Whether you are a novice or expert at website speed optimization, Fast or Slow provides insights to help you better understand how your site’s performance is experienced by your users no matter where they’re located. If you have feedback on Fast or Slow or issues to report, contact us at: fos-feedback@fastorslow.com. Note that this address is not intended to provide assistance with optimizing your site’s performance.

  • What is Fast or Slow?
    Fast or Slow is a global performance monitoring solution for your websites that detects and reports on the most important factors in your users’ experience of your website. Fast or Slow measures this experience from a variety of locations, giving you a wide overview of how site visitors experience your site.
  • How does Fast or Slow work?
    Fast or Slow measures a number of different parameters that help you understand your website’s performance from various locations around the world. These parameters include:

    • First Contentful Paint
    • First Meaningful Paint
    • Time to Interactive
    • Total Blocking Time
    • First CPU Idle
    • Speed Index
    • Round-trip Time
    • Transfer Size

    In addition, Fast or Slow performs audits on your website content to determine where speed improvements can easily be made, and keeps a history of past audits for each site. Fast or Slow is based on Lighthouse, an open-source, automated tool for testing web pages.

  • Where does Fast or Slow measure from?
    Fast or Slow measures your website’s performance from about 18 locations around the world. The number of available locations may vary at times. The tests are the same no matter which server is initiating the test.
  • Why do I need to use Fast or Slow?
    Your own experience of your website can be vastly different from someone else’s. A number of important factors contribute to a website’s overall performance and a user’s experience, including network round trip times dependent on where a site visitor is browsing from, and other factors.
  • Why do I need to monitor performance over time?
    Site performance can vary based on a number of factors including the amount of traffic to the site, recent changes, issues at your web hosting provider, and overall network conditions. By monitoring performance over time, you can get an idea of your site’s “baseline” performance, view the results of any changes you make, and look for patterns that may indicate problems with your web host.
  • How do report subscriptions work?
    Fast or Slow allows you to subscribe to regular performance reports for a site by entering your email address, agreeing to the Fast or Slow terms and privacy policy, and clicking Subscribe. You will receive a confirmation email, and will need to click “Confirm Monitoring” in this email to begin receiving reports. Note that since Fast or Slow does not require you to login, it uses Cookies to keep track of whether or not you have subscribed to reports, so visiting a performance report from a different device will not show that you have already subscribed. You can stop monitoring individual sites or unsubscribe from all reports via a link at the bottom of every report email.
  • How often are subscribed reports sent?
    Subscribed reports are sent once a week, starting from the first time any user subscribes to reports for a site. If another user has already subscribed to reports for a site, you may receive your first report less than a week after subscribing, and then in one week intervals thereafter.
  • How are metrics scored?
    All scoring is based on how the audited site compares to other sites that have been measured with Lighthouse, the technology used by Fast or Slow. For any score from 0-100, higher is better, while for any actual measurements of time, lower is better – that is, the less time a metric takes, the higher its score will be.
  • What is the Performance Score?
    The Performance Score is a weighted average based on the First Contentful Paint (FCP), First Meaningful Paint (FMP), Total Blocking Time (TBT), Speed Index, Time To Interactive (TTI), and First CPU Idle Scores. Higher scores indicate that your site is likely faster than average, while lower scores indicate that your site may be slower than average.
  • What does First Contentful Paint measure and how is it scored?
    First Contentful Paint (FCP) is triggered when any content within the DOM (Document Object Model) is displayed on the visitor’s screen. A “Paint” specifically refers to the browser displaying any part of the webpage – this could be text, an image, or a canvas render.

    This measurement attempts to be more representative of your user’s experience on your site. It measures when actual content has been loaded into the page, not just a change in the page in the browser. As the focus is on the content loaded in a user’s browser, it gives you an idea of how quickly a user receives content like text or images. It is a more useful indication of a user’s experience than measuring when a style or background image or color happens.

    A First Contentful Paint that occurs in 900 milliseconds or less will result in a score of 90 or higher, whereas a First Contentful Paint that takes more than 1.6 seconds will result in a score of 49 or lower.

