The 8-second attention span claim is a myth. Explore what research actually shows about website first impressions, scanning, credibility, and page load speed.
A visitor clicks a link. Before they read any word, they can start forming an opinion of your brand, your business, based on how your website looks. This is not a bias we have because our website is beautiful (and it is VERY beautiful, if we do say so ourselves).
Researchers have spent years studying what happens in those first moments: how quickly people form visual impressions, how they scan pages, what influences trust, and how load time affects whether they stay.
This is where the famous eight-second attention span idea usually enters the conversation.
You have probably seen the claim that human attention span has shrunk to eight seconds, shorter than a goldfish. It shows up in conference talks, marketing decks, and articles almost exactly like this one. It is also unsupported, and tracing why the number became so widely repeated turns out to explain more about visitor psychology than the myth itself ever did.
If that claim was true, website owners, writers and creators would have a simple problem to solve. Capture attention before an invisible clock runs out.
People do not arrive on a website with a universal countdown running in their heads. Rather, there are different things that account for how long you're able to capture the attention of your audience: website loading speed, website aesthetics, mobile responsiveness, clarity and simplicity of the information on your website.
Let's look at why this eight-second claim became famous and what was actually said.
The number that never existed
In this section
The report simply repeated the eight-second figure alongside other figures it attributed to Statistic Brain.
The eight-second claim traces back to a 2015 Microsoft Canada report on consumer attention. The report presented figures saying that the average human attention span had fallen from 12 seconds in 2000 to eight seconds in 2013, with a goldfish listed at nine seconds. But those figures were attributed to a company called Statistic Brain rather than to Microsoft's own research.
In 2017, the BBC investigated the statistic and found that Statistic Brain could not provide a source for the eight-second figure. The other sources it listed, including the U.S. National Library of Medicine, had no record of the research either. The figure therefore had no identifiable underlying study supporting it.
Regardless, the number stuck. Even after the story behind it had been questioned, people kept repeating it. Eight seconds was easy to remember, easy to put in a presentation, and even easier to make memorable when a goldfish was part of the story.
And it worked. A 2022 survey by King’s College London found that half of UK adults still believed the average adult attention span was eight seconds, five years after the BBC investigation. Only a quarter knew that the figure was false.
Meanwhile, Microsoft's own research that has been used as a reference point consistently was looking at something different.
The report looked at how people use digital media, how often they switch between different activities, and how well they can keep their attention when they are dealing with several things at once. It was interested in how our increasingly digital environment affects attention and how people respond to different kinds of information.
That is very different from testing whether a person has exactly eight seconds of attention when they open a website.
Microsoft did not conduct a study that put a timer on someone's visit to a webpage and found that their attention stopped at eight seconds. The report simply repeated the eight-second figure alongside other figures it attributed to Statistic Brain.
So when the number is presented today as though Microsoft discovered that humans have an eight-second attention span, that goes further than the research actually did.
Attention simply does not work like a countdown clock.
Someone can leave a webpage after two seconds because it is taking too long to load, then sit through a three-hour film without checking the time. The two situations are not really measuring the same thing. What someone pays attention to depends on what they are doing, why they are doing it, how interested they are, and what is happening around them.
So there is no evidence that a person's attention simply switches off when eight seconds pass.
But if eight seconds is not the number we should be worrying about, is there a number that tells us how quickly people can react to a webpage?
There is.
And it is 50 milliseconds.
That sounds almost made up. It is 0.05 seconds. But unlike the eight-second figure, this number comes from a peer-reviewed study that specifically tested how quickly people could form an impression of a webpage's visual appearance.
What the 50-millisecond study gives us is: people can start judging how a webpage looks before they have had time to process what it actually says.
In 2006, a research team led by Gitte Lindgaard at Carleton University in Canada set out to find out how quickly people could form an impression of a website's visual appeal.
In the first two experiments, participants were shown images of website homepages for 500 milliseconds, or half a second. They were then asked to rate how visually appealing they found each page.
For the third experiment, the researchers made the viewing time much shorter: just 50 milliseconds, or one-twentieth of a second. The participants saw the same kind of homepage images and rated them again.
