Usability Testing Mastheads: A Case Study with General Motors

This post was guest written by John Ribbing, Sr. User Experience Architect at McCann with over 20 years experience in Digital Content and User Experience.

Some call them rotating carousels, some call them sliders, some call them mastheads. They’re the graphics and images that take up a large chunk of prime real estate at the top of a website, polarizing in that they are loved by site owners but hated by the UX community

The argument for them is that they allow you to have multiple messages on the home page. The argument against is that they’re bad as they’re both distracting and ineffective, often causing them to be tuned out by users.

One of my clients at McCann is General Motors; their corporate site has a rotating masthead with 4 slots for messages. Despite occupying prime real-estate on the homepage, the masthead was receiving very few clicks and we wanted to learn why it was so unpopular with users.

Working with UsabilityHub, we conducted a number of click tests on the home page. Using a set of variation tests (a Pro Subscriber feature) to ensure no tester would take more than one of the five tests. Testers came from all around the world with the majority in the U.S. and U.K.

The first test was a benchmark using the current General Motors home page and asked testers to click where they would expect to find the most helpful information.

The majority of the testers (50%) selected somewhere in the navigation menu. Others (24%) selected the 3 promotional tiles below the masthead and the smallest group (16%) selected the masthead. This was reassuring as it mirrored what the actual analytics from the website were saying.

Why was the current website layout testing poorly with users?

In order to test where people expected to click on a homepage for helpful information we ran a test of a basic wireframe. It was as generic as possible — no branding, no specific navigation terms, no images.

We asked users where they would expect to find the most helpful content… with a couple of twists.

Since we knew most would go to the navigation (and because we wanted to test interaction with the promotional areas of the site) we highlighted 3 areas and asked testers to make their selection in one of those areas.

Testers selected the masthead area most frequently (61%), the promotional tiles under the masthead and the navigation promotional tile were equally selected (15%).

Without seeing any content, testers expected the most helpful information to be in the masthead area — basic Visual Hierarchy based on size and placement.

Our next test used an actual screen grab of the General Motors website which had a message about the vehicles in the General Motors portfolio.

The results were remarkably similar to the wireframe test. Users selected the masthead area most frequently (60%) followed by the promotional tiles (32%) and the navigation promotional tile (6%).

And then things changed.

We ran a test showing a masthead with a financial results message. Even though it was a positive message (3rd Quarter Net Income of $2.8B) users were not so interested. Most of the clicks (54%) were on one of the promotional tiles. Fewer were in the masthead (34%) and even fewer were in the navigation promotional tile (2%).

The last test featured a unique masthead image and a story about how GM has been recycling water bottles and turning them into engine covers and coats and other things. It’s a great story about how GM is striving to be a leader in Sustainability and Recycling.

Once again we saw the majority of clicks in the promotional tile area (46%) followed by the masthead (34%) and then the navigation area (6%).

For all both of these tests, users looked at the biggest, most prominent area of the site and said, “Eh. No thanks.”

So what did we take away?

Content and context matter.

When there was no visible content people expected the mastheads to be most helpful. When there was automotive content users expected the mastheads to be most helpful.

But when the message was about corporate profits and sustainability, interest waned and users thought information in the promo tiles would be most helpful.

Not that corporate profits and sustainability aren’t important factors for a corporation to promote — they are. But they are not the primary message one would expect to see on a corporate site for an automaker.

Years ago I saw the same thing testing a grocery chain’s emails. Users gravitated more toward the food and less toward the non-food content. It made sense then too, most users are coming to a website for a specific reason.

In the real world it works like this — every holiday my local grocery store puts up a big display in the entrance to the store. It’s the first thing you see when you walk in the door and it takes up a lot of space. For Thanksgiving they had all sorts of pumpkin spice products, decorations, drinks, napkins and accessories. But what they didn’t have was turkeys…or potatoes…or stuffing…or green beans.

They had a lot of nice-to-have items but they didn’t have the things I was most interested in buying at the time.

As I was preparing this article I took the time to visit the store and watch people as they walked in the store. A few people grabbed an item or two, but only a small percentage. Maybe they didn’t need the items. Maybe they didn’t like the items. Maybe they were on autopilot and breezed right by them to go look in other areas of the store. (Shopper Science Pioneer Paco Underhill calls this area of a retail store the Decompression Zone in his book “Why We Buy”)

The same thing seems to be happening with these mastheads.

So in mastheads as in grocery store displays, in order to get the most visitors to interact with them you need to display the things that most visitors most want to interact with.