Rage clicks and dead clicks
A rage click is when a user clicks the same small area several times in a row within a very short time. A dead click is a click that gets no response within a reasonable time. The screen doesn’t change, and no other page opens either. Microsoft Clarity flags both automatically. According to Clarity, a dead click can point to a broken element, a slow request or a misleading design.
In short
Both signals say the same thing: the user expected something, and it didn’t happen. The difference lies in their reaction. With a dead click, they click once and nothing happens. With a rage click, they press the same spot again and again.
These signals are produced by session recording tools. When I watch recordings, I look at these two filters first, because they quickly pull the problem moments out of thousands of sessions.
The signal shows you a place, not the reason. You find the reason in the recording, in the code or in a test with users.
What the tool counts
I give the definitions from the tools’ documentation. I rely most on the Microsoft Clarity documentation, because it describes the definitions in detail.
Rage clicks in Clarity
According to the Clarity documentation, a page view or session is flagged as “rage click” when a user clicks several times in quick succession in a small area. Each click in the cluster also counts as a rage click.
According to the documentation, a rage click can be a strong sign of frustration when the user expects an interactive element in that spot.
Dead clicks in Clarity
If a user clicks an element and gets no feedback within a reasonable time, the click is flagged as a dead click. The visual state doesn’t change, and no other page opens either.
According to the documentation, a dead click can point to broken elements, high-latency requests or a misleading design. The Clarity FAQ also gives the typical example: an element that looks clickable but has no coded response on click.
Error clicks
Clarity has a third signal. If a JavaScript error follows right after a click, the session is flagged automatically. In the click heatmap, these clicks have their own view called “error clicks”.
How the numbers are shown
According to Microsoft’s Clarity training module, dead and rage clicks appear on the dashboard as a percentage of sessions. So you see the answer to the question: in what percentage of sessions was there at least one rage click?
The Clarity documents I read give no numeric threshold. They say “quick succession” and “reasonable time”, but not how many clicks or milliseconds that means.
Contentsquare
Contentsquare’s blog post defines a rage click as repeatedly clicking a specific area or element within a short time. It writes that the tool has a rage click heatmap and a filter for repeated clicks in recordings.
I couldn’t open the Contentsquare and Hotjar help pages for this text. That is why I don’t give their thresholds. For choosing a tool, see the Clarity vs Hotjar comparison.
What causes them
The Clarity documentation names three possible reasons for dead clicks: a broken element, a high-latency request and a misleading design. I read rage clicks under the same three headings.
An element looks clickable but isn’t
Users click based on how an element looks. Underlined text, a card with a shadow, an image with a magnifier in the corner all invite a click. These visual cues are called signifiers; more on that on the page affordance and signifier.
The Clarity blog puts it plainly: a phrase or word on the website can look like a link. I say it briefly like this: what looks clickable gets clicked.
Slow response
The button works, but the result comes late. If the user sees no change in the meantime, they think the click didn’t register and press again.
According to Jakob Nielsen’s response time limits, 0.1 second feels like an instant reaction. 1 second is the limit for the flow of thought staying uninterrupted. For operations over 1 second, you need to show that the system is working.
According to Google’s web.dev, the response to clicks, taps and key presses is measured with INP. An INP up to 200 milliseconds counts as good, over 500 milliseconds as poor. Details are in the guide to Core Web Vitals.
Broken button
The button really doesn’t work. That can be a JavaScript error, a wrongly bound event or an area sitting under another layer. The Clarity blog names a CSS selector that didn’t work as an example.
The same post describes a finding from the Bing team. For 4% of the users who clicked into the search box, there were clicks that changed nothing on the screen. The team reduced the top, left and bottom spacing of the box. According to the post, these clicks then almost completely disappeared.
Live example: session recording of a product card
The card below is made up. It has three elements, and each behaves differently. As you click, the session recording next to it lists the events. It gives a badge to any click that meets the demo rule. On a phone, the recording sits below the card.
Rage click (demo rule): 3 or more clicks on the same element within 1 second.
Dead click (demo rule): when nothing changes on the screen in the second after the click.
The magnifier in the corner says the image can be enlarged.
Handmade ceramic mug
$29
In cart: 0 items.
Session recording
0 clicks · 0 dead clicks · 0 rage clicks
- No clicks yet. Click the elements on the card.
Without JavaScript the recording doesn’t work. The card is then just a sample picture.
A made-up product card. The rules are written for this demo and aren’t a real tool’s threshold. The image doesn’t respond, Add to cart stays silent for 2 seconds, Save works. Nothing is stored or sent.
Try this. Click the image once: after 1 second the dead click badge appears. Press Add to cart three times quickly: the rage click badge appears, and after 2 seconds there are 3 items in the cart.
Then press “Fixed version” and do the same. The image enlarges, and the button shows “Adding …” right away and doesn’t accept a second click. The operation still takes 2 seconds, but the user now knows what they are waiting for.