The researchers then compared the ratings from the 50-millisecond and 500-millisecond exposures to see whether people were making similar judgments even when they had barely had time to see the page.
The ratings from the 50-millisecond and 500-millisecond exposures were highly correlated. In simple terms, people tended to rate the pages in much the same way whether they saw them for 50 milliseconds or half a second. That led the researchers to conclude that people could form an impression of a webpage's visual appeal within 50 milliseconds.
But 50 milliseconds is not enough time to read a webpage, understand what it says, or decide whether the information is useful or trustworthy. The study was not testing any of those things.
It was testing: how quickly someone can look at a webpage and decide whether they find its appearance appealing.
The 50-millisecond finding tells us something quite specific. A visitor can form an impression of how a webpage looks before they have had time to read what is on it.
It does not tell us what happens after that. The study did not test whether people stayed on the site, trusted the information, or ignored problems because they liked how the page looked.
Other research has looked at website credibility and found that visual design can influence how people judge whether a site is trustworthy. But that is a different finding from a different study.
What the 50-millisecond study gives us is: people can start judging how a webpage looks before they have had time to process what it actually says.
What the eyes actually do once the page loads
In this section
In its 2006 eye-tracking research, researchers observed 232 users looking at thousands of webpages across different sites and tasks.
If the first milliseconds tell us how quickly people can judge a webpage's appearance, the next question is: what do they actually look at once they start moving through the page?
Researchers have studied this using eye-tracking, which records where people's eyes move and where they stop as they look at a screen. This is useful because people are not always able to accurately remember or describe how they browsed a page after the fact.
One of the best-known studies comes from the Nielsen Norman Group. In its 2006 eye-tracking research, researchers observed 232 users looking at thousands of webpages across different sites and tasks. They found a common scanning pattern that looked roughly like the letter F.
People tended to make a horizontal scan across the upper part of the page, followed by a shorter horizontal scan further down. After that, their eyes moved more vertically down the left side of the page.
The F-pattern does not mean people are biologically programmed to read websites in the shape of an F. It is a pattern researchers observed in how people scan web pages, particularly when they are trying to find useful information quickly.
And the way a page is organised can change how people scan it.
Nielsen Norman Group's later research found that when pages have clear headings and subheadings, users can follow those instead. Their eyes move from one heading to another, and when they find a section that interests them, they are more likely to read the text underneath it. The researchers call this the layer-cake pattern.
That makes the F-pattern easier to understand. It is not a rule that every website visitor follows. The way you organise information gives people different ways to move through a page. If everything is presented as one long block of text, scanning is more likely to concentrate around the upper and left parts of the page. Clear headings give visitors obvious places to pause, find what they need, and decide what to read next.
There is also another research that we need you to see: In 2008, Nielsen analyzed how much text people were likely to read during an average webpage visit. He estimated that visitors could read at most 28% of the words on a page if they spent all of their available time reading. About 20% was considered more realistic. The pages in the dataset contained an average of 593 words.
This does not mean that every visitor reads exactly 20% of every page. It was an estimate based on how long people stayed on pages and how quickly they could theoretically read the available text.
But it lets us know that: people are often deciding what deserves their attention before they have read everything on the page.
What visitors are actually judging
In this section
In a 2003 study, 2,684 people evaluated two live websites covering similar subjects.
Scanning behaviour explains where the eyes go. It does not tell us everything about what people are thinking while they look.
For that, one of the most detailed studies comes from the Stanford Persuasive Technology Lab, led by psychologist B.J. Fogg. In a 2003 study, 2,684 people evaluated two live websites covering similar subjects. Participants were also asked to explain why they found the websites credible or not, and the researchers grouped those comments into different categories.
The category mentioned most often was what the researchers called “design look.” It appeared in 46.1% of the comments. This covered the site's visual presentation, including things such as layout, typography and other aspects of how the site looked. The next most frequently mentioned categories were information structure and information focus.
But there is an important limit to what the 46.1% figure tells us.
The researchers were not asking participants to rank the things that make a website credible from most important to least important. Instead, they asked people to explain why they found a website credible or not, then looked at the features people mentioned in their answers.
