--- title: What Is a Good TTFB? Benchmarks From Our Own Hosting Accounts description: "A good TTFB is 0.8 seconds (800 ms) or less, according to Google's rough guide. That is the bar most sites should clear for most of their visitors. On four hosting accounts…" url: "https://hostingguider.com/blog/good-ttfb-benchmarks/" canonical_url: "https://hostingguider.com/blog/good-ttfb-benchmarks/" date: "2026-10-19T01:22:00+00:00" modified: "2026-10-12T08:46:28+00:00" author: Rantideb Howlader type: article site: HostingGuider language: en-US --- # What Is a Good TTFB? Benchmarks From Our Own Hosting Accounts A good TTFB is 0.8 seconds (800 ms) or less, according to Google’s rough guide. That is the bar most sites should clear for most of their visitors. On four hosting accounts we paid for, 30 of 34 city readings cleared it. Distance from visitor to server shaped the results more than the type of plan. You’ll find expected numbers here, not fixes. If yours is already too high, the [server-level fixes for slow TTFB](https://hostingguider.com/blog/fix-ttfb-hosting-level/) are covered separately. Table of Contents ## What Google says a good TTFB is Google treats a TTFB of 0.8 seconds or less as good, measured at the 75th percentile of real page loads. TTFB (time to first byte) is the time between a visitor requesting a page and the first byte of the response arriving in the browser. Google’s [TTFB article on web.dev](https://web.dev/articles/ttfb?utm_source=hostingguider.com), updated November 18, 2025, puts it this way: “As a rough guide, most sites should strive to have a TTFB of 0.8 seconds or less.| TTFB | Google’s label | | --- | --- | | 0.8 s or less | Good | | 0.8 s to 1.8 s | Needs improvement | | Over 1.8 s | Poor | **Two details matter here:** - **TTFB isn’t a Core Web Vital.** The same article says that, because of this, “it’s not absolutely necessary that sites meet the ‘good’ TTFB threshold. - **It’s a rough guide.** Google says it must be weighed against how your site delivers its main content. A server-rendered WordPress page can have a higher TTFB and still load well. **For your site, this means:** treat 800 ms as a ceiling for most visitors, not a score to chase to zero. A low TTFB simply leaves more of the loading budget for the page itself. TTFB is one of the server speed signals in the [website performance and server response guides](https://hostingguider.com/topics/website-performance/), and the only one this article benchmarks. ## Our TTFB data: four accounts, one test site We measured TTFB on four hosting plans, all running the same WordPress test site, hostingguider.online, with the same theme and plugins. We paid for every plan (Namecheap’s first month was a $0.00 promotional rate). The [receipts for every test account](https://hostingguider.com/purchased-hosting-evidence/) are public. | Account | Hosting type | Server location | TTFB tool | When measured | | --- | --- | --- | --- | --- | | Hostinger Business Web Hosting | Shared | Not stated in the review | GTmetrix (1 location), DebugBear (10 cities) | Test date not stated in the review; the account was about 18 months old | | Hostinger Cloud Startup | Cloud | Singapore | KeyCDN (8 cities) | Several months into use; review published September 24, 2026 | | Kinsta WP 2 | Managed WordPress | Region not named; the review points to Asia-Pacific | KeyCDN (8 cities) | First 3 days only, from August 27, 2026 | | Namecheap Stellar Plus | Shared | US data center | 8-city test (tool not named in the review) | Late August 2026, short history | Each city figure is a single test run from that location, not an average of real visitors. Treat them as snapshots. ## TTFB by city and account Where our tests share a city, the table puts each account side by side. All values are in milliseconds (ms). | Visitor city | Hostinger Business (shared) | Hostinger Cloud Startup (Singapore) | Kinsta WP 2 | Namecheap Stellar Plus (US) | | --- | --- | --- | --- | --- | | Singapore | 392 | 156.86 | 82.79 | 650 | | Sydney | 1,692 | 674.16 | 376.28 | 604 | | San Francisco | 516 | 514.54 | 593.89 | 126 | | New York | Not measured | 388.2 | 797.66 | 232 | | London | Not measured | 453.82 | 759.04 | 523 | | Frankfurt | Not measured | 282.8 | 780.22 | 538 | | Amsterdam | Not measured | 317.36 | 798.92 | 489 | | Bangalore | 755 | 247.17 | 513.05 | 707 | “Not measuredmeans that city wasn’t in that account’s test. The Hostinger Business DebugBear run also covered six cities outside this table: | Visitor city | Hostinger Business TTFB (ms) | | --- | --- | | São Paulo | 229 | | Atlanta | 485 | | Brussels | 709 | | Helsinki | 905 | | Tokyo | 1,242 | | Dubai | 1,316 | GTmetrix, testing from one location the review doesn’t name, recorded 191 ms on the same account. Each account’s full results, including speed scores and uptime, are in its review: - The [Hostinger Business shared hosting test](https://hostingguider.com/reviews/hostinger/), on an 18-month-old account - The [Singapore Cloud Startup account](https://hostingguider.com/reviews/hostinger-cloud-startup-review/), after several months of use - The [three-day Kinsta WP 2 test](https://hostingguider.com/reviews/kinsta/), with a short data window - The [Namecheap Stellar Plus results](https://hostingguider.com/reviews/namecheap-hosting/), from a US data center **Source:** HostingGuider TTFB benchmarks, October 2026. You may quote this table with that credit and a link back to the article. ## What the numbers show Three patterns stand out in the 34 city readings above. ### Distance to the server dominates Each account with a known or likely server region was fastest near it and slowest far from it. - **Namecheap (US data center):** 126 ms from San Francisco, but 650 ms from Singapore and 707 ms from Bangalore. - **Hostinger Cloud Startup (Singapore):** 156.86 ms from Singapore, but 514.54 ms from San Francisco and 674.16 ms from Sydney. - **Kinsta WP 2:** 82.79 ms from Singapore, rising to 798.92 ms from Amsterdam. - **Hostinger Business (location not stated):** 229 ms from São Paulo, but 1,692 ms from Sydney. The gap between an account’s fastest and slowest city ranged from about 4 times (Hostinger Cloud) to about 10 times (Kinsta). Our Kinsta review reads that pattern, with its SSL report, as a Google Cloud origin likely in Asia-Pacific. The review doesn’t name the exact region. **For your site, this means:** pick a server close to most of your visitors before comparing anything else. ### Most readings were “goodBy Google’s rough guide, 30 of the 34 city readings were under 800 ms. Kinsta’s Amsterdam (798.92 ms) and New York (797.66 ms) results sat just under the line. All four exceptions came from Hostinger Business: Helsinki (905 ms), Tokyo (1,242 ms), Dubai (1,316 ms) and Sydney (1,692 ms). Each of those falls in “needs improvement, below the 1.8 second poor line. **For your site, this means:** a WordPress test page on a decent plan can pass Google’s guide from most places. Real visitors on slow mobile networks can see higher numbers than these lab runs. How real websites score in Chrome’s field data is collected in a separate set of website speed statistics for 2026. ### Same-region results look alike When the visitor was near the server, the numbers bunched together across very different plan types. Namecheap’s shared plan scored 126 ms near its server. Hostinger’s cloud plan scored 156.86 ms near its server. Kinsta, a managed WordPress plan, scored 82.79 ms from Singapore. Those are different cities and different tests, so they aren’t a ranking. They do show that a cheap shared plan can respond fast when the visitor is close by. ### Where a CDN can help A CDN (content delivery network) keeps copies of files on servers in many cities. If it also caches full HTML pages, distant visitors can get the first byte from a nearby edge server. Our records don’t say whether a CDN served the HTML during these runs. We therefore read the city spread as mostly the effect of distance to the origin server. Kinsta says its CDN is powered by Cloudflare and reaches “300+ cities in more than 100 countries, according to its [Kinsta CDN documentation](https://kinsta.com/docs/wordpress-hosting/wordpress-cdn/kinsta-cdn/?utm_source=hostingguider.com). Hostinger lists “CDN includedon Cloud Startup on its [cloud hosting plans page](https://www.hostinger.com/cloud-hosting?utm_source=hostingguider.com), checked October 4, 2026. **For your site, this means:** if