Miratech participated in an international study involving a total of 510 users in 17 countries:

France, Australia, Belgium, Brazil, Chile, Denmark, Germany, Mexico, the Netherlands, Poland, Russia, South Africa, Spain, Sweden, Switzerland, Turkey, and the USA.

The Zurich-based Institut für Software-Ergonomie und Usability AG headed the study, with support from the IUTP (International Usability Testing Partnership), a network dedicated to user testing. Miratech is a founding member of the IUTP.

The study used eye tracking technology and questionnaires to measure how user behavior and opinions differ from one country to another.

The full report is available for download: IUTP_Marriott_Study_Report.pdf (PDF 5,6 Mo)

Methodology

Each user was placed in front of a computer screen displaying a mock-up of a booking page for Marriott Hotels.

They were then asked to perform tasks based on five separate scenarios. We measured how each user interacted with the page. Once the tasks were completed, we asked the users to answer a questionnaire.

The exact same testing protocol was used in all 17 countries.

Page used for testing

Behavior varies considerably from country to country

Country-related behavioral differences were much more significant than age- or gender-related differences.

We found major differences for each of the hypotheses we tested.

As an example, we found that the Germans only gazed at the navigation tabs after being on the page for 42 seconds, whereas the Poles gazed at them after only 15 seconds. That’s three times faster!

Germans (on the far right) gaze at the navigation tabs rather late and don’t pay much attention to them. The Poles (far left), on the other hand, gaze at them very early on

Here’s another example: few of the Dutch users were distracted by the main picture in the middle of the page, yet the Russians were quick to click on it.

Few Dutch users (far right) click on the picture in the middle of the page, whereas the Russians (far left) are drawn to it.

Overall, we found that users from different countries don’t always look for information in the same places. The heat maps below show the Belgian and Brazilian gaze patterns for the same scenario.

Belgians and Brazilians look in completely different places to perform the same task

Opinions also vary from country to country

The differences don’t end at behavior patterns: the opinions and perceptions of users also vary considerably from one country to another.

In this study, we compared Americans with non-Americans and found major differences.

As an example, 54.8% of Americans remembered seeing the phone number on the page, compared with only 33.2% of non-Americans.

As another example, 93.3% of Americans answered that they would book a hotel on the Marriott page, compared with 63.9% of non-Americans.

Opinions differ between Americans and non-Americans

Understanding the main functionalities

Despite the differences in behavior patterns and opinions, users from all countries had no trouble finding and understanding the main functionality of the site: the search tool.

Users from all countries find and understand the search tool.

This is something we often see in our international studies: regardless of what country they’re from, users have no trouble identifying website functionalities such as search tools, ‘add to cart’ buttons, filters and sorting functions, forms, etc.

Conclusion

People often ask us if translating a website is enough to roll it out worldwide. The answer is no.

Two parameters vary considerably from one country to another:

• the opinions and perceptions of users
• the areas on the page where users expect to find the information

On the plus side, users from all countries are able to understand and use the main website functionalities. The basic principles of usability remain largely the same from one country to another.

For more details about the study, feel free to download the full report: IUTP_Marriott_Study_Report.pdf (PDF 5,6 Mo)

Miratech won First Prize of the Be-UX Paris event for the creation of the international website Wonder.Legal, an inovative online tool for creating contracts, letters and legal documents in many countries.

The context

Miratech previously conducted two studies on the way users look at Google search results on a computer.

The first study analyzed how users behave on a Google search results page and talked about the famous Google Golden Triangle.

The second study focused on the visibility of Google ads.

Three years on, we have some new questions:

• About mobile phones: how do users look at Google search results on a smartphone screen?
• About computers: has rich content (images, videos, etc.) changed user behavior?

Our methodology

We conducted a user test with 18 participants: 10 French users and 8 Americans.

