Kerry Yu | Product Designer
Artboard+1+%E2%80%93+1.jpg

Pratt Website Analytics

Redesign The Experience of Donate And Volunteer

Screen Shot 2019-12-16 at 8.18.03 PM.png
 

Pratt Website Digital ANALYTICS

OVERVIEW

Pratt.edu is the main website for Pratt Institute. We conducted a study to identify problems with data support in order to help with the future website update.

We gained greater insight into Pratt.edu’s users and best understand the actual use of the existing website, focusing on the Overall audience & session, Site performance, Search terms, Email campaigns, and Google Analytics Implementation in order to help with further evaluation of its audience and vast usability issues.


MY ROLE

I analyzed the data, created the Interactive Dashboard, identified findings, and designed mockups for recommendations.

This was a 6-week project in collaboration with a team of three analysts from Pratt’s School of Information. Insights from the data allowed us to identify recommendations that will support the future redesign of Pratt’s website. The dashboard is useful for staff at Pratt to keep track of traffic to the website in the future.


METHODOLOGY

Screen Shot 2019-12-16 at 4.01.30 PM.png

For the client meeting, we defined the site’s most important users as prospective students and their parents, followed by current students, faculty and staff, and alumni.The most important pages to be analyzed were the Homepage, Admissions, News, and site Search.

My notes from client meeting

My notes from client meeting

Google Analytics was used to analyze behavioral and quantitative data.

We analyzed data from the period August 1 - November 30, 2019. I created separate dashboards in Google Analytics to analyze the data and identified problems.

Google Data Studio was used to create an interactive dashboard.

It is a free tool that can attractively displays key data metrics from Google Analytics over a default period of the previous month. This will be a useful tool for Pratt administration to refer back to periodically in the future.

Key Metrics

  • The Data Studio dashboard includes visualizations of key Google Analytics metrics like sessions, pageviews, bounce rate, new vs. returning users, traffic sources, browsers and devices used, and user locations.

  • More granular information is displayed in tables, like email campaigns broken down by session and bounce rate; landing pages from emails; and number of pageviews by event tracking where on a page users clicked into their next path.


THE PROBLEM

The last design update of the website was 10 years ago in 2009. Since it has been used for years, the demand for updating the Pratt website is high.

Data support is needed to locate the issues for both desktop and mobile in order to help with the future website update.

Pratt’s administration is also concerned about its usability and mobile-friendliness, and there have been reports that some site content was hidden and users were having a hard time finding information.


GOALS

To gain greater insight into Pratt.edu's users and better understand the actual use of the existing website in order to help with further evaluation of its audience and vast usability issues, our findings focused on: 

  • Audience overview

  • Site performance

  • Search terms 

  • Email campaigns

  • Google Analytics Implementation


FINDINGS - AUDIENCE OVERVIEW

Who are the users?

Audience overview

Audience overview

  • 60.1% users access from desktop and 36.9% from mobile.

  • Bounce rate for mobile devices is pretty high (57.27%) compared with desktop (36.01%)

  • 78.8% new visitors, 21.2% returning visitors.

  • 82.2% of users are from the US, 45.93% are from New York.

  • Top 3 international visitors are from India, China, and Canada.

  • Bounce rate for France is really high (85.39%).

 

When do they visit the website?

Sessions by week

Sessions by week

Sessions

Sessions

  • Users visit the website more on weekdays than weekends.

 

What pages do they visit?

Landing Page

Landing Page

New Visitors

New Visitors

Returning Visitors

Returning Visitors

  • New visitors are interested in checking the faculty and staff

  • Accessing MyPratt accounts is one of the most important motivations of returning visitors.

 

 

RECOMMENDATION

  • Add MyPratt quick links to the top of the home page and turn them into a second header menu.

Make the quick links visible will help users quickly access to their user account.

Screen Shot 2019-12-16 at 7.04.03 PM.png
  • Promoting Pratt’s information during the weekday when user traffic is the highest.

Putting the information, such as hero image slide, annoucement, NEWS, and EVENTS on the popular time will receive more attention from visitors.

 

 

FINDINGS - SITE PERFORMANCE

Desktop and mobile page load times are very slow, resulting in a high bounce rate

We found that the load time for Pratt’s Home page and Admissions page are 8.54s and 7.32s respectively, and overall load time on the site is 6.02s on desktop and 5.76s on mobile.

Load time

Load time

Mobile page speed score

Mobile page speed score

Desktop page speed score

Desktop page speed score

 

Research from Google in 2017 indicates that 40% of users will leave a page that takes longer than 3 seconds to load (Google/SOASTA Research 2017).

About 40% of users visited the website using mobile devices from August 1 - November 30, but the bounce rate for mobile devices is 57% compared with 36% on desktop. This could be a direct result of the mobile site’s load time, among other mobile usability factors.