So when we say that “design look” appeared in 46.1% of the comments, we are saying that 46.1% of the comments mentioned something about the site's visual presentation. It does not mean that design accounted for 46.1% of a person's trust, or that design was 46.1% responsible for the credibility judgment.
What makes the finding useful is that people were not only talking about the information they found on the websites. They were also talking about how those websites looked. Things such as the layout, typography and overall presentation were part of the reasons people gave for deciding whether a site seemed credible.
That does not mean appearance can replace good information. A website can look polished and still contain inaccurate information, and this study did not test whether attractive design could make people ignore those problems.
It tells us something more specific: the way a website looks can become part of the judgment people make about whether that website is credible.
Now put that together with the 50-millisecond study. People can form an impression of a website's visual appearance extremely quickly, and separate research shows that visual presentation can play a role in credibility judgments. The two studies did not test the same thing, but together they help explain why the first moments on a website matter.
A visitor may start deciding whether a page looks worth their attention before they have had time to read the evidence, assess the expertise behind it, or examine the quality of the information itself.
Those things still matter. The first impression simply happens before most of that evaluation does.A visitor can form an impression of how a page looks extremely quickly. Separate research shows that visual presentation can also become part of how that visitor judges a site's credibility. But neither finding means people stop caring about evidence, expertise or the quality of the information itself.
Those things still matter.
The point is that a visitor may begin deciding whether a page deserves their attention before they have taken the time to examine any of them.
The other filter: how long the page takes to even appear
In this section
The numbers therefore show a pattern: slower-loading mobile pages were associated with more visitors leaving.
Everything described so far assumes the page has actually loaded. For a large share of mobile visitors, that assumption can fail.
Google's 2016 mobile-speed research found that 53% of mobile site visits were abandoned when a page took more than three seconds to load. The figure came from Google Analytics data collected from a group of mobile websites that participated in Google's benchmarking programme. Google combined the data from those sites and removed information that could identify individual users before analysing how page speed related to visitor behaviour.
That distinction is important because Google was analysing websites that were already operating. It was not taking the same website, changing its loading time and watching exactly how many more people left. The numbers therefore show a pattern: slower-loading mobile pages were associated with more visitors leaving.
The figures should not be read as a formula that applies exactly to every website. A page taking five seconds instead of ten does not guarantee a specific percentage increase in visits, session length or engagement. Other things about the website and the visitor also affect what happens.
The data is also from 2016. So the 53% figure is not a statement about what exactly 53% of mobile visitors do today. It is evidence from that period showing that loading time was closely associated with whether mobile visitors stayed or left.
What matters for this article is the basic finding: page speed can affect whether someone gets far enough to experience the page at all.
The basic finding is still useful: the longer a mobile page takes to load, the more likely visitors are to leave before they experience it.
And that changes the order of events we have been discussing.
Before someone can form the kind of visual impression measured in Lindgaard's 50-millisecond study, the page has to appear on their screen. If it takes too long to get there, the visitor may leave before they ever see the design, read the headline or begin scanning the page.
In other words, the first thing your website has to do is actually show up.
If the page is still loading, the visitor has not reached that stage yet. They may simply leave.
So the first few seconds on a website are not only about what someone sees once the page appears. They can also determine whether the visitor gets to see the page at all.
What happens when someone lands on a website?
In this section
What the research does show is that several things can happen very quickly when someone lands on a webpage.
The studies we've looked at examine different parts of the experience of visiting a webpage. None of them followed one person from the moment a page began loading to the moment they decided to leave. Each study was asking a different question.
First, the page has to become usable. Google's mobile research found a strong relationship between slower loading and abandonment, including the finding that 53% of mobile visits were abandoned when pages took more than three seconds to load in its 2016 dataset.
If the visitor stays and the page appears, they can form a visual impression very quickly. Lindgaard and her colleagues found that ratings of visual appeal after 50 milliseconds were highly correlated with ratings after 500 milliseconds.
Then the visitor begins looking through the information. Nielsen Norman Group's eye-tracking research found a common F-shaped scanning pattern. Later research showed that when pages are organized with clear headings and subheadings, people can scan from one heading to another, reading the text underneath when they find a section that interests them. The researchers called this the layer-cake pattern.