your visitors are spread across continents, ask whether your host or CDN can cache full pages at the edge. Caching only images and scripts won’t change the page’s TTFB. ## What to expect by hosting type With four accounts, our data shows what is possible on each type, not what every host of that type delivers. This is not a market survey. Other hosts, plans, server loads and test dates can produce very different numbers. | Hosting type | What our accounts showed | What it suggests | | --- | --- | --- | | Shared | 126 ms near the server (Namecheap), 191 ms in GTmetrix and 229 ms from São Paulo (Hostinger Business); 600 to 1,700 ms from the far side of the world | Fine for a local audience | | Cloud | 156.86 ms in Singapore; under 520 ms to Europe and the US West Coast | A good fit when most visitors share a region | | Managed WordPress | 82.79 ms from Singapore; under 800 ms to London, Frankfurt, Amsterdam and New York over 3 days | Fast near the server; far cities still cost hundreds of milliseconds | If your audience is mostly in Asia or Europe, comparing [cloud hosting plans by region](https://hostingguider.com/best/best-cloud-hosting/) is a sensible next step. For WordPress sites that need a choice of data centers, the [managed WordPress hosts we compared](https://hostingguider.com/best/best-managed-wordpress-hosting/) list where each one runs servers. ## How to measure your own TTFB Test from the cities where your visitors live, run each test more than once, and compare like with like. - Check your real visitor locations in your analytics. - Run a multi-location TTFB test from those cities. - Test a cached public page and an uncached page separately. - Repeat the test at a few different times of day, because one run is only a snapshot. Compare your numbers with the city table above, not with one headline figure. The [guide to reading a hosting speed test](https://hostingguider.com/blog/how-to-read-hosting-benchmarks/) explains which tools to trust and why single runs mislead. If your TTFB comes out high, the Website Speed Checklist sets the order to fix a slow site, starting at the server. ## Change log - **October 4, 2026:** Added all 10 Hostinger Business DebugBear cities and Kinsta’s Bangalore and Amsterdam results, taken from the published reviews. - **October 3, 2026:** First published with four accounts: Hostinger Business, Hostinger Cloud Startup, Kinsta WP 2 and Namecheap Stellar Plus. We plan to add longer Kinsta and Namecheap data as their monitoring history grows. ## Frequently asked questions ### What is a good TTFB in milliseconds? Up to 800 ms is good by Google’s rough guide. From 800 ms to 1,800 ms needs improvement, and above 1,800 ms is poor. In our tests, each account’s fastest city came in between about 80 ms and 230 ms. ### Is a TTFB of 200 ms good? Yes. 200 ms is well inside Google’s guide. In our data, numbers that low appeared only in each account’s single fastest location. ### Why is my TTFB high in some countries but not others? Each request has to travel to the data center and back, and a new connection needs several of those round trips. Visitors far from your server pay that cost every time. ### Is TTFB the same as server response time? Not quite. Server response time usually means only the server’s own work. TTFB, as web.dev defines it, also includes redirects, DNS lookup, and connection and TLS setup. ### Does TTFB affect SEO? Not directly. TTFB isn’t a Core Web Vital, according to Google’s web.dev guidance. It does come before the page starts drawing, so a high TTFB makes Largest Contentful Paint (LCP), which is a Core Web Vital, harder to pass. ## How we researched this The thresholds and definition come from Google’s web.dev TTFB article, rechecked on October 4, 2026. Kinsta’s CDN reach comes from its own documentation, rechecked the same day. Every city figure comes from a hosting plan we bought and set up with an identical WordPress install, following [how we set up each test account](https://hostingguider.com/methodology/). Test windows differ by account. Hostinger Business has run since December 2024 and Cloud Startup since November 2025. Kinsta covers only 3 days, and Namecheap’s history starts August 25, 2026.