  • What does First Meaningful Paint measure and how is it scored?
    First Meaningful Paint, often called FMP, is the time it takes for a page’s primary content to appear on the screen. Coined by Google, FMP is often used as a primary measurement for user-perceived loading experience. Paint refers to the time at which the browser displays web page content onto a user’s screen. A first meaningful paint suggests that the content of the page, versus a header or navigation, is shown to the user.

    A First Meaningful Paint that occurs in 900 milliseconds or less will result in a score of 90 or higher, whereas a First Meaningful Paint that takes more than 1.6 seconds will result in a score of 49 or lower.

  • What does Total Blocking Time measure and how is it scored?
    Total Blocking Time (TBT) measures the amount of time between First Contentful Paint (FCP) and Time To Interactive (TTI) where the site was unresponsive to input. Any task during this time that takes more than 50milliseconds is considered “Blocking,” since this is about the maximum amount of time a user can wait for their input to be processed before delay is noticed. For any task that takes longer than 50milliseconds, the extra time it took beyond 50milliseconds is added to the Total Blocking Time.

    A Total Blocking Time of 200 milliseconds or less will result in a score of 90 or higher, while a Total Blocking Time greater than 600 milliseconds will result in a score of 49 or lower.

  • What is the Speed Index and how is it scored?
    The Speed Index Score is based on how quickly the page becomes visibly populated, compared to other sites that have been audited. A Speed Index score over 90 indicates that the page became visibly populated within 1.2 seconds, while a score under 49 indicates that it took more than 2.3 seconds.
  • What does Time to Interactive measure and how is it scored?
    The Time to Interactive (TTI) metric measures how long it takes a page to become interactive to the site visitor. A web page is interactive when it has displayed useful content, event handlers are registered for most visible page elements, and the page responds to user interactions within 50 milliseconds.

    Useful content means the primary content of the page such as article contents and headlines and does not include header images, navigation, and other elements of a web page.

    A Time To Interactive score of 2.3 seconds or less will result in a score of 90 or higher, while a Time To Interactive of more than 4.8 seconds will result in a score of 49 or lower.

  • What does First CPU Idle measure and how is it scored?
    First CPU Idle measures how long it takes a page to become minimally interactive. A page is considered minimally interactive when:

    • Most elements on the screen are interactive
    • Page responds, on average, to most user input in a reasonable amount of time (less than 50milliseconds).

    Both First CPU Idle and Time to Interactive measure when the page is ready for user input. First CPU Idle occurs when the user can start to interact with the page; TTI occurs when the user is fully able to interact with the page.

    A First CPU Idle that occurs in 2.3 seconds or less will result in a score of 90 or higher, while a First CPU Idle of more than 4.8 seconds will result in a score of 49 or lower.

  • What does Round-Trip Time measure?
    Round-trip time (RTT), often referred to as network latency, measures the length of time it takes for a request to travel through the network to its destination plus the length of time it takes for an acknowledgement of that request to travel back.

    Measured in milliseconds, the measurement of RTT begins when a browser first sends out a request and ends when a response is returned. RTT is a critical aspect of your website’s page load time, since the browser cannot begin to parse content until it receives a server response.

    Factors influencing RTT include the physical distance between the visitor and server, server traffic volume and the number of intermediate nodes the request needs to pass through. Fast or Slow measures the RTT for each host a request is sent to when a page is loaded, and displays the average in the Performance overview as well as a more detailed breakdown in the Audits section.

  • What does First Byte measure?
    First Byte measures the amount of time between the start of profiling and the time that the first byte arrives over HTTP/HTTPS in the visitor’s browser. The variables measured by Fast or Slow include DNS, socket connection time, the time taken to send the HTTP request, and the time taken to get the first byte of the page. The First Byte timing is dependent on network latency as well as server responsiveness, as it can include the time for a CMS to respond to a request with headers and content. A very low First Byte time is observed with statically served or cached web pages that do not require extensive computation, database connections, or other backend processing. A higher duration is often seen with larger or dynamic websites leveraging database connections.

    Improving the First Byte timing requires improving your network availability as well as web server performance.

    Moz has found that websites with lower First Byte scores rank better in search engine result pages.

    Note that some tools and sites use varying definitions of First Byte or Time to First Byte (TTFB). Some will count only the duration between the end of the HTTP request and the first byte of the response received, ignoring DNS lookups, redirects, and some network latency.