How to find them in recordings
- Look at the share of rage and dead clicks on the dashboard. If one page is clearly above the others, start there.
- Set that page’s click heatmap to the dead click and rage click views. How to read a heatmap is covered in the heatmap guide, so I won’t repeat it here.
- Choose the element with the most flags and watch the recordings in which it was clicked. Clarity’s click heatmap lets you open the recordings for a selected element.
- Look at mobile and desktop recordings separately. A small touch target can create dead clicks on phones only.
- Look at error clicks separately. Here you look at the code first, because a JavaScript error occurred right after the click.
How to fix them
The fix depends on the reason. I map every finding to this table.
| Reason | What it looks like in the recording | Fix |
|---|---|---|
| Element looks clickable but isn’t. | Dead clicks on the same element, sometimes rage clicks. | Make the element work, or take away its clickable look. |
| Slow response. | The first click is dead, then rage clicks on the same button. | Respond visibly to the click right away, and block a second click while the operation runs. |
| Broken button. | Dead click or error click. | Find and fix the bug, then record the same flow again. |
Two ways with an element that looks clickable
If users click an image, they probably want to enlarge it. Then the simplest fix is to do the job they expect.
If the element has no job, take away its clickable look: underline, hand cursor, shadow and magnifier icon. Use the link colour for links only.
Instant feedback for a slow response
Even if you can’t speed up the operation, acknowledge the click right away. The button text changes, or an indicator appears next to it. According to Nielsen, a response under 0.1 second feels instant, so keep the first reaction under that limit.
While the operation runs, the button doesn’t accept a second click. As the demo shows, an unprotected cart button turns three clicks into three items. For button states, see the CTA button pattern.
Testing a broken button
Watch the recording with the error click together with the developer. Take the browser, device and steps from the recording. After the fix, look again with the same filter for a few days.
Not every flag is a problem
The rule doesn’t know intent. The Save button in the demo works, but if you press it three times quickly, it gets the rage click badge too.
Real products have moments like these as well. Pressing the button to increase the quantity several times, or clicking quickly through a gallery, is normal. Contentsquare’s post also writes that some repeated clicks come from habit. Some users click on the page while reading, for example.
So don’t turn the share into a success metric on its own. For a flagged element, watch a few recordings and confirm that the user really got stuck.
My note
To me, the dead click is a more valuable signal than the rage click. The rage click comes after the user has lost patience. The dead click shows on the very first try that you created a wrong expectation.
The most often overlooked cause is the slow response. The team usually tries the button on a fast connection and never sees the delay. The user, on the other hand, stares at an unchanged screen for two seconds on a slow mobile network.
If you want an outside look at where users click into nothing in your product, I go through your product from start to finish with the UX Audit. You get what I find as a list of fixes, by priority.
Questions
What is a rage click?
A rage click is when a user clicks the same small area several times in a row within a very short time. Microsoft Clarity flags these clicks, and the session they occur in, automatically.
What is a dead click?
A dead click is a click on an element that gets no feedback within a reasonable time. The screen doesn’t change, and no other page opens. According to the Clarity documentation, this can point to broken elements, slow requests or a misleading design.
How many clicks does Clarity need to count a rage click?
The Clarity documents I read give no numeric threshold. They speak of clicks in quick succession in a small area. The rule of 3 clicks in 1 second in the demo on this page applies to the demo only.
What should I do if the share of rage clicks is high?
First find the page and the element with the highest share. Then watch a few recordings in which that element was clicked. Whether the reason is an element that looks clickable, a slow response or a broken button, fix it accordingly.
Sources (8)
- Microsoft Learn, Clarity: Semantic metrics. accessed November 4, 2026; definitions of rage click, dead click and click error
- Microsoft Learn, Clarity: Frequently asked questions. accessed November 4, 2026; questions on rage and dead clicks, example of an element that looks clickable with no coded response
- Microsoft Learn, Clarity: Click maps. accessed November 4, 2026; heatmap types dead, rage and error clicks, recordings for an element
- Microsoft Learn, Explore your website's usage data with Clarity (training module). accessed November 4, 2026; dead and rage clicks on the dashboard as a percentage of sessions
- Clarity Staff, Microsoft Clarity Blog, Rage clicks: what do they tell you about user behavior?. accessed November 4, 2026; June 23, 2021; text that looks like a link, CSS selector, Bing search box 4%
- Dana Nicole, Contentsquare, Rage clicks (blog). accessed November 4, 2026; last updated November 9, 2025; definition, rage click heatmap, clicks from habit
- Jakob Nielsen, Nielsen Norman Group, Response Times: The 3 Important Limits. accessed November 4, 2026; 1993, updated 2014; 0.1 s, 1 s, 10 s
- Jeremy Wagner and Barry Pollard, web.dev (Google), Interaction to Next Paint (INP). accessed November 4, 2026; updated September 2, 2025; up to 200 ms good, over 500 ms poor