Answering voters’ questions on websites

When links, labels, and headings on websites match how voters talk and think, they find what they’re looking for, and they are less likely to call asking questions.

My team spent the better part of autumn 2012 conducting a research project on county election websites. We were wondering what counties were putting on their websites, whether voters went to the websites, and if voters did go to the sites did they find what they were looking for.

We started just by documenting what was on the main page for elections for 147 county websites. We collected the text of about 8,000 links, headings, and graphics. Then we conducted 41 usability test sessions in which we asked voters what their questions about the election were and then had them try to look up the answers on their county or parish or borough) election website.

What we learned surprised us. Most of us inside elections think of an election as a process. Things happen in a predictable order, and for the voter, the common wisdom is that the process starts with registering and ends with voting. So it makes sense that most election websites reflect this process.

But voters don’t start with registering. They start with the ballot. Here’s the basic thought process, in this order:

  1. What’s on the ballot? [I know about the big thing, but what else should I know about?]
  2. What are my options for voting? Do I have to vote in person on Election Day?
  3. How do I apply for an absentee ballot? What are the deadlines for that? Are there any other options?
  4. Who is in office now?
  5. Do I need voter ID? What can I use to get it? Where do I have to go?
  6. Am I registered? How do I register? What if I’ve missed the deadline?

Yep, almost the complete opposite of what you might think. So, if your county can answer these questions on the election website, chance are, fewer calls will come in. And, if you can answer the first question without using the term “sample ballot” you get extra points. Even when voters in the study came across links for the sample ballot, they ignored them, expecting that it was show and wouldn’t tell them what was on their ballot.

By the way, the people in our study had been to their county websites and were much more likely to go to the county election website than to visit the Secretary of State’s website about elections. No pressure.



This research was made possible with support from the John D. and Catherine T. MacArthur Foundation.