*Listen to my recent “IMA Leader” interview on this topic

One year after Chapman University's new website launched, we reviewed the data, analyzed the KPI, and saw dramatic movement in the right direction.

• The average time spent on the homepage has gone up 15%.
• The bounce rate has decreased by 10%.
• Within the first six months following the launch, we saw the average time on the site increase by 5%.
• Within the first six months following the launch, our site’s overall bounce rate decreased by 4.5%.
• In February of 2013, only 1% of the traffic to our undergraduate application information page came from the homepage. Now 11% of the traffic to that page comes from the homepage.

Through this data, we can see that our new homepage is an improvement from the previous version. This sentiment is reflected in the positive feedback we have received from Chapman Family members. We also get praise from outside organizations in the form of honorable mentions in blog posts and awards.

I am often asked, “how do you decide what goes on the homepage?” Chapman’s web team bases decisions primarily off of data. Data is our favorite tool because it allows us to determine what to do without human emotion, ego, or rank getting in the way, and it produces the best results for our users. Naturally, there are exceptions to our ‘data-driven decision making’ rule, usually driven by politics or institutional aspirations.

Using Chapman University as an example, here are three things you can consider when deciding how to allocate that precious homepage real estate.

1) Data-Driven Decision Making

Data-driven decisions produce the best results for our users, and the digital marketing community considers it a best practice. At the 2014 IMPACT conference presented by the Internet Marketing Association, data-driven decision making was one of the primary topics of conversation at the CMO roundtable event, which included marketing directors from Microsoft, Adobe, MGM Grand, and many other major companies.

Data-driven decision making is just that: decisions that are based on unbiased facts and statistics. Using Chapman University as an example, let’s say we are trying to determine what should be listed in the “global navigation” (the bar across the top of Chapman’s primary website). At Chapman, we look at this from multiple vantage points:

• Focus groups and surveys
• Usability testing
  • “Rocket Surgery Made Easy: The Do-It-Yourself Guide to Finding and Fixing Usability Problems” by Steve Krug
• Tobii eye-tracking
• Google Analytics

Here, you’ll see the results of a heat map generated by clicks taken by survey respondents.

According to the survey data, most users utilize the Academics link, but few are hitting the Athletics link. When we look from another vantage point (Google Analytics), we see the same pattern.

Academics is probably one of the more popular pages on our site, so let’s dig a little deeper.

The data above reveals that after hitting our homepage, the “Degrees and Programs” page and “Academics” page are top ‘next step’ actions.

So what does all of this mean? It means that the Academics link has more than secured its spot in our global navigation. It also means when I’m prioritizing projects, the speed/presentation/overall user experience of those Academic pages will be examined, and if we need to allocate time to improve them, we will.

Unfortunately, this also means that at some logical point in time, I’ll have an awkward conversation with Athletics about removing them from the global navigation and better utilize that real estate with something else — something that the data tells us is more meaningful to our users.

The drop-down menu items in the global navigation also came to us after examining the data.  According to our Web Design and Technology Manager, Meghan Farrington:

The footer and the roles-based navigation are both influenced by data in the same way the global navigation is, but there are just a few exceptions. Any items that are in the global navigation, the roles-based navigation, or the website footer that are not supported by data are there because of legal requirements, University policy, or politics in general.

How we use data to influence story content on Chapman’s homepage:

Ross Loehner, Web Content Manager, put it best by saying:

Our homepage showcases beautiful HTML5 video prepared for us by Panther Productions, the video production unit of Chapman’s Strategic Marketing and Communications team. We call this section of the homepage the hero panel.  According to the mastermind designer of Chapman’s homepage, Meghan Farrington:

A pitch committee selects the stories featured in this hero panel. This group of marketers and public relations experts gather to ‘pitch’ stories that they feel communicate the University’s key marketing messages. Some of the content considered at the pitch meetings comes directly from our community VIA the inside.chapman.edu algorithm.

Chapman’s robust blog network (blogs.chapman.edu) allows anyone with valid Chapman login credentials to submit blog posts for consideration. Inside.chapman.edu’s algorithm then scores stories and the trending stories percolate to the top of the page. The Chapman web team uses those stories in the official ‘pitch’ meetings. What does all this mean? If you write for blogs.chapman.edu, and if the data tells us that you have an excellent story (via the algorithm) then your story could end up on the homepage.  Even if a story does not make it to the hero panel, it could be featured as a secondary story elsewhere on the homepage.

2) Aspirational Considerations

If we stick with our global navigation example, you’ll notice that there were a few other ‘low-scoring’ links according to the survey data. Arts, research, and support were also a little warm on the “what are you least likely to use” heat map. You’ll also notice that when users were asked “click on the resource that you have never used, but might be interested in exploring in the future” arts and research do well, while support is clicked less often.

Despite the data, support and research are both safe as they are both critical to the University’s aspirations and senior leadership’s strategic plan. This is a great example of something that we’ll elect to keep in the global navigation despite the data suggesting that it’s not heavily utilized there.

3) Politics

As long as we work with humans, there will be a political element for what goes on our websites. Generally speaking, we may have to include things that the data doesn’t support. Here are some examples of political justification:

• Let’s say one of Chapman’s schools or colleges is far more popular than the least popular “school or college.” We’re not likely to favor the more popular school or college just because it generates more hits.  That wouldn’t be fair to the other schools and colleges.
• In our roles-based navigation, the current student section is incredibly popular, but other audiences don’t use the website as much. We’re still not going to leave one of our key audiences out, even if the data says that we should.
• If University leadership strongly mandates that something needs to be on the website in a particular way, it shall be done.

In Conclusion

Your homepage is precious, and deciding what to include is important. Don't rush your decisions. Take the time necessary to consider the data, the institutional aspirations, and the organizational politics. A well-organized homepage is likely your most visible and critical marketing tool, and it should be given more thought and care than the billboard in the student union. Remember to ask yourself "who is this homepage for, and what do we want them to do once they get here?" Remember that as a digital marketer or a web designer, part of your job is to delight your users and community members. 

View the original blog post on Chapman's blog network.