The simplest form is the total number of bounces on a page divided by the total number of entrances of that same page. 

For this study, I took data from September 1, 2019 to November 25, 2019. This is the portion of the Fall 2019 semester till November 25th, the day I began this.

Bounce Rating

Here is a general ranking of bounce rates.

Out of the 18,426 pages that make up the CSP website, 922 pages have a recorded 100% bounce rate!

Cleaning Data

Here is the collection of all the raw program page data. Within the yellow column describing bounce rates, I highlighted pages with a 100% bounce in red and highlighted pages with an excellent rate, 26-40%, in green. From there I took out those rows and separated them into their own tables with respect their their bounce rank.

These are the tables I organized to display the rates I am focusing on that have 20+ pageviews. Anything under 20 isn’t reliable, obviously, because the values are so low. From here, I matched up 10 pages to study their comparison.

Calculating Correlation

A correlation coefficient is a numerical measure of some type of statistical relationship between two variables.

This is the formula I followed to solve for r, the Pearson Correlation, that measures a linear dependence between x and y. In this case, date and bounce rate.

Below is a display of the calculation steps I did in excel.

I first converted the dates into number formats and change the percentages into decimals. I then was able to calculate each mean, find the difference between each variable, perform multiplication, and total the sums.

Plugging in all the results into our equation, the answer displays r = -0.573.

On the spectrum from -1, being a negative correlation, to 0, no correlation at all, to 1, a positive correlation, we have a relationship that leans negative. Meaning, one variable increases while the other variable decreases. In this case, as the date of a page’s last update increases, it’s bounce rate decreases!

This graph shows the relationship of the date increasing while the bounce rate is decreasing. You can see in the graph that I included the y-forumla to showcase the rate of which the variables are changing together by the slope and y-intercept.

R^2, read “r squared”, is the variable used to represent a statistical measurement representing the variance of the dependent variable explained by the independent variable in a regression model.

Variance implies the average of the squared differences from the mean.

Regression Model is a set of statistical processes for estimating the relationships between variables.

Average Load Time

From September 1st to November 25th, there was an average of 590 CSP web page visitors per day. Who, on average, had to wait 5.57 seconds for a page to load. 

5.57 seconds. That is a lot considering the average user only waits 3 seconds before leaving site to view a competitor. Anything more than 3 seconds causes frustration and makes the visitor exit the page increasing the bounce rate.

But, who is waiting for us? Who are our users and what are they accessing us with?

User Demographics

User Age

This table displays our users by age. These two categories, by far, documented higher values of users compared to other age groups.

25 - 34: 25 - 34 year olds make up about 800 out of the 3,127 Fall 2019 CSP undergrads. This age of web users typically are seeking degree completion or to explore Masters level programs. Notice majority of this group is new compared to returning. That shows 77% of 25 - 34 year olds do not visit our website again.

18 - 24: 18 - 24 years olds are the traditional perspective, college-aged students. Compared to the other group, 18 - 24 year olds return to site more. These are most likely current students.

User Devices

Desktop: 55% Vistors; 54% Bounce Rate; 43,607 Chrome; 30,039 Safari

Mobile: 42% Visitors; 60% Bounce Rate; 25,838 Apple iPhone; 2,911 Apple iPad

The turnout between desktop vs. mobile is about equal. There is a 60% bounce rate for mobile users most likely because the screen space is limited, that is not as easy to organize flow and thought compared to the larger viewing offered from desktops. 

CSP seems like to favor Apple. Although we follow google platforms, the majority of people’s experiences are through Apple’s products and brower.

This is important to consider because users bounce and provide data from whatever device they are using at that moment. Desktop and mobile offer different perspectives and opportunities in experience.

Math Pages

BA - Mathematics

Last Update: 1/4/19

Bounce Rate: 40%

User Next Action: 40% Academics; 18% BS - Mathematics

Secondary Education Math Major

Last Update: 4/26/18

Bounce Rate: 100%

User Next Action: 22% Academics; 24% Admissions; 7% Dr. Krueger; 7% Dr. Jahn

In comparing the math pages, these programs both offer lots of content and areas to explore further.

Good: 

• Good use of white space, which help promote balance, user focus increases by 20% just because the ability to have headings and text stand out. The white space is fresh, modern, and open.

• Next Action is frequently to the admissions page which leads to 18% to the application page.

• Showcasing faculty allows users to learn and get to know their potential educators. Visitors advance in that direction, when available.

Bad:

• Admissions link leads to an error page.

• Negative white space - user is shown blank, empty space where useful information could be portrayed.