Visits by device

Visits by device

Screen Shot 2019-12-16 at 7.37.27 PM.png
 

 

RECOMMENDATION

Reduce page load time speeds by optimizing both desktop and mobile sites.

This could include reducing the size of the hero images across the site, simplifying the site architecture, and minimizing the number of referenced CSS and JavaScript files.

Screen+Shot+2019-12-16+at+7.46.27+PM.jpg
 

 

FINDINGS - SEARCH

4.59% of users used the “search” feature on the website.

The top search words are shown in the table. “Tuition,” “calendar,” and “starfish” are the three most searched terms by users. A possible reason could be that those three pieces of information are hard to find in the website.

Screen Shot 2019-12-16 at 7.56.56 PM.png
 

Users also search for similar information in different terms

Such as “tuition,” “Tuition,” “fee,” “fees,” and “tution” for tuition, “calendar,” and “academic calendar” for calendar.

Screen Shot 2019-12-16 at 8.00.00 PM.png
 
 
 
Screen Shot 2019-12-12 at 10.51.11 PM.png

“starfish” is the third most searched term but has 0 matched results.

 

 
 
 

RECOMMENDATION

  • Make the most searched information easier for users to find through the navigation 

  • Provide better search results for popular search terms

  • Provide some related text to the search field to help users locate the information under results 

 

 

FINDINGS - CAMPAIGNS

Email campaign has the most visits (45.88%) among all campaigns, but some landing pages from email campaigns have a very high bounce rate (92.01%).

A possible reason could be that the target audience is right but the landing page is not working well.

WeChat Screenshot_20191216222642.png
 

Email Landing Page “pratt.edu/admissions/request-information” has a high bounce rate (69.25%)

A possible reason could be that the design of the page is hard for users to acquire or navigate to the information they need

WeChat Screenshot_20191216222742.png
 

 

RECOMMENDATION

We recommend redesigning the landing pages that have a high bounce rate so that users will be able to easily navigate to the information they need.

screencapture-pratt-edu-admissions-request-information-2019-12-16-22_30_46.jpg

Current Page

All the links on the request information page are merged in sentences, it’s not an effective way for users to acquire information. Similar situations were found in other landing pages too.

Group%2B50.jpg

Recommendation

Reduced the size of the header image to make the important information above the fold;

Redesigned the links that were buried in the paragraph;

Redesign the button style to increase discoverability.

 

 
 
 

FINDINGS - GOOGLE ANALYTICS IMPELEMENTATION

The News page is receiving bot traffic and event tracking is only enabled on the Admissions page

Part of our initial research was into the News page, where we discovered that only 4.93% of all users visited the News page from August 1 - November 30, and almost 64% of New Users to the News page were from Paris. The period also has a 96% average bounce rate, raising some flags for unusual behavior on this page, likely from bots.

WeChat Screenshot_20191216232602.png
WeChat Screenshot_20191216232445.png
 

Event tracking is only enabled on the Admissions page

Event tracking is only occurring on select links and buttons on the Admis- sions page. This leads to irregular analysis of the actions us- ers are taking on the site, and omits some very important user behavior.

WeChat Screenshot_20191216233820.png
 

 

RECOMMENDATION

  • Follow established guidelines to reduce bot traffic

Include a robots.txt file to instruct or prevent bots on the siteAnalyze server logs and block bot IP addressesConsider throttling to limit how many times a client can hit the website before being denied

  • Ensure that the Google Analytics implementation is setup correctly 

Setup Search analytics to capture user searches 

  • Implement Goals and Funnels

Track specific user interactions and paths

  • Add Google Tag Manager to track events from all pages

Track tags & tracking pixels (Adwords, Facebook pixels), triggers (pageview, link click, etc), and variables


 
 

REFLECTION

Google Analytics can tell us where users click on a site and how often, what devices, browsers, and methods they come to the site by among other quantitative data. However, we cannot directly know user moti- vations or usability issues, so this data will be more powerful when combined with a usability study.

Google Analytics data is very detailed and granular for metrics like user acquisition and behavior, however, it tends to skew user demographics like age and gender towards young people. Google infers these traits based on a user’s other web behavior, but cannot definitely identify us- ers of a specific gender or age cohort. For this reason, we were unable to definitively track user behavior of the specific user groups identified in our initial meeting.

Our analysis of the News page also yielded some curious results that we interpreted as bot activity, thus limiting the insights that we were able to gather on that page. Lastly, because of the settings in Pratt’s Google Analytics implementation, we were only able to analyze data from August 2019 going forward. However, we were still able to gather valuable insights about user acquisition, site performance, and behavior on other pages of the site, detailed in the next section.

Final presentation to the clients

Final presentation to the clients