Each participant was set up in front of an individual computer screen. For the mobile phone tests, we displayed a smartphone emulator. For the computer tests, we simply displayed a Web browser. In both cases, we asked the users to perform some organic searches on Google. Using our eye tracking technology, we measured how participants looked at the pages as they browsed.

Instead of using their fingers as they would on a real smartphone, participants navigated around the emulator using a mouse. We were aware of the methodological bias this would cause and took this into account when calculating our results. We didn’t measure scroll and click times, for instance, because we know that these can be very different on a touch screen interface.

Visibility of Google ads on a smartphone

Google ads proportionally take up much more space on a smartphone screen than they do on a computer screen. So it’s only natural that Google ads get much more gaze time on a smartphone than they do on a computer. In some cases, users even spend more time looking at the ads than at the first organic result!

Because a Google ad takes up much more space on a smartphone screen than on a computer, it gets looked at a lot more.

Two Google ads take up almost a third of a smartphone screen.

In our previous study, we had calculated that users looked at a single ad at the top of the page on a computer screen for 0.98 seconds.

For an ad with an equivalent position on a smartphone, the average gaze time is 1.57 seconds. That’s 60% longer than on a computer!

A Google ad at the top of a page gets 60% more gaze time on a smartphone than on a computer!

When there are two ads at the top of the page, the first gets more gaze time than the first organic search result!

The Golden Triangle appears on smartphones too

The “Google Golden Triangle” shows that Google search results get less and less gaze time as a user moves down the page.

The heat map of users’ cumulated eye activity forms a triangle on the page. You can see this result in our previous user test.

We have found that this natural phenomenon occurs on smartphone search results pages too.

The tops of search results pages naturally get more gaze time. This is what we call the “Golden Triangle”

The “Google Golden Triangle” can be seen on smartphones in about the same proportions as on a computer.

The Golden Triangle now disappears on a computer

On a computer, the “Golden Triangle” now tends to disappear because rich content in search results changes user behavior on the page.

The presence of images, videos, thumbnails or lists shifts user focus in an unpredictable way.

The presence of a list of images can sometimes draw attention to the link located just beneath the images

In other cases, a list of images can create a “visual barrier” that stops users from exploring the rest of the page.

The added features are not seen on a computer

On a computer, Google has trouble drawing attention to its added features because the left and right columns get very little gaze time.

In the test, they were never even clicked on, despite the fact that some of the scenarios presented to the users strongly encouraged their use.

As is the case for all websites, the right and left columns were not looked at or used.

Conclusion

We used eye tracking to test how users viewed Google search results pages on mobile phones and computers.

To sum up:

• On mobile phones: Google ads (AdWords) got much more gaze time on a smartphone than on a computer.
• On mobile phones: the famous “Golden Triangle” appears for organic results. In other words, the first results get the most gaze time.
• On computers: rich content (images, videos, etc.) changes user behavior and eliminates the Golden Triangle.
• On computers: as is the case for all other websites, the right and left columns get little attention and are seldom used.

Surprisingly, the generational digital divide had never been precisely measured in terms of usability. When it comes to performing tasks on the Internet, just how wide is the gap between young people and older people?

Common sense already tells us that young people are more at home on the Web. We wanted to prove it.

So we studied the differences in performance between a group of over-55s and a group aged 18 to 25.

First, we conducted remote user testing on 257 respondents.

Participants performed the test on their own computer. They were asked to perform 8 tasks on 4 websites, including 2 that were very familiar and 2 that were unfamiliar.

We used a software program to collect the browsing data. The Miratech teams then compiled and analyzed the results.

To supplement the remote testing, we then conducted individual user testing using eye tracking technology.

Failure rate

The task failure rate (error or abandonment) for the youngest group was 18%; it was 34% for the oldest group.

The failure rate was 1.9 times greater for the over-55s than it was for the under-25s!

Time to successful completion

Compared with the youngest participants, the oldest took 40% more time to complete a task successfully.

Click here to view the embedded video.

On an unfamiliar site, an older person takes longer to accomplish a task

Click here to view the embedded video.

