Definition:
TTFB, or Time to First Byte, is the time elapsed from the start of a resource request until the client receives the first byte of the response.
For a web navigation, it may include redirects, DNS resolution, connection establishment, TLS negotiation, transmission of the request, server processing and the beginning of the response. It therefore does not measure server speed alone and does not by itself represent total loading time or the complete user experience.
Table of contents
Time To First Byte Measurement
TTFB can be measured in a laboratory through a controlled test or through real-user data. The value changes according to location, network, device, protocol, cache status and the infrastructure serving the request.
A measurement begins when the client starts the navigation or request and ends when the first byte is received. Requesting an IP address directly removes DNS resolution, but the connection, network and processing still contribute. Diagnosing a delay therefore requires these components to be separated instead of automatically attributing it to the hosting service.
Tools may report different results because they use different locations, connections and methodologies. Real-user data should generally be assessed across the distribution and percentiles of many visits rather than through a single test.
Time to the optimal first byte
There is no universal optimal value for every resource. As a web performance principle, a low TTFB allows the browser to start receiving HTML or other data sooner, but its importance depends on the page, caching, personalisation and the rest of the rendering process. TTFB is not itself a Core Web Vital.
Common causes of a high TTFB include:
- Load and capacity: High request volume, insufficient resources or shared-environment limits may delay the response.
- Network and distance: Latency, packet loss, redirects and the distance between the user and origin increase the time. A CDN can reduce part of the journey when content is served from a nearby location.
- Dynamic processing: Slow queries, calls to external services, calculations, plugins or missing caches may extend response generation.
- Technical configuration: DNS, TLS, the web server, application, database and redirect rules can add delays or bottlenecks.
Improvement should begin with repeatable measurements and identification of the slow component. Optimising caches, queries, code, infrastructure and distribution can reduce TTFB, but it should be evaluated with other WPO metrics to determine the actual effect on loading and interaction.