• Long - especially on mobile preview

Religion Pages

Theology Major

Last Update: 11/20/19

Bounce Rate: 37%

User Next Actions: 16% Academics; 16% Academics; 9% Koschmann; 8% Deressa; 8% M. Schuler; 8% R. Schuler

Religious Minor

Last Update: 2/26/18

Bounce Rate: 100%

User Next Actions: 36% Academics

Good:

• People are interested in the faculty! Especially in theology because religion is a close, personal connection.

Bad:

• Mobile version - Due to a phone’s limited real estate, the first sight the user expereincesis a picture of a window.

• Mobile version - Very long

• No featured faculty

• 34% Exit Rate

Social Pages

Political Science Minor

Last Update: 9/24/18

Bounce Rate: 33%

User Next Actions: 24% Academics; 15% Admissions

Criminal Justice Minor

Last Update: 2/8/18

Bounce Rate: 100%

User Next Actions: 38% Academics; 19% Admissions

Good:

• Clean white space, again visually appealing and results in a trustworthy connection with webpage and therefore organization.

• User is spending 2 and a half minutes on this page - shows interests before moving on. They must be reading word-for-word instead of skimming content.

• Admission to 18% Application page

• 2.32 second load time

Bad:

• As it seems users are interested in learning about the faculty, these pages offer no opportunity to do that.

• No related programs showcased.

Health Pages

Forensic Mental Health Certificate

Last Update: 2/8/18

Bounce Rate: 33%

User Previous Actions: 27% “forensic” Search

User Next Actions: 80% Admissions

Gerontology Minor

Last Update: 7/11/18

Bounce Rate: 100%

User Next Actions: 47% Academics

Good:

• Load time is 1.39 seconds but because there is unfortunately no content.

Bad:

• Not presenting anything to user and asking too much of them.

• No faculty, again.

Writing Pages

BA - English with Journalism Emphasis

Last Update: 2/8/19

Bounce Rate: 40%

User Next Actions: 38% Admissions; 21% Academics

Communication Studies Minor

Last Update: 7/12/18

Bounce Rate: 100%

User Next Actions: 50% Academics

Good:

• Best page performance: BA - English with Journalism Emphasis

  • Leads to admissions most of the time

  • Fast load time (1.74 sec)

  • Includes faculty and related programs

  • Happy hero image - girl laughing

Bad:

• Imperfections are unprofessional and interpreted as untrustworthy by the user and there interpret our institution as untrustworthy.

• Communication Studies Minor has a long load time (3.68 sec)

What can we do with this?

Encourage departments to include more content such as inserting faculty profile links and showcase related programs because CSP web page visitors do show interest in the educators and similar programs. It would also be important to frequently update pages to clean up and add new information.

The marketing department, and all departments, can use this to determine whether or not they are meeting their user’s expectations. This information can also be used to make informed desicions regarding content marketing, paid acquisition, SEO, and CRO or user experience.

Resources & References

• csp.edu (in reference with google analytics)

• CSP enrollment data

• https://www.searchenginejournal.com/what-is-bounce-rate-how-to-audit/240555/#close

• https://www.digishuffle.com/blogs/10-ways-to-analyze-high-bounce-rate-advanced-guide/

• https://www.gorocketfuel.com/the-rocket-blog/whats-the-average-bounce-rate-in-google-analytics/ 

• https://www.socialmediatoday.com/content/web-design-11-characteristics-user-friendly-website 

• https://blog.hubspot.com/marketing/improve-your-websites-user-experience 

• https://www.bluecorona.com/blog/how-fast-should-website-be 

• https://kickofflabs.com/blog/bounce-rate-14-ways-you-are-driving-people-away-from-your-landing-pages/ 

• https://www.oho.com/blog/2019-google-analytics-benchmarks-higher-education-websites 

• https://www.broadbandsearch.net/blog/mobile-desktop-internet-usage-statistics 

• https://www.dotnetcatch.com/2017/07/14/zeros-in-google-analytics-site-speed-metrics/ 

• http://www.thearda.com/Denoms/D_887_t.asp 

• https://iop.harvard.edu/survey/details/demographic-and-political-profile 

• https://ezgif.com/effects/ezgif-6-a7142680d0de.gif (tool used)

• https://screenshot.guru (tool used)

• Larsen, Richard J., and Morris L. Marx. An Introduction to Mathematical Statistics and Its Applications. 5th ed. Boston: Prentice Hall, 2012.

• https://www.mathsisfun.com/data/standard-deviation.html

• https://cxl.com/guides/bounce-rate/