  • What are page size and transfer size?
    Page size measures the full size of the content on the page after it has been decompressed, and it is shown on the profile detail for each location. Transfer size measures the actual size of the data transferred when a page is requested. This includes HTTP headers but it can be much smaller than the Page size, since most sites will compress data before transferring it. It can also vary slightly between locations because the site might respond with different headers or include different scripts depending on where the page is requested from. Transfer size is shown on both the main profile summary and on the detail for each location.
  • What does Requests measure?
    Requests measures the total number of network requests made when a page was loaded. It includes requests for images, CSS, JavaScript, and any other resources that are included as separate files from the main page.
  • What is more important, First Meaningful Paint or First Contentful Paint?
    For measuring an actual user’s experience, First Contentful Paint is more important. Usually, however, these two metrics end up being very similar. There are specific situations where you might want to look at one metric over another.

    On large, complex websites with many different assets, you may see First Meaningful Paint and First Contentful Paint as different values. This is because there are many different assets and elements of your site that need to be loaded into the browser. First Contentful Paint would be the best metric to use on such sites.

    On smaller and lighter websites, these two metrics are often the same value because the browser is able to quickly render page elements. As these values are the same, either will suffice.

  • What are CSS and JavaScript files?
    CSS files are used to store information on how a website is styled, and include instructions that impact fonts, colors, widths, and more. JavaScript files are used to add functionality to your site and can be used for almost anything that a website can do, up to and including rendering the entire site.
  • What does minify CSS/JavaScript do?
    JavaScript/CSS minification is the process of removing extra characters, such as whitespaces, line breaks, and comments from CSS and JavaScript files. This not only makes these files smaller, causing them to be transferred more quickly, but can also allow the browser to process them more quickly in some cases. Combining multiple CSS or JavasScript files also results in fewer overall requests, further improving performance.
  • How do I optimize my images?
    Anything you can do to minimize the transfer size of images without visibly diminishing their quality can be considered optimization. There’s generally 3 steps you can take to ensure that images are optimized:

    • Resize image files to fit the space they’ll take up on the webpage – if you’re going to be displaying a header banner that’s 900 pixels wide by 20 pixels tall, make sure that the file you’re using is of the same dimensions.
    • Compress image files as much as possible without visibly impacting quality
    • Use next-gen formats for your images, such as JPEG2000, JPEG XR, and WebP, all of which offer better compression than GIF, JPEG, and PNG formats.

    There are a number of plugins that can assist with these measures, but be aware that not all browsers support next-gen formats, so you may need to implement a system to fall back to older formats for these browsers, which several plugins can also assist with.

  • How do I improve caching efficiency?
    Generally, it’s important to set a long TTL (Time To Live – the amount of time a browser should keep a cached copy before requesting an update) on any resource that won’t change very often, such as Images, CSS files, and JavaScripts. There are a number of plugins that you can use for this, and many CDNs also allow you to control caching for certain resources. Bear in mind that this may not always be possible with advertising scripts or other resources loaded from a third party, though there are plugins and other methods for keeping an up-to-date copy of third party scripts locally on your server, which will allow you to control caching for them.
  • What does bootup time mean?
    Bootup time is the amount of time spent parsing, compiling, and executing scripts on the site. A high bootup time often correlates with a poor Total Blocking Time and can negatively impact First CPU Idle and Time To Interactive as well.
  • How do I improve bootup time?
    The easiest way to improve bootup time is to minify and compress all JavaScript, and defer the parsing of any JavaScript that is not critical. Note that you should be careful not to defer scripts that are necessary to render the site (such as jQuery on most WordPress sites). Depending on your comfort level with JavaScript you can also use code-splitting to serve critical portions of your scripts first, and to prune unnecessary code.
  • What does DOM size mean?
    DOM, the Document Object Model, is a way to describe and manipulate web pages as a tree of nested elements such as div tags, images, scripts, and hyperlinks. DOM size measures the number of distinct elements on the profiled page, the maximum depth at which they are nested, and the maximum number of “child nodes” any element has. Each of these has an impact on how the webpage performs. Pages with more than 1,500 elements can cause performance issues, as can pages nested more than 32 elements deep. Pages that contain elements with more than 60 “child nodes” can also cause problems.
  • I have failed the DOM size test, how do I fix this?
    3rd party scripts and iframes, especially ones designed to embed advertising or other content, can add a large number of DOM nodes to your site that aren’t necessarily apparent from viewing the page source. While they may be necessary to your site, it can help to check that duplicate versions of these scripts are not present, that you’re not using multiple providers for a given purpose (advertising, retargeting, analytics, etc.), and that you’re following your ad provider’s guidelines for the number of ads to display on a given page.
  • How do I defer parsing of JavaScript?
    If you’re using a WordPress site, there are a number of plugins available that can defer JavaScript. Be careful about which scripts you defer, since some are not designed to be deferred, and they may not run correctly. If you’re not using a WordPress site, you can change your site’s “<script src=” tags to “<script defer src=” in order to defer loading of the scripts in question. You can find out more about this at:


  • What does the map indicate?
    The map indicates all of the locations that Fast or Slow is measuring your website from, as well as the location(s) where content on your website is being hosted and the distance between these locations. Note that if your site is using a CDN, each testing location may show content being loaded from a different region. Hovering over a location shows the combined number of requests served from that location to all testing sites, not the number of requests served to any individual testing site. Clicking a source location will display the URLs loaded from each location.

    Some sites such as Facebook use advanced methods to deliver content regionally from a single IP, but the IP’s location can only be displayed in one place. If content delivered from one of these sites does not cause slow results in the report produced by Fast or Slow, the location should not be a concern.

  • Why do the location details show different information than the summary?
    In addition to taking different amounts of time for different network-related tasks, your site (or scripts and resources loaded by it) may be configured to display different content when accessed from different locations. This can lead to different performance scores. Additionally, the location detail shows more details for certain metrics, including Loading Progression Graphs, which are extremely useful for determining the cause of issues, as well as less impactful metrics such as Time To Last Byte.
  • How do I interpret the Loading Progression Graph?
    The Loading Progression Graph breaks down each network request made when visiting a site, displaying when each request started and how long each portion took. You can also hover over each request to view exact timings. It is normal for some advertising and tracking scripts to begin after the “load” event. Some relatively uncommon features like large auto-playing videos may cause some events in this graph to take longer than expected, and these may also affect the performance scores.
  • How do I interpret the Content Graph?
    The Content Type graph breaks down requests by content type, displaying the size, transfer size, number of requests, and percentage of total requests used for each type of content. The HTTP Status graph displays which status codes were received when loading your site as well as how many requests resulted in each status code, possibly including redirects and various HTTP errors. The Secure graph displays the number of requests that were made over a secure (https://) connection vs. an insecure(http://) connection.
  • How do I interpret the Hosts Graph?
    The Hosts Graph displays how much total time was spent loading content from each host. Note that this graph is cumulative, and in many cases requests are made simultaneously, so total times may be longer than actual site loading times.
  • What can I do to improve my site’s performance?
    Typically the most impactful steps you can take to improve your site’s performance include:

    • Minifying CSS and JavaScript
    • Optimizing Images
    • Improving Caching efficiency
    • Using a CDN to reduce network delays
    • Minimizing the use of externally loaded scripts and resources

