What are frustration signals? Rage clicks, dead clicks and error clicks
The click patterns that mean a visitor is stuck, how to tell the three apart, when they are false alarms, and where they show up on a store, a SaaS site, a blog and a service site.

The complaint nobody writes down
A frustration signal is a pattern in a visitor's clicks, taps or mouse movement that usually means the page did not do what they expected. Nobody fills in a complaint form. The behaviour is the complaint. A session replay tool watches for these patterns automatically and flags the visits where they happen, so you find the stuck visitors without watching every recording.
The three most common signals are rage clicks, dead clicks and error clicks. They sound alike and they overlap, but each one points at a different kind of problem. A rage click says "I clicked and I am losing patience." A dead click says "I clicked and nothing happened." An error click says "I clicked and something broke."
Frustrated visits are common enough to be worth a regular check. One large 2026 benchmark, built from sessions on about 6,500 websites, found that just over a third of visits ran into some kind of friction.
Rage clicks
A rage click is several fast clicks in the same spot. Picture someone jabbing a button that will not respond, or tapping "Continue" three times because the page seems frozen.
Every tool draws the line a little differently. A common rule is three clicks within a second or so, landing within a few dozen pixels of each other. Some tools wait for five clicks. Some only count clicks on an element that then fails to respond. The exact numbers matter less than the shape: a burst of clicks, close together, in one place.
The usual causes are a slow page (the click worked, but nothing has shown up yet), a button that is disabled without looking disabled, a broken link, and an element that only looks clickable.
Dead clicks
A dead click is a click or tap that changes nothing. No page loads, no menu opens, no message appears. Most tools mark a click as dead when nothing on the page changes within a few seconds. One waits a full seven seconds before deciding.
Dead clicks are quieter than rage clicks. The visitor tries once, gets nothing, and moves on or leaves. That makes them easy to miss, and many visitors will not try twice.
They tend to land on things that look interactive but are not: a product photo that will not zoom, a heading that looks like a link, a logo in the middle of the page, a greyed-out button, a card with a hover effect and no destination. Nielsen Norman Group's eyetracking research from 2017 found that when clickable elements lacked clear cues, people spent about a fifth longer finding what they needed. Flat, minimal designs are a reliable source of dead clicks for exactly that reason.
Error clicks
An error click is a click followed straight away by a JavaScript error in the visitor's browser. JavaScript is the code that makes a page interactive, and when a piece of it fails, the browser logs an error the visitor never sees. Tools define the signal as a click or tap right before one of those errors. Some call it a click error.
Error clicks are the most useful of the three for a developer, because the recording comes with the error message attached. From the visitor's side, an error click usually looks like a dead click. The difference is that the tool can tell you why.
The other signals
Most tools go beyond clicks. A u-turn, sometimes called a quick back, is when someone opens a page and returns to the previous one within a few seconds. It usually means the page was not what the link promised.
A thrashed cursor is a mouse moving fast and erratically, or in circles. Tools read it as a sign the visitor is confused, lost or waiting for a page to load. It only exists on desktop, since there is no cursor on a phone.
Excessive scrolling is a visit with far more up-and-down scrolling than normal for that page. It often means the visitor cannot find something they expected to be there.
Rage clicks vs dead clicks: telling them apart
The two get confused because a rage click is often a pile of dead clicks. The simplest way to keep them straight:
Some tools count rage clicks as a subset of dead clicks, so do not worry if the same moment shows up under two headings. What matters is where the signal happened and what the visitor did next. The post on how to analyze session recordings covers the routine for that.
- A dead click is one click that did nothing. It tells you an element looks clickable and is not, or is broken.
- A rage click is a burst of clicks in one spot. It tells you the visitor cared enough to keep trying, which usually means a slow page or a broken control on the way to something they want.
- An error click is a click that triggered a browser error. It tells you where the code failed, and it is the one to hand to a developer.
When a signal is a false alarm
Not every burst of clicks is anger. Some people double-click everything out of habit from desktop software. Some click on the text while they read. On a phone, a double-tap to zoom, a fat-fingered tap and a tap meant to stop the page scrolling all land as repeated clicks. Some controls invite fast clicking by design: a quantity stepper, a photo carousel, a calendar's next-month arrow, a "next page" button. At least one vendor admits in its own documentation that next and previous buttons trip the rage-click rule even when they work perfectly.
So treat a single flagged session as a hint. A signal that repeats across many visitors on the same element is the thing to act on. A few recordings settle it fast. A working carousel looks nothing like a checkout button that refuses to respond.