…on a familiar website as well

Perceived difficulty

1.9 times more participants found the task difficult or very difficult in the over-55s group than in the under-25s.

Influence of familiarity with the site

For both groups, familiarity with the site reduced the time taken to accomplish a task.

• The under-25s take 17% more time to accomplish a task on an unfamiliar site than on a familiar one
• The over-55s take 25% more time to accomplish a task on an unfamiliar site than on a familiar one

Familiarity with the site has more influence on the under-25s than the over-55s. It is clearly easier to learn a website at 20 years of age than it is at 60.

Conclusion

For the first time, Miratech has measured the differences in performance between under-25s and over-55s.

Older people:

• experience twice as many failures when browsing
• take 40% longer to accomplish a task
• have more difficulty learning to use a site

For the many researchers who follow our work, the raw data is available for download: Miratech_publication_ergonomie_difference_age.xlsx (xlsx file, 197 Ko).

Miratech led a study involving men and women in equal proportions, for a total of 210 participants from 7 countries:

• France – Miratech
• Denmark – Eyefact
• Netherlands – Concept7
• England – Bunnyfoot
• Brazil– Checon Pesquisa
• Spain – Université Pompeu Fabra
• USA – Customer Experience Labs

Some of these partners are members of the International Usability Testing Partnership – IUTP – a network dedicated to user testing and consumer research. The IUTP is supported by Tobii, the world’s leading developer of eye tracking equipment.

The testing was conducted individually. Each participant was simply asked to look at the photo of a young woman for 20 seconds, with no specific instructions.

While they looked at the photo, their eye movements were measured using eye tracking technology.

Miratech then compiled and analyzed the results. The detailed results are available in a white paper:
Men_are_pervs.pdf (PDF 5.9 Mb)

Original photo

Differences between men and women

In all countries there are some behavioral differences between the men and the women.

1 – The men spend 37% more time looking at the chest than the women do.

Click here to view the embedded video.

Men are all the same

2 – The women spend 27% more time looking at the ring than the men do.

Click here to view the embedded video.

Similar behaviors between countries and genders

Certain behaviors are identical, regardless of the country or gender:

1 – The face is the area most looked at

2 – The time spent looking at the pelvis, right hand and hair is identical for both the men and the women

For all countries and genders combined, the face is the most looked at

Differences between countries

There are major behavioral differences between the countries.

1 - Out of the groups of women, it is the French women who look at the chest the most! They look at the chest for 2.7 times longer than the English women, who come last in that particular ranking.

International ranking of the time the women spend looking at the chest

2 - The French and Danish men spend the most time looking at the chest – almost twice as long as the other countries!

International ranking of the time the men spend looking at the chest

3 – The English women are the most materialistic: they spend 149% more time looking at the ring than the English men do, and 2.2 times longer than the French women.

International ranking of the time the women spend looking at the ring

4 - France is the country with the least difference between the men and the women. The men and women from France are those who spend the most time looking at the chest. The French men only spend 9% more time looking at the chest than the French women do.

The benefits of eye tracking

Once they had looked at the picture, we asked the participants to fill out a questionnaire to tell us what they had seen.

When answering the questionnaire, it was the men who most mentioned the eyes and the ring, and yet it was the women who had looked at these the most.

This is a good illustration of the difference between what is self-reported (questionnaire) and what is measured (eye tracking). With eye tracking, it is possible to measure behaviors that are impossible to reveal using conventional research methods.

Conclusion

Miratech and its international partners conducted the first international eye tracking study to measure how men and women look at the photo of a provocative-looking young woman.

Conclusion : men are pervs and women are gold diggers.

The detailed results are available in a white paper: Men_are_pervs.pdf (PDF 5.9 Mb)

News: Miratech won First Prize of the Be-UX Paris event for the creation of Wonder.Legal, a website for generating contracts and letters with a simple form.

Our study measures web browsing on different devices, using different interfaces.