  • Why might I get different scores from Fast or Slow than from similar tools?
    The score generated by Fast or Slow may differ from the score generated by similar tools because it audits a page from multiple locations and uses an average of the scores from these locations. It also takes into account metrics that other tools may not (such as Total Blocking Time, which will be used in Lighthouse v6 but is not currently used in v5) or it may weigh them differently. Additionally, the Lighthouse audit in Chrome throttles CPU and Network by default in order to simulate lower-end devices, and the speed of your own computer and network connection will also influence these scores.
  • Do I need to worry about a low performance score for countries where I don't have any customers?
    This largely depends on whether you want customers in these countries. Most search engines preferentially return location-specific results, so poor performance in a given region will impact both conversion and search engine performance in those regions, but will likely not impact these metrics in other regions.
  • What overall minimum performance score should I have?
    A Performance score of 49 or lower is considered “poor” – although we cannot recommend a performance target, be aware that site speed can influence Search Engine rankings as well as conversion rates. While it’s typically not possible to obtain a score of 100 across all metrics, a score of 75 or greater is considered “good”.
  • If I have a bad performance score will it affect my rankings in all search engines or only certain ones?
    Site speed directly impacts search engine rankings on Google. Additionally, Bing now also considers page load time as a factor when determining search engine rankings. Yandex does not directly consider site speed, but they do consider time spent on the site and whether or not a visitor goes to a competing site afterwards, both of which are heavily impacted by site speed.
  • Does it matter if my score is okay one day and not good the next day? Will it affect my rankings in all search engines or only certain ones?
    This depends largely on your site’s speed at the time it was crawled by a given search engine and the degree of difference in speeds. As a general rule, search engines do not crawl sites every day unless they are very highly ranked, but you can typically request a re-crawl after making optimizations.
  • Why do similar tools have more recommendations for performance improvement than Fast or Slow?
    Fast or Slow focuses on audits that can typically be implemented within a CMS without directly editing code, such as by editing content or using plugins. Additionally, it focuses on audits that most directly impact the performance score and hence search engine rankings, while other performance audit items may not.
  • How do I resize and compress my images if there are no plugins available for my content management system or because I am not using a content management system?
    In most cases, you can download the images from your site, resize them using a photo editor, and use one of a number of online and offline tools to compress them. Once you have done this, you can re-upload your optimized images.
  • How do I create an image fall back system for JPEG2000, JPEG XR, and WebP if I can't use a plugin for this?
    If you have full control over your site’s code, you can often wrap a <picture> element around your <img> tags to define a webp version of an image and a fall back version. More information about this can be found at:


    You can also use the site’s .htaccess file to setup an image fall back system, though this may not work on all hosts.

  • Why might I get different scores when running the same test on Fast or Slow?
    Scores can vary based on a number of factors, including network congestion, various types of caching, and load on the site being audited. One of the reasons Fast or Slow keeps a history of past audits is to make it easier to keep track of overall trends, which matter more than individual scores at any given point in time.
  • How do I minify CSS and JavaScript files if there isn't a plugin available?
    There are a number of online tools to minify both CSS and JavaScript files – in most cases, you can simply download a copy of the script you wish to minify, run it through the minifier, and re-upload the minified version. Many of these tools will also contain more detailed instructions on their use. You can find more information about this at:


  • I've combined/deferred my JavaScript files and now my site is broken, how do I fix this?
    If possible, restore a backup of your site from before the changes. If you do not have any backups available, you may be able to restore functionality by removing the “defer” directive from scripts, by sourcing the combined JavaScript files from the header, rather than the footer, or by changing the order of scripts in the combined file, as some JavaScripts rely on others running first in order to function properly.
  • How do I implement caching if I am not using a content management system or the content management system I am using doesn't have a plugin for this?
    If you’re using a CDN, many of them allow you to manage caching from their control panel. Additionally, on most hosting accounts, you can configure caching using the site’s .htaccess file (for sites running Apache), the NGINX configuration file, or the IIS manager.
  • How long should I cache each type of resource?
    As a general rule, you should set a long expiration for files that you don’t expect to change for a long time, such as images and fonts, with a shorter expiration for files that may be changed occasionally. Here are some recommended guidelines:

    • Images, Video files, and Fonts: Access plus 1 year
    • JavaScript, CSS, and PDF files: Access plus 1 month
    • Most others: Access plus 1 week

    When using any CMS, be careful with setting cache times for JavaScript and CSS files. These files may have a version number in the URL after the filename, but some plugins or CDNs may strip or ignore the version. If the JavaScript or CSS file is updated while visitors still have a cached copy without a distinguishable version, their browsers may still use the old cached version, which may not be compatible with the rest of the page.

