This case study explores how Digital Loop team collaborated with a leading e-commerce company (health industry) to address the specific challenge of enhancing website performance and load speed.
The project aimed to optimize page load time through targeted technological improvements (specifically by transitioning from HTTP/1.1 to HTTP/2. Details provided further), with the goal of elevating user satisfaction and ultimately increasing conversions.
The primary challenge our client faced was enhancing page load speed across their network of websites. Although the sites already achieved relatively high-speed performance, the client sought every possible improvement to reduce load time further.
This challenge was significant because in the e-commerce industry even minor delays in page loading can lead to increased bounce rates, decreased user engagement, and potential loss of conversions.
For the client, competing on speed was essential; studies showed that page load times have a direct impact on conversion rates, with faster sites encouraging more purchases and higher retention.
The Digital Loop team explored various opportunities to enhance page speed, carefully analyzing each aspect of the website’s infrastructure and performance. We assessed areas such as resource loading, server response times, and data transfer efficiency to identify bottlenecks affecting load speed.
By considering multiple optimization strategies, from reducing file sizes and leveraging browser caching to exploring advanced protocols like HTTP/2, we crafted a tailored approach to achieve faster load times.
The next step for our team was to undertake a structured analysis process aimed at understanding the bottlenecks and potential gains from switching protocols from HTTP/1.1 to HTTP/2.
The Digital Loop team first measured the current page load times for the company’s websites, particularly focusing on the time required to load static files (e.g., images, CSS) over HTTP/1.1 (the HTTP/1.1 protocol was used across the client’s entire website network).
Testing without network throttling revealed an average load time of 2.013 seconds on landing pages.
However, during simulations under slower 3G conditions, load times exceeded 15 seconds in some cases. This variation highlighted the impact of HTTP/1.1’s limitations under different network conditions.
In parallel, our team reviewed relevant studies on page speed and user behavior, using findings from sources like Google and SOASTA to understand the broader impacts of load time on user engagement and conversions.
Conversion Rates
Bounce Rates
User Engagement
These studies helped underscore the business case for improving page speed by illustrating how even small gains could yield significant engagement and conversion benefits.
Recognizing that the e-commerce company’s sites were loaded with a high volume of static files, the team identified HTTP/2’s multiplexing, server push, and header compression features as potential solutions:
To help our non-technical readers better understand the difference between HTTP/1.1 and HTTP/2, we found an example that clearly illustrates the results of HTTP/1.1 compared to HTTP/2.
In the following simulation, a “main” document(big square) was fetched. This could represent the index of a website.
After retrieving the “main” document, the simulator fetches 99 linked items, which could include images, scripts, or other documents.
HTTP/1.1 is limited to 6 concurrent requests. The big box is the initial index or collection.
HTTP/2 can fire off many parallel requests over 1 TCP connection
Source: https://evertpot.com/h2-parallelism/
To ensure the integrity of the results, our team requested that the client refrain from any other site optimizations during the testing period.
This created a controlled environment to isolate the impact of the HTTP/2 implementation on load times and user behavior.
Given the client’s high mobile traffic, the Digital Loop team emphasized optimizing load speed specifically for mobile users.
Mobile sessions are particularly sensitive to delays, and improvements here were expected to drive significant gains in engagement.
HTTP/2’s multiplexing and server push functions were chosen with mobile users in mind, where each fraction of a second matters for conversion and retention rates.
The switch from HTTP/1.1 to HTTP/2 yielded measurable improvements for our client, demonstrating both immediate and promising long-term benefits.
Faster speeds can directly influence purchase decisions, especially in e-commerce, where a seamless experience is critical.
By transitioning from HTTP/1.1 to HTTP/2, the project focused on addressing the limitations of HTTP/1.1 that slowed down static file loading and caused delays under constrained network conditions.
The implementation of HTTP/2 led to a 5.6% increase in page views per session for mobile users and a 2.68% reduction in bounce rate, lowering it below 20%. These results demonstrate the direct impact of faster load times on user engagement.
By meeting users’ demand for speed, the project improved customer experience, making users more likely to stay on the site and explore multiple pages, laying the groundwork for future loyalty.
HTTP/2’s multiplexing, server push, and header compression capabilities provide a scalable solution that can support future growth, making the client’s sites more competitive in a fast-paced e-commerce landscape.
The project’s success lies in its targeted approach, which identified and resolved specific performance issues using a tailored solution (HTTP/2), proven to increase speed and engagement.
The results not only met but exceeded the client’s expectations by achieving measurable improvements in engagement metrics and setting a foundation for long-term growth.
This case illustrates how focusing on performance optimization can create a ripple effect, benefiting both user experience and business outcomes.
You can find our office in the heart of Munich:
We are social