Improving your website doesn’t have to involve a major redesign that takes months to complete. Instead, by using the findings from live testing, you can continuously update your site and make easy, effective changes that move you toward specific goals—like boosting online giving or recruiting more members.
But in order to make the most of live testing, nonprofits need to stay acutely focused on those goals, says Shanta Mali, director of membership and digital strategy at Central Park Conservancy. Live testing lets you determine which suggested changes to a website perform better by trying them out, one at a time, on people using the site. The Conservancy, for example, has run tests to see which website features increase the number of new members and which ones encourage more people to sign up to receive emails.
Staying focused on goals prevents you from getting sidetracked by options that look flashy but may not help your nonprofit accomplish its mission. For example, changing online content so that it gets more clicks, but not more donations, may not be the best use of your charity’s resources.
Central Park Conservancy worked with M+R, a firm that specializes in nonprofit marketing, to build a one-month testing plan by first identifying goals for its website, then developing ideas for reaching those goals, and testing those ideas. Using this method, the Conservancy increased its membership conversion (the number of people who follow through on an action after encountering the opportunity to act) from 10 percent to 16 percent during the first 11 months of 2014.
After thinking carefully about your nonprofit’s broad goals, follow these steps, shared by Ms. Mali and Cameron Lefevre, senior strategist for M+R, to develop a plan for running live tests focused directly on your group’s needs.
Set goals for what you hope the website changes will accomplish during the next six to 12 months. For example, your organization may seek to increase the number of online membership renewals or attract more monthly donors.
Research how your website currently performs, and then set testing objectives for improvement. Use digital tools that track how people use your website—including Google Analytics, which monitors website visitors and other metrics; Crazy Egg, which shows where people are clicking on your website; and Lucky Orange, which records videos of how people use your site—to figure out your website’s baseline data on traffic, conversions, and other measurements of user behavior.
Decide which new targets you’d like to hit. Then identify the challenges, opportunities, and intended audiences for each target, and figure out the steps you need to take to achieve it. For example, to increase the number of people who become members via the website, you’ll first need to drive more traffic to the online membership form and then increase the percentage of people who become members after reaching the form.
Brainstorm test ideas and create a testing roadmap. For each test, develop hypotheses about which website changes might help you meet the goal by looking at other websites and brainstorming with colleagues. Depending on the target, you may want to think about ways to get more attention for a particular webpage, simplify certain processes, or increase users’ incentive to follow through on a specific action.
For example, if you want to increase the number of people clicking your “donate” button, you may decide to run experiments to see what happens when you change its color, size, or position or rewrite the text that surrounds it.
Create a testing roadmap to determine which tests to run first, taking into account which targets are top priorities; which might have the greatest impact; and which require the most time, money, or tools to test. Running tests one at a time allows you to identify which changes caused which effects.
Here’s a sample website testing roadmap, created by M+R, that your nonprofit can use. The spreadsheet will calculate scores for each kind of test depending on the values you assign (high, medium, or low) in the following categories, “priority of the goal,” “potential impact,” and “resources needed.” Tests with higher scores should be run first.
Start to run tests. It’s important to note that the amount of time it takes to accumulate statistically significant results will vary depending on the amount of traffic the website receives. Use tools like the A/B testing platform at optimizely.com or the sample size calculator at evanmiller.org to figure out how long you’ll need to run your tests.