We tested a representative sample of 20 users in our user testing labs. All of them were familiar with the iPad. We had them browse through five well-known French websites that all have their own iPad app.

We tested how each participant interacted with:

• the website on a computer
• the website on an iPad
• the iPad-optimized app

To reduce the risk of bias caused by familiarity, we rotated the order in which the interfaces were used. First, we let the participants browse freely around each of the sites/apps, and then we assigned them specific tasks. Our team compiled and analyzed the results using technologies such as eye-tracking.

Eye activity on a website and its corresponding app

Web browsing is still more efficient on a computer

It is much easier to perform specific tasks on a computer.

Participants looked at fewer pages on a computer and they made far fewer errors to achieve the same result, especially when searching for and buying a product.

What’s more, users are wary of m-commerce sites and still reluctant to use the iPad for online shopping. For example, many users think that the products available on the website are not all available on the iPad.

Apps are more fun

When users were free to browse as they pleased and didn’t have a set task to perform, they were faster and viewed more pages on an iPad-optimized app. This is because the iPad interface is more fluid and the swipe feature encourages users to flick through content.

These findings are consistent with those of earlier studies: it’s easy for users to flip through the pages of an iPad looking for content.

Click here to view the embedded video.

It’s fun to swipe!

It’s hard to browse the web with an iPad

Standard websites are not easy to use on a tablet. We noticed that many of our participants experienced frustration due to a high number of tap errors. As a result, tasks took much longer and there was an obvious risk the user would just give up.

Click here to view the embedded video.

It’s hard to browse the web with an iPad

What’s more, the users hardly used the zoom function, which actually would have made it easier to browse with the iPad. We’ll take a closer look at this result in our next study.

Finally, the number of tap errors was just as high on the apps. But the fun of swiping and the overall ease-of-use largely made up for the resulting frustration.

Conclusion

Miratech revealed the differences between using an iPad and using a computer:

• The computer is still the most efficient technology for performing specific tasks such as online shopping.
• The iPad, on the other hand, is great for discovering content, as long as you use its specially-designed apps.

The Miratech team wanted to know how reading patterns differ between an iPad and a printed newspaper.
Our conclusion : people read articles in a printed newspaper and only skim through them on an iPad.

We conducted a study to analyze the differences between reading on an iPad and reading a printed newspaper.

For the purposes of the study, we used the iPad and printed versions of the 20 Minutes newspaper. We used special eye tracking technology to monitor the eye activity of participants, and we measured how well they remembered the articles and ads.

In our previous article, we revealed that reading on an iPad is more superficial than reading in a printed newspaper.

We continued to analyze the results by focusing on the average overall reading time, the number of articles read, and the number of ads looked at.

Eye activity on the front pages of the iPad and paper versions

More articles are read on the iPad

The most significant difference was in the number of articles read.

Readers read twice as many articles on an iPad than in a printed newspaper (35% of articles read on the iPad, compared with 18% on paper).

This is because reading on an iPad is more superficial: people skim through the information (see previous results). They concentrate less and less information is retained.

No difference in ad visibility

There is no marked difference between ad visibility on an iPad and in a printed newspaper. 26% of ads are looked at on the iPad, compared with 21% in the printed paper.

Ads get just as much attention on paper…

…than on an iPad.

The average time spent reading is greater on an iPad

We noticed that readers spend more time on the iPad, taking 2’30″ more to go through the whole newspaper!

The readers noticed that the brightness of the iPad screen made text and images stand out. As a result, the iPad content looks more appealing and readers want to spend more time on it. On the downside, they also noticed that reading on the iPad put more strain on their eyes.

Conclusion

Miratech revealed the differences between reading on an iPad and reading a printed newspaper:

• On an iPad, reading is efficient but superficial. Readers look through more articles, but retain less information.
• People assimilate and remember information more when they read it in a printed newspaper.

In short : we skim through articles on an iPad and we actually read them on paper.

