Server Response Time explained: measuring and improving server response time

Fast load times are a crucial factor for a website's success, and server response time forms the foundation. A slow server response can deter visitors and undermine SEO efforts. In this article we explain why Server Response Time is so important, how it is measured and how it can be improved.

What is Server Response Time (SRT)?

Server Response Time (SRT), also known in German as Serverantwortzeit, is the period a server needs to respond to a client's request. It starts when the client sends the request and ends when the first byte of data from the server reaches the client. This time includes the duration for transmitting the request to the server, the server's processing of the request and the start of the response transfer.

Server Response Time is by no means the Page speed of a website. Instead, Server Response Time refers precisely to the duration required for a client to receive the first "sign of life" from a server after a request. Server response time is therefore highly dependent on the performance of the server (and its physical distance from a visitor). The faster it processes the information and responds, the shorter the response time and the sooner the transfer of the requested data can begin.

Distinction between server response time and server speed

Server response time and server speed are two distinct but related concepts.

Server response time

• This is the time that elapses until the server sends the first response to a request.
• It includes the network transmission time of the request, the processing time on the server and the transmission time of the first byte back to the client.
• The server response time is heavily influenced by the server's performance, the network infrastructure and the distance between server and client.

The server response time is therefore closest to the performance metric TTFB (Time to First Byte).

Server speed

• This is the total time required to fully load a web page in the user's browser.
• It starts when the request is sent to the server and ends when all of the page's resources (HTML, CSS, JavaScript, images, etc.) are fully loaded and rendered.
• Server speed depends not only on server response time, but also on the size and number of resources, the efficiency of client-side scripts, network speed and other factors such as caching and Content Delivery Networks (CDNs).

Server speed is therefore closest to the performance metric LCP (Largest Contentful Paint).

While the server response time represents an early, specific part of the loading process, server speed covers the entire process up to the full display of the page in the browser. Both are important metrics for web performance optimisation, but they address different aspects of load times.

How can the server response time be measured?

There are now numerous tools for measuring a server's response time, some of which can be used online. One of these — certainly the oldest — is the ping command, which has been included in all operating systems since MS‑DOS in the 1980s and remains part of the standard toolkit on Microsoft Windows and Linux. For decades it served as the primary tool for measuring and improving networks and is still in use today.

With server monitoring the response time can be recorded and measured regularly.

In addition, there are now many other tools that can often be operated in a browser. These offer several advantages that make measuring and improving considerably easier than in earlier times. A very important one in an age of global connectivity is the ability to measure the server response time from different locations around the world. Purely theoretically, a data packet transmitted at the speed of light through a fibre‑optic cable from Europe to Australia requires at least 100 ms. As an informal threshold for a fast server response time, roughly 200 ms is often cited. This includes delays caused by processing the information in multiple routers and the receiving computer.

Some examples of popular and reliable platforms for measuring response time are GTmetrix or Pingdom.

Graphic: Measuring server response time with GTmetrix

Which factors can improve server response time?

Server response time is influenced by a variety of factors, both technical and infrastructural:

Server location and geography

The physical location of the server can significantly affect response time. The further the server is from the user's location, the longer data transmission takes. Therefore, geographically distributed servers or Content Delivery Networks (CDNs) can reduce response times by placing servers closer to users. For this reason, the server location in Germany is recommended for websites aimed at domestic visitors.

Server performance and hardware

The server's hardware, including processor performance, memory (RAM) and disk speed, has a direct impact on the processing time of requests. More powerful servers can process requests faster, resulting in shorter response times. Using Solid-State Drives (SSDs) instead of conventional hard drives can also improve performance.

Web hosting type

The type of web hosting plays a significant role:

• Shared web hosting: Multiple websites share the same server and its resources. This can lead to overloads and longer response times.
• Virtual Private Server (VPS): Provides dedicated resources within a shared server, which can lead to better response times compared with Shared web hosting.
• Dedicated Server: All server resources are available to a single website, enabling optimal performance.
• Cloud Server: Uses a network of virtual servers to flexibly handle traffic spikes and ensure stable response times.

Tip: Read about measuring, comparing and optimising CPU performance on a VPS as well as measuring, comparing and optimising disk performance on a VPS

Network conditions and bandwidth

The network connection between the server and users affects response time. Higher bandwidth and stable networks can reduce the transmission time of requests and responses. Network outages or bottlenecks, on the other hand, can lead to delays.

Application code and database optimisation

The website's source code and the configuration of the database can strongly influence server response time. Inefficient code, unnecessary database queries or unoptimised database tables can lengthen processing time. Through code optimisations, database indexing and caching mechanisms, performance can be significantly improved.

Content Delivery Networks (CDNs)

Content Delivery Networks (CDNs) store copies of website content on servers distributed around the world. By serving content from a server that is geographically closer, response times can be reduced significantly. This is particularly useful for global websites with users in different regions.

Caching strategies

By caching frequently requested data at multiple levels (browser cache, server cache, proxy cache), the need to process the same requests repeatedly can be reduced. This leads to faster response times because cached content can be delivered more quickly.

Minimising HTTP requests and optimising resources

The number and size of resources a server must deliver affect response time. By minimising HTTP requests, compressing files (e.g. images, scripts, stylesheets) and using modern web technologies such as HTTP/2, overall load times can be improved.

Together, these factors determine the efficiency and speed with which a server responds to requests and should be carefully monitored and optimised to ensure an optimal user experience.

A fast server response time is essential for optimising the user experience, increasing conversion rates and improving search engine rankings. Targeted measures such as optimising server location, hardware, code and caching strategies can effectively reduce server response time. Overall, optimised server performance leads to more satisfied users and better visibility on the web.

Our article is based on our own experience and research as well as information from external sources.

Article update:
Update 12.06.24: Content updated and expanded.

Sources & further links on the topic:
https://tool.pingdom.com/ (Tool to measure server response time)
https://gtmetrix.com/ (Tool to measure server response time)