Tools for measuring core web metrics
Posted: Sat Feb 08, 2025 10:45 am
One experiment showed that in animations, the illusion of one event causing another begins to break down at a delay of about 100 milliseconds. It has been suggested that the illusion of direct manipulation in user interfaces will break down with this amount of delay. (This experiment also apparently references earlier suggestions that actions should have visible effects within 100-200 milliseconds.)
As with LCP, there is no “hard number”, only markers that represent the distribution. Some people are much more sensitive than others and may notice shorter delays when tactile or auditory feedback is involved. In addition to changing the feel of the UI, delaying something that users expect can cause them to make mistakes. They may repeat an action because luxembourg mobile database they think it doesn’t work, but the second action may have a negative effect. For example, they may click the “Add to Cart” button twice without realizing they are now purchasing two items. The responsiveness associated with these experiences is measured by First Input Delay (FID), and we recommend that sites keep the FID below 100 milliseconds for 75% of page loads .
Influence
We analyzed millions of page renders to understand how these metrics and thresholds impact users. We found that users were 24% less likely to abandon a page load (leave the page before any content has rendered) when sites hit the thresholds above. We also looked specifically at news and shopping sites, whose businesses depend on traffic and task completion, and came up with similar numbers: 22% fewer abandonments for news sites and 24% fewer abandonments for shopping sites . Few interventions show this level of improvement for online businesses, and results like these are part of the reason we and our ecosystem partners are prioritizing Web Vitals metrics. Providing a smooth experience for users is one of the most effective ways to grow online traffic and web-based businesses. We hope that the Web Vitals metrics and thresholds will provide publishers, developers, and business owners with clear and actionable ways to make their sites part of a fast, smooth experience for more users.
As with LCP, there is no “hard number”, only markers that represent the distribution. Some people are much more sensitive than others and may notice shorter delays when tactile or auditory feedback is involved. In addition to changing the feel of the UI, delaying something that users expect can cause them to make mistakes. They may repeat an action because luxembourg mobile database they think it doesn’t work, but the second action may have a negative effect. For example, they may click the “Add to Cart” button twice without realizing they are now purchasing two items. The responsiveness associated with these experiences is measured by First Input Delay (FID), and we recommend that sites keep the FID below 100 milliseconds for 75% of page loads .
Influence
We analyzed millions of page renders to understand how these metrics and thresholds impact users. We found that users were 24% less likely to abandon a page load (leave the page before any content has rendered) when sites hit the thresholds above. We also looked specifically at news and shopping sites, whose businesses depend on traffic and task completion, and came up with similar numbers: 22% fewer abandonments for news sites and 24% fewer abandonments for shopping sites . Few interventions show this level of improvement for online businesses, and results like these are part of the reason we and our ecosystem partners are prioritizing Web Vitals metrics. Providing a smooth experience for users is one of the most effective ways to grow online traffic and web-based businesses. We hope that the Web Vitals metrics and thresholds will provide publishers, developers, and business owners with clear and actionable ways to make their sites part of a fast, smooth experience for more users.