Dear Vernon City Council: about your website

There are a great number of people who don’t navigate a mobile device. Option A has a broader appeal.

The challenge with building a new website is that every stakeholder has an opinion. We all have ideas on what makes a website good.

Recently, the Vernon Morning Star reported on council’s review of the city’s new website: City uploads new design.

Naturally, each councillor had an opinion about the two design options.

I’d like to suggest a better way.

Look at the data

Those of us who design websites, web applications, and mobile apps have learned a truth about human behavior. There’s a big difference between what people say they want and how they actually act.

You have to be aware that people make confident but false predictions about their future behavior, especially when presented with a new and unfamiliar design. There’s a huge difference between imagining using something and actually using it. In addition, human preferences are rather unstable.

The only way to truly know what people want is to observe what they actually do. The great thing about the web is that we have tools to do this!

Google Analytics

Start with Google Analytics. It’s installed as a default by almost every web development firm. Viewing the source code on Vernon’s current website, I can see that it’s installed.

First, the city should see which pages are currently the most popular. You can do that by going to the Behavior > Site Content > All Pages menu. It will look something like this:

In this example, we can see the homepage is the most popular (naturally). Next is the Employment page, followed by Pay a Fine. In this case, the city may want to highlight those pages on the homepage so that people can find them faster. This exactly what Chatanooga did with their website:

Next, let’s address that statement made by Coun. Scott Anderson above. How many of Vernon’s residents are browsing the site on a mobile device?

To find out, go to Audience > Mobile > Overview in Google Analytics. Here’s an example:

Here, nearly 53% of website sessions happen on a mobile device (phone, iPod touch) or a tablet (iPad). I’m guessing that the City of Vernon’s stats are similar. PC sales are down, and smartphone usage continues to climb. For many people, their smartphone is their primary computing device.

How to test a design before you launch

There are also tools you can use to gather data on new website designs before you launch.

If the designs are in mockup stage (image files not yet turned into code) you can use a tool like Usability Hub to test how well each design helps users accomplish different tasks. Verify is another tool that does this well.

Another place to get data is to test how users interact with the current website. A tool like Peek User Testing will have anonymous users visit your site, and record their experience. This would allow the City to watch where people currently get stuck, or frustrated.

Test two variations

If there’s too much writing, you get discouraged.

Statements like this need data to back them up. This could be true, but we won’t know until we run a test.

One way to test these assumptions is by using an A/B test. These tests split a website’s traffic into two groups, and allow you to compare the results. For example, you might find that more text on the page improves engagement, and helps people achieve their desired result.

For example, when 37signals ran a split test for Highrise, they found that the homepage with more text resulted in more people signing up for the service:

A tool like Optimizely will help you run A/B tests.

Don’t make decisions based on assumptions

I hope the City of Vernon doesn’t make website decisions based on opinions, feelings, or assumptions.

My advice? Form a hypothesis and then test it. See how people are actually using the vernon.ca website, and make decisions based on that.

For more reading on this, check out Gizmodo’s article.

Justin Jackson is the author of Marketing for Developers, and writes at justinjackson.ca