In our next newsletter we will present the results of our study on how computer mice and touch screens are used.

Important: Miratech won First Prize of the Be-UX Paris event for Wonder.Legal, a solution for online legal documents.

At Miratech, we are interested in the difference between reading on an iPad and a printed newspaper.
Readers remember an article read in a newspaper better than an article read on an iPad. Readers are more likely to skim over articles on an iPad than in a newspaper.

Our study analyzes the differences between iPad and newspaper reading patterns. We asked a representative sample of participants to read similar information from a printed newspaper, and from its iPad version. Half the participants were asked to read the iPad first and the other half were asked to read the newspaper first.

Participants in the study were already iPad users, so we avoided any bias associated with learning how to use the iPad (discovering the functions, playing with the zoom, etc.).

They were allowed to freely manipulate both the paper and iPad versions, and we monitored their behavior as they read. We used eye tracking technology to follow their gaze path. After reading, we measured how well they remembered the articles and ads.

Click here to view the embedded video.

Click here to view the embedded video.

No significant difference in reading time

The average time taken to read an article on each medium is very similar.

A user takes an average of 1 minute 11 seconds to read an article on paper, compared with 1 minute 13 seconds on an iPad. Thus the length of time for reading an article on paper or iPad is very close.

Identical gaze patterns on an iPad and a printed newspaper

For concentration and retention, paper is the clear winner

A more detailed analysis shows that the eyes linger longer on the paper version (275 ms on paper versus 231 ms on the iPad).

This means that people concentrate more when reading an actual newspaper.

This increased concentration results in a better retention of printed articles. After reading, only 70% of participants recall an article read on an iPad, compared with 90% for paper!

Conclusion

Miratech demonstrated the difference between reading an article in a newspaper and on an iPad.

The study showed that:

• The type of medium doesn’t influence reading time when the text is short (like an article).
• It is easier to assimilate and retain information read in a newspaper than on an iPad.

Our next newsletter will present some more surprising results, this time on ad visibility.

1st criteria: test the characters that appear in the ads

Define your characters according to two criteria:

• gender
• sensuality

2nd criteria: test ad layout

Define ad layout according to four criteria:

• the presence of text
• the presence of a face
• visual integration: the image colors blend in with the rest of the website (integrated) or clash completely (not integrated)
• density: the background of the ad is busy with lots of detail (dense) or, on the contrary, it is quite blurred and simple (not dense)

Conclusion

We at Miratech have made a tool available online so you can compare the impact of different criteria on ad visibility.

This tool compiles the results of an eye tracking study conducted by our teams. For a more detailed explanation on how this study was performed, download our white paper (in French only, for now.)

Our study focused on how the content of an ad influences its visibility. We tested a sample of users in our eye tracking labs. The users were asked to navigate freely on 8 blogs that had been redesigned for the purposes of the study.

For the first time, the detailed results of the study are now available in a white paper:
eye-tracking-etude-publicites.pdf (PDF 3,5 Mb – in French)

The study is divided into two parts. The first part covers graphical elements. The second examines criteria relating to people featured in ads.

Using graphics effectively

1 - The “Text” criteria has the greatest impact on ad visibility. If you add text, you can expect the visibility to almost double!

Here is an example of a heat map to illustrate our point:

2 – After text, the presence of a face is the next most influential factor on the visibility of your ad:

The example above is obvious: ads that feature a face attract a lot more attention!

3 - The last criteria we studied, that was shown to have a strong impact on ad visibility, is visual integration or how well the ad blends in with the web page

Here we can see clearly that the complementary ad block (same tones) is more eye-grabbing than the ad on the left with its loud colors:

Winning graphical elements

The presence of both text and a face are the two most influential factors affecting ad visibility.

Here is a pie chart to illustrate the relative influence of each element:

The winning combination for graphics

So the winning combination to achieve optimum visibility is…

…text + face + good page integration

Example of an ad that combines the winning criteria

Effectiveness of criteria relating to people featured in ads

1 - The results are not surprising: a sensual-looking person attracts more looks.

