Legacy FID vs Total Blocking Time (TBT)
You'll learn: How the lab metric Total Blocking Time (TBT) historically served as a proxy for the legacy First Input Delay (FID) metric during performance testing.
The Field vs Lab Challenge
Note: On March 12, 2024, Google officially replaced First Input Delay (FID) with Interaction to Next Paint (INP) as a Core Web Vital. FID is now a deprecated, legacy metric. Please refer to this course's INP lessons for current responsiveness metrics and ranking guidance. However, understanding how FID related to TBT provides important historical context for how lab and field metrics interact.
Historically, FID was a field metric—it only existed when real users interacted with a page in the wild. When developers were testing locally or in a CI/CD pipeline, there were no real users clicking buttons. To predict field responsiveness in a controlled lab environment, developers relied on Total Blocking Time (TBT).
What is TBT?
Total Blocking Time measures the total amount of time between First Contentful Paint (FCP) and Time to Interactive (TTI) where the main thread is blocked long enough to prevent input responsiveness. Specifically, it sums up the blocking portion of all long tasks (JavaScript tasks exceeding 50ms).
The Historical Relationship
Think of TBT as the historical "training dummy" for FID. While FID measured the actual delay of a single user interaction in the field, TBT quantified how much potential for delay existed during the page load in the lab. A high TBT strongly predicted poor FID scores because:
- Both metrics measured main thread blocking.
- Long tasks that increased TBT were the same culprits that caused high FID.
- TBT correlated strongly with real-world FID data.
Practical Example
If your page had three long tasks during load:
- Task 1: 70ms (20ms blocking beyond the 50ms threshold)
- Task 2: 100ms (50ms blocking)
- Task 3: 55ms (5ms blocking)
Your TBT = 20 + 50 + 5 = 75ms
This 75ms of blocking time indicated a congested main thread. In the era of FID, developers optimized TBT during lab testing to reduce the likelihood of real users experiencing input delays. Today, keeping TBT low remains a valuable practice for diagnosing main thread congestion, though your field measurement focus should now be on INP.
Key Takeaway: TBT is a lab metric that sums up main-thread blocking time; historically, it was used to predict and optimize the deprecated FID metric, laying the groundwork for how we diagnose responsiveness issues for modern metrics like INP.