A website redesign can be a very challenging job for a communications lead. You’re responsible for the most important, public, and comprehensive telling of your organization’s story. You also walk the tightrope of managing expectations internally. Staff, leadership, and board members are all scrutinizing how their programs and initiatives are described. You’re also probably doing this on top of another job. Many communications leaders have never done a big website redesign before and don’t have a huge team to support them. 

In a previous Insights post we wrote about what we call the “ingrown website,” the condition where your site’s information architecture starts to mirror your org chart. Department names become top-level labels. The “About Us” grows more pages than the programs do. This often results from a hundred reasonable small decisions made for the sake of expediency and internal peace because you can’t demonstrate why things should be otherwise. 

We’re here to encourage you to invest some time and energy into engaging with internal colleagues to gather what will become excellent information to create a durable, effective website information architecture and user experience for your audiences. 

Here are the same five steps we take with clients to get to a site that really works for the most important people — your audience.

1. Listen with an open mindset

As the person leading this effort, it can be difficult to truly listen openly to hear answers or feedback you do not want. None of the next steps will work if you’re only half-listening while holding firm to your own ideas about the answers. It takes confidence and courage to let go and allow a creative process to unfold, especially if there are internal tensions or conflicts with people you need to bring into this process. Take it from us – it’s much better to let go and allow others in than to end up with a product and experience that fails. 

2. Listen to the evidence you already have

You have information about your visitors already. Google Search Console shows the literal words people type to find you. If you have a search feature on your site already, you can look at the search logs to show where people had a hard time finding the information they were looking for. If “how to donate” is a top query on your own site, your structure buried that option. It’s true that some people will default to search instead of navigating, but we’ve found that great navigation is more vital than search. One very important reason is that it proves to your visitors that you really understand them. 

Look at your landing page analytics and see where people go next. That’ll show you how people are using your site. Are people using Google search to find specific things and using deeper links? That data can guide you to potential information that should be in your navigation. 

So use the tools you already have to understand your visitors’ vocabulary, in their own words. 

3. Decide who the site is really for

A navigation that treats every audience equally serves none of them well. Do some initial work to talk to your staff and emphasize the importance of your different audiences, using your organization’s strategic goals as your north star. You might have multiple priority audiences. That’s fine. A website can accommodate many different pathways. They just can’t highlight too many or they become ineffective.  

Do this: Build a one-page audience map. Put your audiences down the first column, and then add a column for how they help you advance your strategic goals. And more columns for what actions you want them to take, what each is trying to do, what motivates them, and what worries them. 

This exercise gives you a stable basis for the hundred small decisions still ahead. Once this document exists, you have a map for what the navigation needs to do and, in turn, redirects questions of preference into questions of fit. 

Talk to your staff and emphasize the importance of your different audiences, using your organization’s strategic goals as your north star. 

4. Ask people what they were trying to do

Once you’ve done that internal exercise, do some pressure testing. You have to ask people. This is the step in the process that takes some time, but if you give yourself time, you can spread it out. 

You don’t need dozens of meetings. Conduct about five 30-minute conversations per priority audience. You’re not asking them what they think of the website. Ask them the more fundamental questions that help you test your (informed) assumptions you made in your audience map. 

Ask them instead to walk you through what they want to do and what their experience was like. Try not to talk too much, and they’ll give you a ton of great information. Write down the verbs they use. The verbs can become your navigation labels.

Then go back and revise your audience map. Where the interviews contradict what you assumed, the interviews win or require more investigation. Every cell they overturn is a mistake you didn’t build into a sitemap.

This gives you the actual shape of the work people are doing, which almost never matches your org chart. When a grantseeker is trying to figure out whether they’re eligible for something, they might touch a program page, an FAQ, and an application portal, and those might be three different functions at your organization. The visitor experiences it (or should) as one continuous task. 

This is the information that will convince your colleagues because hearing someone struggle with a page you built is more persuasive than any argument you could make on your own.

5. Test before you build

You can also do some card sorting. Take all of your content and ask people who aren’t part of your organization to sort them into natural buckets. That gives you another data point and can suggest ways you haven’t even thought of before. 

Another activity that has proven useful is to take the different candidate structures you have and put them on post-it notes. Ask people “Where would you find ______?” That will tell you so much about how much work your labels are doing. 

All of these exercises permit you to be wrong early, while it’s still cheap to pivot. And it gives you evidence to show internal people who want to maintain the status quo. It's a lot more comfortable and useful to learn from fifteen strangers in week three than to learn from your core audiences after launch that a label doesn’t work.

It's a lot more useful to learn from fifteen strangers in week three than to learn from your core audiences after launch that a label doesn’t work.

Replace Guessing With Listening

This up-front process mostly replaces guessing with listening. What comes out the other side is a structure you can explain, improve, and stand behind, built on something sturdier than anyone's best instinct, including yours. That’s a better website, and it's also a much lighter thing to carry.