Websites and Tech

Time to First Byte (TTFB)

Also called TTFB, server response time

The delay between a browser requesting a page and the first byte of the server's reply arriving.

Quick facts: Time to First Byte (TTFB)

Category
Websites and Tech
Also called
TTFB, server response time
Level
Advanced
Affects
Page speed, Core Web Vitals, crawl rate, bounce
Where to see it
PageSpeed Insights, Chrome DevTools network panel, Search Console crawl stats, WebPageTest
In this article4
  1. What time to first byte measures
  2. Why time to first byte matters
  3. Common mistakes with TTFB
  4. How to act on it

What time to first byte measures

The clock starts when the browser asks for a page and stops when the first byte of the reply arrives. Everything in between is included: finding the server through the domain name system, opening the connection, negotiating the secure handshake, the server’s own thinking time, and the journey back. In practice two things dominate — how far the visitor sits from the server, and how much work the server has to do before it can say anything at all.

On a database-driven site such as WordPress, that work means running theme and plugin code and querying the database to assemble the page. When the page is cached, the server hands over a file it prepared earlier and answers almost immediately. When it is not, it rebuilds the page from scratch for every visitor. The gap between those two states is usually the single largest number in the whole measurement, which is why caching is the first thing to look at.

Why time to first byte matters

Nothing on the page can begin until the first byte lands. The browser cannot parse markup it has not received, cannot discover the stylesheet, cannot start downloading the hero image. Every later milestone inherits the delay, including Largest Contentful Paint, which is one of the three Core Web Vitals.

It also shapes how much of your site gets crawled. Search Console reports an average server response time in its crawl statistics, and a server that responds slowly is visited less enthusiastically. For a large catalogue or a site publishing regularly, that is a real constraint rather than a theoretical one.

Common mistakes with TTFB

The most common is measuring from the wrong place. A test run from a data centre near the origin server flatters it; the same page requested on a mobile connection in Kathmandu, or by a customer in Sydney, tells a different story. Test from where your customers actually are.

The second is misdiagnosis. Teams compress images and defer scripts for weeks while the real delay happens before a single pixel is requested. The third is trusting a content delivery network to fix everything: if it only caches images and stylesheets while the HTML is still built on a distant, overloaded origin, the first byte still waits for that work. Redirect chains add a whole extra round trip each, and on shared hosting a neighbouring site’s traffic spike can slow your responses without anything on your side changing.

How to act on it

Measure the same page cached and uncached, from a location that matches your audience, and compare. If the cached response is fast and the uncached one is slow, the fix is caching and code efficiency. If both are slow, the problem is the server or the distance, and no amount of front-end work will help.

From there the moves are ordinary: cache full page output rather than assets alone, put a delivery network in front that will serve HTML, remove unnecessary redirect hops, keep the database and plugin list tidy, and keep server software current. When the plan is simply undersized or in the wrong region for the audience, the honest answer is a better plan or a move, which is what hosting and migration work exists to handle.

Do and do not

Do

  • Measure from where your customers actually are
  • Cache page output so the server skips repeated work
  • Check the server response before blaming front-end code

Do not

  • Judge hosting on a single test from one location
  • Add a CDN while leaving slow origin responses unfixed
  • Ignore response times creeping up as traffic grows

Questions people ask about this

What counts as a good time to first byte?

There is no single figure that suits every site, and the honest answer depends on the platform, the hosting plan and the distance between server and visitor. What matters more is the comparison and the trend: whether a cached page answers far faster than an uncached one, and whether responses are getting slower as traffic grows. Close those gaps first.

Is time to first byte a Google ranking factor?

Not directly, and it is not one of the Core Web Vitals. It matters because it delays everything that follows: Largest Contentful Paint cannot happen until the first byte has arrived, and that measurement does form part of the page experience signals. A slow server also reduces how much of your site search engines are willing to crawl.

Will a CDN fix a slow server?

Only partly. A content delivery network shortens the trip for the files it stores, so images, stylesheets and scripts arrive from somewhere nearer the visitor. If the page markup itself is still assembled from scratch on a distant or overloaded origin server, the first byte still waits for that work. Cache the page output, not just the assets.

Related terms

Found this useful?

Share it, or ask an AI to summarise it

Back to the glossary

Knowing the term is the easy part

Applying it to your own site and budget is the work. Book a call and I will tell you what actually applies to you.