The ad on the left showing a man with a bare chest got more looks from participants.

2 - Surprisingly, the users (a mix of both men and women) tended to look more at men than at women:

In this example, the man was looked at much more than the woman:

Winning criteria when featuring people

Of the two criteria, sensuality has a far greater impact than gender:

Pie chart showing the relative importance of each criterion

The winning combination when featuring people

The winning combination to achieve optimum visibility is…

…a sensual man

Example of an ad that combines the winning criteria

Conclusion

Miratech has exposed the criteria that have the most significant impact on ad visibility.

Our study revealed the following key points:

• Users looked at men more than women
• The most effective ad is one that features text and a sensual man, and blends in well with the web page.

Detailed results are available in the white paper: eye-tracking-etude-publicites.pdf (PDF 3,5 Mb – in French)

News: Miratech won First Prize of the Be-UX Paris event for Wonder.Legal, a solution for creating contracts and legal documents with a simple form.

We used eye tracking technology to study the behavior of a representative sample of French users.

Participants were asked to navigate freely on sites such as lemonde.fr, jeuxvideo.com, allocine.fr and lequipe.fr, which are rich in content and advertising. A total of 1,746 ads were studied on 411 visited pages.

We tracked the dwell time and the number of gazes for each ad, as well as the time spent on each page during navigation.

The results were surprising!

We found that users spent 6.58% of their time looking at ads, which was much more than we expected.

In other words, for every 10 minutes of navigation, a user spends an average of 40 seconds on ads!

This is even more surprising when you consider the commonly held belief that web advertising on the internet doesn’t work well. Our study proves that even when the ads aren’t clicked on, they still register!

Content influences effectiveness

Content type has a strong influence on how well an ad performs: the simpler the ad is, the more it is looked at.

For example, plain text ads (typical Google ads) are viewed twice as much as ads that contain an image:

Average viewing time for plain text ads compared with ads that contain an image

Example from the l’Equipe website:

On one same page of l’Equipe website, the plain text ad is looked at and the more complex ad is completely ignored

We also found that, static ads are twice as effective as animated ads:

Average viewing time for static and animated ads

This result illustrates a paradox: the more an advertiser tries to increase the visibility of an ad, the more the user ignores it!

Here is a classic example from the lemonde.fr website:

Click here to view the embedded video.

The plain text ad works much better than the animated one!

Different formats have different degrees of effectiveness

For the purposes of the study, we grouped the different types of ads into three very general categories:

• Wallpaper ads (replaces the background of a website or page)
• Horizontal banner ads (rectangular blocks that generally take up a significant part of the page)
• Square ad boxes (a broad category that includes all square format ads)

This page from allocine.fr contains all three advertising formats

The best-performing advertising format is the wallpaper ad. Horizontal banner ads come second, followed by square ad boxes:

Time spent looking at advertising, in relation to time spent on the page.

Example from allocine.fr:

On this allocine.fr page, the wallpaper ad is the most effective.

Position influences effectiveness

The effectiveness of an ad varies according to its position on the page. Unsurprisingly, the higher the ad is placed on the page, the more it is looked at.

Time spent looking at square ad boxes, in relation to their position on the page.

Interesting finding: there is no significant difference in effectiveness between an ad positioned on the right of a page and one placed on the left.

Conclusion

Miratech conducted the first statistical study on ad visibility.

The study revealed the following major points:

• Web users spend 6.58% of their time looking at ads. This is a considerable amount of time.
• Simple ads perform better than flashy ads: plain text ads are more effective than image-based ads, and static ads are more effective than animated ads.
• Wallpaper ads are the most effective format.
• Ads placed high on the page attract more attention than ads lower down on the page. Positioning to the right or left doesn’t influence viewing.

Our next study will examine how content such as text, images, animation and graphics influences ad effectiveness. You’ll finally have all the information you need to make your web ads profitable.