While doing this, visitors can also begin forming an opinion about whether the website seems credible. In Fogg and colleagues' study, “design look” was the most frequently mentioned category in participants' explanations of their credibility judgments, appearing in 46.1% of comments.
These findings do not mean that every visitor goes through these steps in exactly this order. Someone might notice a visual problem before consciously looking for information. Someone else might leave because the page is irrelevant before credibility becomes a question. Another person might stay because they already trust the organisation and are simply looking for one specific piece of information.
What the research does show is that several things can happen very quickly when someone lands on a webpage.
They may be waiting for the page to load, taking in how it looks, scanning for something useful, deciding whether the site seems credible, or deciding that the page is not what they need.
There is no single number of seconds that determines when someone will stop paying attention. What happens in those first moments depends on the page, the visitor and what they came there to do.
And that gives us a much more useful question than “How do I capture someone's attention for eight seconds?”
What does my website communicate before someone has decided to stay and read it?
So, what should a website do well?
In this section
Visual design deserves attention because people can evaluate it extremely quickly.
None of these studies tells you exactly what your website should look like. They do not prescribe a particular layout, font, colour palette or page structure. They answer different questions, so the practical lessons need to stay within what the research actually shows.
The page has to load before anything else can happen. If a visitor leaves while the page is still loading, they never get to see the design, scan the information or decide whether the site is credible. Google's mobile research found that slower-loading pages were associated with higher abandonment.
Visual design deserves attention because people can evaluate it extremely quickly. Lindgaard's research found that people could form judgments about a webpage's visual appeal after just 50 milliseconds. Fogg's credibility research separately found that “design look” was the category mentioned most often when participants explained why they considered a website credible or not. Neither study shows that good design guarantees trust or keeps people on a page, but together they show that appearance is part of the experience and should not be treated as decoration.
Structure helps people find their way through the page. Eye-tracking research shows that visitors do not necessarily read webpages from top to bottom like a printed book. They scan for information that looks useful to them. Clear headings and subheadings give them places to stop, understand what a section is about and decide whether they want to read further.
And the content still has to deliver. None of this makes accuracy, evidence or useful information less important. A good-looking page can get someone's attention, but it cannot make poor information useful. If the page earns a second look, the content has to give the visitor a reason to stay.
At first, some of these findings can sound surprisingly superficial. Why should it matter so much whether a page looks organised or whether it takes a few extra seconds to load?
Because those things affect what the visitor is actually able to experience. A page that takes too long to appear may be abandoned before it is seen. A page that looks difficult to navigate may make it harder to find relevant information. A page with clear structure can make it easier to decide where to look next.
And none of this requires an eight-second timer.
There is no scientifically established point at which every person's attention switches off after eight seconds. What the research gives us instead are several different parts of the website experience that can be observed and tested: how quickly the page loads, how quickly people form visual impressions, how they scan the information, and what they notice when judging credibility.
Those are different questions, but they are much more useful than one number.
They give website owners specific things to look at, measure and improve.
Sources referenced in this piece: Lindgaard, Fernandes, Dudek & Brown, "Attention web designers: You have 50 milliseconds to make a good first impression!" Behaviour & Information Technology, 25(2), 2006. Nielsen Norman Group, "F-Shaped Pattern For Reading Web Content" (2006) and later eye-tracking research; "How Little Do Users Read?" (2008). Fogg et al., "How Do Users Evaluate the Credibility of Web Sites?" Stanford Persuasive Technology Lab, 2003. Google/DoubleClick, "The Need for Mobile Speed," 2016; Google research on 900,000 mobile ad landing pages across 126 countries, 2017. King's College London, "Are attention spans really collapsing?" 2022. BBC investigation into the Microsoft eight-second attention-span claim, 2017.
Want to put this into practice? TheStyle Vaultto take a FREE hands on course,How to Create a Landing Page that Converts, that walks through building a page around clear messaging, page structure, trust signals, and simple design principles, from a blank canvas to a page ready for real visitors.
