Profiling React component performance with Chrome devtools

Ben Schwarz

November 28, 2017

Updated: October 27, 2020

React was one of the first major front-end frameworks to not only name, but tout its rendering performance. React’s virtual dom is famous for efficiently rendering components—but what happens when those components suddenly don’t feel fast anymore? Where do you look? How do you fix it?

In this walk-through, we’ll show how to find and fix slow React component code using Chrome devtools.

While this article is focused on React, you’ll learn concepts that can be applied to React, Angular, Vue or just about anything written in JavaScript.

There are three things you’ll need to follow along:

  • Use React 16 or newer, in development mode
  • Use Google Chrome or Chromium’s devtools (available in all current releases)
  • Export JS with source-maps (This is optional, but highly recommended)

👷‍ Setting up your audit

Firstly, open Chrome's devtools. You’re going to need some room to breathe, so undock devtools and maximise it to be as large as your screen allows.

When doing performance audits it’s essential that you're working towards realistic real world device targets. After all, not everyone has a powerful developer computer or current model phone.

Thankfully, Chrome devtools can synthetically throttle JavaScript execution. Throttling JavaScript execution in this manner makes performance issues far more apparent, so it’s a good idea to always debug performance using slowed performance.

Remember: Any improvements that you can make for slow hardware will also mean that fast devices get an even better experience. Everyone wins!

Slowing your likely powerful developer machine by at least 4x will have it perform similarly to a Motorola Moto G4.

🔬 Recording and viewing a performance trace

In development mode, as React renders, it creates ‘mark and measure’ events for each component. Those marks and measures can be viewed within Devtools.

To record a performance profiler trace, navigate to the page that you’d like to test (on localhost) and press the ‘Start profiling and reload page’ button.

This will record a performance trace for the current page. Chrome automatically stops recording the trace after the page has settled, though you can end it earlier by pressing the stop button.

Or, use the Mac keyboard shortcut: ⌘ + ⇧ + E

Once you’ve got a trace, your window will look something like this:

A timeline displaying page render, painting and main thread scripting duration.

Two items worth highlighting, that may not be immediately obvious to someone who is new to the Performance tab.

  1. This red bar indicator shows that there were significant long tasks around this part of the trace timeline. We probably want to investigate there.

  2. The colours used in the graph at the top of the performance window correspond to different types of activity. Each category has various causes, fixes, and analysis required.

In this article, we’re focusing on “Scripting” (JavaScript runtime performance).

Expand the window, click open user timing, and use the screenshot timeline to investigate how your page painted 👀

Now we’re going to want to investigate that red CPU-burn area. We can see that the page rendered elements on the page during that period of the trace.

Refine your zoom level by scrolling your mouse wheel while moused over the flame chart.

Zooming immediately displays user-timing information and a component labelled Pulse (that takes 500ms to render).

Beneath the Pulse component, there appears to be child components rendering, although the size of those items indicates that they aren’t costing a lot of execution time.


⏳ Discovering slow functions

  1. Click the component you’d like to investigate, in this case, Pulse. This scopes the lower portion of the window to focus only on the Pulse component.
  2. Choose the “Bottom-Up” tab.
  3. Sort by total time descending. In some cases, you may want to sort by Self time, or group by something other than the URL. Experiment what works best for what you’re investigating.
  4. Expand functions until you’ve located a point in the code you’d like to investigate. In this case the map function appears suspect. It amounts to 90ms of execution time.
  5. This is why you need sourcemaps: Clicking MetricStore.js:59 in the line number gutter will take you to that spot in the code. Let’s go!
Code that took longer to execute has a time indicator down the left hand side

🤝 Improvement is straight-forward when you know where to look

Using this approach, I’ve managed to shave seconds from areas of code that I thought “well, it’s complicated, so I guess it takes a while”. Now I know exactly how and where to look for janky JavaScript performance.

🛣 Sidenote: Visualise the impacts of third party scripts

Easily investigate which scripts or third parties are costing your customers time using Calibre’s Main Thread Execution Timeline explorer

Quickly identify which third party scripts attributed to main thread long tasks

🤔 Can I export these metrics? Or add my own?

When in development mode, React publishes component runtime metrics to the User Timing API, which is why it’s visible in devtools without any additional browser extensions.

You can use custom User-Timing events to instrument your sites and applications.

Here's a few examples that you may want to consider adding to your projects:

  • Webfonts loaded
  • Advertisement initialised
  • Add item to cart
  • Registration page: Time to register
  • Cart page: Time to checkout

All major browsers support the User Timing API, so it makes sense to add instrumentation to critial user-actions and flows to gain a deeper insight into the experiences people are having on your pages.

User timings can be delivered to Google Analytics, or to a custom collection endpoint using navigator.sendBeacon().

Profile your apps!

Being a successful performance-minded developer requires great focus and knowledge. The good news is you’re already half way there! Best of luck!

Ben Schwarz

Ben is the founder and CEO of Calibre. He regularly speaks at conferences worldwide and has been involved in far-reaching Open Source projects as well as web standards. Find him on Twitter or LinkedIn.

Related posts

How We Improved the Performance of Our Live Chat by 30%

Don’t miss another article

Be notified about new product features, releases and performance research.