On an ecommerce store
Product pages collect dead clicks on photos. Shoppers expect images to zoom or swipe, and when they do not, the tap goes nowhere. Size and colour swatches that are out of stock but still shown cause both dead clicks and rage clicks. A 2025 benchmark built from about 14 billion recorded sessions found rage clicks rising on product detail pages and filter controls, most of all on mobile.
In the cart, the classic rage click is a slow "update quantity" control, so the shopper clicks again and ends up with two of everything. Discount code fields collect rage clicks when a code is rejected without a clear message.
At checkout the signals get expensive. A pay button that appears to do nothing while the payment request runs in the background is the most common home for a rage click on a store. Error clicks cluster on address lookup and third-party payment widgets. Over Cyber Monday 2023, the same research team measured about 1.2 rage clicks per shopping visit across 100 large retailers, up from just under one the year before.
What to check first: the pay button, then any other button that submits a form. Watch the recordings where a rage click happened on checkout and see whether the visitor ever finished. The post on reducing checkout abandonment picks up from there.
On a B2B SaaS site
On a pricing page, the dead clicks land on the plan cards. Visitors click the plan name or the price expecting to start, and only the small button at the bottom does anything. A monthly-to-annual toggle draws rage clicks when it is slow to redraw the prices. Comparison tables get dead clicks on feature names that look like they should expand.
On the signup form, error clicks show up around email verification, "sign in with Google" buttons and password rules that reject input without saying why. A submit button that stays disabled until every field validates is a rage-click magnet, because nothing tells the visitor which field is wrong. The 2025 benchmark saw the same pattern in financial services, with rage clicks clustering on dropdowns, document uploads and "prove you are human" boxes.
Inside onboarding, the pattern is the disabled "Next" button and the empty state. A new user clicks a greyed-out control, or a dashboard tile with no data yet, and gets nothing. Dead clicks on a locked feature are useful in their own right: they show which upgrade prompts people actually want.
What to check first: the signup submit button, then the first screen after signup. If you are not sure where in the flow people stall, start with how to see where users get stuck.
On a blog or content site
Content sites have few buttons, so most signals are dead clicks on things that are not links. Readers click headings, author names, images and the logo. Share buttons that open nothing on mobile, or that load slowly, collect rage clicks. Embedded videos, social posts and maps are a common source of error clicks, because the embed's own code fails, especially for visitors who block third-party scripts.
Navigation is the other hotspot. A menu that opens on hover but not on tap gets dead clicks from every phone visitor. Category links styled as plain text produce u-turns and excessive scrolling as readers hunt for the next article.
What to check first: the mobile menu, then any embed on your ten most-read pages. A click map of each page will show whether readers are tapping things that do nothing.
On a lead-gen or local service site
Almost everything on these sites points at a form, a phone number or a booking calendar, so the signals bunch up there. Forms collect rage clicks on the submit button when validation fails silently, and error clicks when a spam check or third-party form service fails to load. A phone number written as plain text gets dead clicks from mobile visitors trying to tap and call. So does an address with no map link.
Booking widgets deserve their own check. Date pickers collect legitimate fast clicks on the next-month arrow and real rage clicks on a date that looks available and is not. Time slots shown but already full are a dead click every time. The same 2025 benchmark found rage clicks concentrating on date pickers and search filters in travel, and a plumber's or dentist's booking page uses the same controls.
What to check first: the contact form's submit button, then the phone number on mobile. If the form is your main source of leads, the post on form abandonment covers what to measure around it.
What to do with a signal
Start from the element, not the session. Most tools rank the elements on a page by rage clicks or dead clicks, and the top of that list is your to-do list. Bigdelta, for example, marks dead clicks and rage clicks on its click maps and lets you filter replays down to sessions with a rage click, a dead click or a JavaScript error.
Then watch three or four recordings for the top element. You are looking for one of a handful of fixes. If the element is not meant to be clickable, make it look that way, or make it work, since visitors clearly expect it to. If it is slow, add a loading state so the first click visibly does something. If it is broken, the error click hands the developer the error. If it is disabled, say why. If it is a false alarm, move on.
Last, check whether the signal sat on the path to something you care about. A dead click on a footer logo costs you nothing. A rage click on the pay button costs you the sale. The 2026 benchmark mentioned earlier found that cutting rage clicks per page by a point and a half added one page view to the average session. Fix the signals on the money pages first, then confirm the count dropped. That loop, from signal to recording to fix, is most of what website behavior analytics is for.


