to even greater improvements in site performance and reliability, customer satisfaction, and end-user retention. The distributed nature of CDNs helps balance traffic loads across multiple endpoints, while many CDN providers offer automated failover features, which allow site content to remain available even if the CDN experiences a failure or outage.

Website speed testing is a necessary practice, as it helps to identify and troubleshoot site pages and content needing performance optimization. Slow-to-load or underperforming sites are more than just an inconvenience or an indicator of inefficiency within a system—increased load time can have a direct impact on a company’s bottom line.

E-commerce sites especially stand to benefit from routine page load speed tests and performance optimization—the faster web servers can deliver content to client browsers, the faster they can generate more sales.

End users and customers typically perceive site load times as being slower than what the results of a website performance test actually suggest. Brands should therefore prioritize using website speed tests to make load speeds as fast as possible to prevent loss of traffic and customer conversion. Slow site loads can also negatively impact a business in non-monetary, but equally substantial, ways, including how the brand is perceived by customers. A poor website user experience can even reduce customer loyalty, and ultimately drive down visitor engagement. Performing page load speed tests is a necessary and foundational part of optimizing website performance and speed.

Website performance testing is also critical for determining the stability of a site or application. It’s important to know the maximum traffic volume a site and web server can handle before it begins to experience performance issues or crashes, as this enables admins to have a better understanding of when to act, like investing in new software and hardware or deploying necessary updates to improve performance.

In addition to leading to more engaged customers and end users, websites optimized for performance provide back-end benefits to brands and companies. Faster traffic requires less bandwidth consumption, which can allow companies to offload servers and reduce the costs associated with web hosting. This has the added advantage of giving admins the ability to scale site capacity for seasonal increases and sudden spikes in traffic volume.

SolarWinds Web Performance Monitor (WPM) provides a useful suite of tools for helping you test, analyze, and optimize webpages and site content. WPM’s web-based application performance monitoring capabilities are built to provide you with quick and actionable insight into web service issues impacting load times and user experience, including for third-party and SaaS applications.

WPM collects performance metrics associated with webpage load times to create detailed TCP waterfall charts to highlight the specific site elements and content contributing to the biggest overall load times. This allows you to more easily keep an eye on HTML, JavaScript, and CSS elements causing site slowdowns, while WPM’s notification and alerting tools make it simple and straightforward to receive updates when the tool detects underperforming or incomplete web transactions.

WPM’s website performance testing capabilities allow you to easily monitor traffic and web interactions from multiple locations—including within network firewalls or from Amazon instances—for deeper understandings of how site performance can vary. WPM installs behind network firewalls, allowing you to safely use the tool’s monitoring functions for tracking the performance of internal web applications, such as help desk, supply chain, and customer relationship management (CRM) solutions.

WPM can also be seamlessly integrated with other SolarWinds products for more holistic visibility into critical infrastructure performance like web apps, allowing you to quickly drill down to identify root cause and generate easy-to-understand visualizations like web transaction and application performance metrics.