Puzzling web habits across the globe - part 2

As any avid Olympics fan can tell you, time is of the essence. In the case of a "fingertip win" like Michael Phelps', even a hundredth of a second can make a difference between a gold and a silver medal.

The same principle applies to those of us with higher BMIs and underused gym memberships. Do you know that every single second you spend staring at the rotating hourglass on your screen has a direct effect on the way you interact with the site? It's a fact.

Part 1 of our Clicktale Web Browsing Habits Report showed that:

  • For every additional second that it takes to load a page, an extra 6 seconds is spent browsing that page.
  • The Dutch and Israelis outsurf all other countries, with extremely fast Internet and in-page browsing speeds. The Chinese, with the slowest load times, spend three times as long on a typical page.

Okay, so page loading is not exactly up for consideration as an Olympic sport. However, while looking at the data we put together, we encounter a few fascinating record breakers. These are summarized at the end of this section.

Time In-Page vs. Page Loading Speed

Figure 1: Cross-country comparison of the average time visitors spent browsing a single web page vs. the average time it takes a web page to load.

One striking anomaly of the study was that Indians, whose Internet speed of 3.5 seconds is slower than average, browse an unprecedented five seconds more quickly than the stats predict. These five seconds matter.

Five seconds can mean the difference between a visitor completing a conversion on your site or being distracted by the US Gymnastics team. Hence, the fact that surfers in India take an average of 24 seconds to browse a site as opposed to the expected 29 is a phenomenon that begs an explanation.

Somehow, we don't think a surprise drug test will give us the answer we're looking for, and so we set off on Part 2 of our research where we try to understand:

Why do most countries surf about as fast as they're expected to while India is in a league of its own?

The first step to answering this question requires the following data: On average, how many pages per site are viewed by people in different countries?

As you can see from Figure 2 below, page views range from an average 1.48 to 2.69 per country, with China viewing the fewest number of pages while Israelis view the most per visit to a single site.

Visitor Pageviews per Country

Figure 2: The average number of pages per website viewed by visitors from across the globe.

Now that we have this info, we can ask: Is there a correlation between the number of pages viewed by visitors and the page loading speed?

The short answer is: Yes. Pageviews enhance performance.

From Figure 3 below, you can see that a linear relationship exists between Internet speed and the number of page views, as most countries are positioned close to the regression line. Only a few countries, including India and Israel, stand apart.

Pageviews per Visitor vs. Page Loading Speed

Figure 3: Cross-country comparison between page loading speed and the number of pages viewed.

If you look at this graph in Olympic terms (and if pageviews = points), most countries performed exactly as they were expected to. A few favorites didn't place as well as we thought they would (France, Netherlands) while some of the underdogs took home winning medals (India, Israel).

How can we explain this finding?

We flexed our brain muscles and came up with these ideas:

  • Perhaps Israelis and Indians are more impatient browsers than others, or maybe they browse more during office hours when time is more limited.
  • Maybe it's the nature of the pages they're visiting; they may be laid out such that visitors can get their information and move on.
  • Why do Indians take gold in the Team Browsing event? Possibly because they multi-tab browse, which means that they browse other pages while a current page is waiting to load. This makes sense since Indians have a higher than average page loading speed.

It would be interesting to compare these figures to the number of computers per capita or per household. Maybe there is a line of impatient family members standing over each other's shoulders and waiting their turn to surf.

How does Internet speed affect the total time visitors spend inside a site?

To answer this question, we measured how much time visitors from different countries spend browsing inside websites. As you can see from Figure 4, the time visitors spent in a site varied from a low of 24.8 seconds (Dutch) to a high of 106.7 seconds (Indian). In other words, on average Indians spend over 4 times as long inside websites than the Dutch.

Visitor Time in Site per Country

Figure 4: Average time spent inside a website across the globe.

Can page loading speed explain the differences in time spent inside a website?

Figure 5 below shows that there is a strong linear relationship between page loading speed and time spent inside a website. In fact, if we apply our favorite exercise - statistics- we can see that for every additional second it takes to load a page, a visitor spends an extra 10 seconds inside the website "an outstanding 10-fold effect!"

Funny, we never before thought of browsing as an endurance sport.

The slope of the line in Figure 5 (below) shows that small changes in the loading speed have a big impact. However, it's hard to see each country's position on this graph so let's zoom in on the x-axis, as shown in Figure 6 below. (Our version of the classic slow motion instant replay.) Notice that Indians spend about 33 more seconds than expected browsing websites while the Dutch spend an average of 23 seconds less than expected.

Visitor Time in Site vs. Page Loading Speed

Figure 5: Cross-country comparison of the average time visitors spent inside a website vs. the average time it takes to load a page.

Visitor Time in Site vs. Page Loading Speed

Figure 6: Cross-country comparison of the average time visitors spent inside a website vs. the average time it takes to load a page.

Why would load speed have such a huge effect on the time that visitors spend in a site?

Our best guess, as we said in Part 1, is that perhaps surfers who experience slower Internet speed feel that they've made such a big investment that they prefer to stay on the site longer to get a bigger return.

Part 2 Summary of Results

  • Page loading speed has a 10-fold effect on time spent in the site. That means that for every additional second it takes to load a page, a visitor will spend 10 more seconds browsing the website on average.
  • Chinese view the fewest pages, while Israelis view the most.
  • Indians spend the most time browsing a typical website and go through more pages much faster than would be expected.

Recommendations (or What Does All This Mean to the Armchair Athlete?)

  • Since Internet speed is becoming faster, visitors tend to spend less time per page but browse more pages. Therefore, websites need to become more effective at grabbing visitor attention quickly, with the goal of converting visitors into customers as fast as possible.
  • Web designers and developers should apply K.I.S.S. (Keep it Short and Simple) and focus on simple messages that don't require much time or effort to analyze.
  • Use tools such as Clicktale to understand visitors' behavior inside the web page and continuously monitor changes that improve usability and increase conversion rates.

You be the Judge

In this blog, we've provided our own explanations to our research findings.

  • Do you agree or disagree with our explanations?
  • Are you surprised at the outcome of the research or the standings of each of the countries?

We'd love to hear your thoughts about our findings and recommendations. Simply record your opinions in our comments section below.

Talk to us to explore how customer experience analytics can improve your business