,

Rephrasing for Results: Matching Your Words to the Way Your Audience Thinks

In a previous post, I discussed the usefulness of “A/B testing”— the practice of running randomized experiments with multiple wording or design options in order to see which one performs better.

One particularly useful application is helping to communicate in ways that match the way your target audience is thinking.

Case Study: Helping People Understand the Child Tax Credit

In 2021, my colleagues at Code for America launched a website to help low- and no-income families claim the Child Tax Credit.

Yet at the end of the year, millions of eligible families were still not claiming the Child Tax Credit. Some of my colleagues dug into this deeper to try to figure out why this was happening and what might change things. They found that messaging may have played a contributing role, writing:

Many campaigns designed to encourage people to claim the CTC did so by telling families they could receive up to $3,600 a year per child. The implicit assumption here was that highlighting the total yearly CTC amount ($3,600) vs. the monthly amount ($300) would increase people’s interest in claiming this benefit.

But $3,600 might be hard to conceptualize if your budget isn’t based on yearly numbers — and for most people, it isn’t. In fact, based on a study we conducted among people who receive government benefits, less than 1% budget on yearly basis. The vast majority of people who use a budget—over 85%—report budgeting on a weekly or monthly basis.

We learned that when we communicated Child Tax Credit amounts in terms that matched how people were already thinking about their budgets — on a weekly or monthly (rather than annual) basis — that 16 to 26% more people showed interest in applying.

Doing It Yourself

There may be opportunities for you to apply this kind of approach in your work! 

Think about forms that you ask people to fill out or information that you publish or send, such as website text, mailed or emailed notices, and marketing campaigns. How are you describing things? For example:

  • Money received through a certain program (such as with this Child Tax Credit example): Per week, per month, per semester, per year?
  • Penalties that someone may face, or benefits they may lose, if they do not take a certain action: Should you give the minimum, maximum, or full range of consequences?
  • The amount of time that a certain process takes: Should you measure in days, in weeks, or in months?

In addition to just looking at how you frame one aspect of a question, you could also look at options for multi-part questions, such as our experimentation to help people more easily report their income.

Using quantitative and/or qualitative methods, you can test out different ways of framing content or asking a question. Then, use the results to inform decisions about what to write—and of course feel free to keep iterating. Even if you don’t have the resources to do research and testing that’s as rigorous as what was done in this Child Tax Credit example, even simple experimentation can provide helpful illumination.

By experimenting with different ways of asking questions and providing information, you can help people take the right actions and better access your services.


Greg is the Associate Director for Human-Centered Government at Code for America, where he is leading efforts to support public servants with resources and training on the organization’s principles and practices for how government can and should serve the public in the digital age.

Images by Mikhail Nilov at Pexels.com and Code for America.

Leave a Comment

Leave a comment

Leave a Reply