  • I have installed a caching plugin, made a change to my style sheet and my customers are seeing the old version of the site. How do I fix this?
    It is important to differentiate between files that are cached on your site and files that are cached in your visitors browsers – if the issue is with files cached on your site, you should be able to purge them so that your visitors can see the change, but if the issue is with files cached in your visitors browsers, they will not see the change until their browser’s cache expires, or unless they visit the site in a different browser. It is possible to work around this by adding a version query string parameter to cached resources, though this may require modifying your website’s code.
  • I have resized and compressed all of my images, minified my CSS and JavaScript files, installed a caching plugin and I still have a bad performance score summary overall, how do I fix this?
    In cases where a site performs poorly despite following all recommendations, it’s important to be aware of which scores are the worst: If your First Byte and Round Trip Times are poor, then you may be able to improve these scores using a CDN, though in some cases it may be necessary to switch hosting providers. If your First Byte and Round Trip Times are fast, but your Total Blocking Time, First Meaningful Paint, or Time to Interactive scores are slow, then this may indicate that your site is relying heavily on client-side JavaScript parsing in order to render content, which may require choosing a different theme or more extensive development efforts to correct.
  • I can't afford a web developer, so how do I fix the things myself that the report is telling me to fix?
    Do it Yourself website maintenance often requires an investment of time to learn the basics, but there are numerous educational resources specifically targeted at fixing each item in the audit list, since our audits are fairly standard.

    Some excellent resources to learn more about these issues are available at https://web.dev/learn/

    For more WordPress-specific information, visit https://wordpress.org/support/article/optimization/#Improve_WordPress_Performance

  • I've signed up to a CDN and now Wordfence is blocking everyone including me, how do I fix this?
    If you are locked out after placing your site behind a CDN, it is likely that Wordfence is no longer correctly detecting visitor IPs. In cases like this, you may need to forcefully regain access to your site. Once you have regained access and Wordfence has been reactivated, you’ll need to go to All Options and update “How does Wordfence get IPs?”. If your site is using Cloudflare, use the Cloudflare “CF-Connecting-IP” HTTP header. If you are using a CDN other than Cloudflare, you will most likely need to use the X-Forwarded-For HTTP header.
  • What can I do if the profiling server was blocked from loading the website?
    This indicates that Fast or Slow was unable to connect to your site from one or more locations. This can happen if your site is blocking access from certain countries or regions, or if your hosting provider is blocking access from devices using the Chrome browser in headless mode. If you do not want users to access your site from certain countries or regions, you can typically ignore results from these regions.
  • I am going to move to a different hosting provider that has better, faster servers. How can I test how good their servers are before I migrate my site?
    Many hosting providers have marketing material showcasing sites hosted on their servers. You can use Fast or Slow to profile these sites, and pay specific attention to the First Byte and Round Trip Times, which are typically good indicators of hosting performance.
  • I've upgraded my hosting plan for more server resources and I still have a low performance score after implementing all recommendations so how do I fix this?
    In cases where upgrading your hosting does not correct performance issues, you may also need to use a CDN, since even a site on a fast hosting account may take longer to load from geographically distant locations. Additionally, certain themes can dramatically impact performance by loading and executing large amounts of unnecessary JavaScript as well as requiring extra processing on the server itself. If temporarily switching to a basic theme improves your performance scores, then this indicates that your theme is likely contributing to poor performance.
  • How can I score 100 on all measurements?
    It is typically not possible (or necessary) to score 100 on all measurements, but by following the recommendations provided by Fast or Slow, most of a site’s metrics can be significantly improved. Some metrics, such as Total Blocking Time and First Byte, are largely dependent on the technology used to build the site and where it is hosted, so it may not be possible to improve these beyond a certain point without significant investment.
  • Does Fast or Slow measure mobile content?
    Fast or Slow does not currently measure how sites perform on mobile devices, but we plan to add this functionality in the near future.
  • How will using a CDN (Content Delivery Network) change my Fast or Slow scores?
    Using a CDN will typically result in improved scores on every metric except Total Blocking Time, since it speeds up all network-related tasks, especially Network Round-Trip Time (RTT). Total Blocking Time is an exception because it is based on how long a site takes to render after the majority of data has been transferred. A CDN will especially improve scores from locations that are geographically distant from where your site is hosted.
  • Why does a site spend so much time on redirection?
    Fast or Slow emulates how a browser loads a site. Based on how the site is submitted, this may involve the site redirecting from http:// to https:// as well as redirecting to (or away from) a www subdomain. Individual resources on the audited page may also require redirects depending on how they are included.
  • Why is my page/site so slow?
    There are many factors that can contribute to a site being slow, but Fast or Slow runs a number of audits that can help you pinpoint these factors, in addition to displaying which parts of the loading process need the most improvement. Additionally, auto-playing videos may result in a slower user experience, especially if they use workarounds to bypass browser auto-play settings.