What’s Next for Your Government Website? Let A/B Testing Tell You.

by Cris Paden

posted on 08-15-2017

By Kumar Rachuri, Director of State and Local Government Solutions

Ever find yourself in a debate with a colleague over what banner will work best? Or what headline or button copy will be most effective? These are both essential questions for every online experience. But sometimes, the loudest person in the room wins the debate.

With A/B testing, you can rely on measured results—not loud voices—to determine the best results.

Through testing, everything can be made into a contest, which is enjoyable for your internal team and motivates people to share data and ideas. The whole testing process is all about working toward an optimized goal. However you get there, a platform that can test everyone’s theories can transform the creative and collaborative energy of the room. (Tip: The best agency employees to work on this are those who naturally question everything.)

For example, let’s take the average job search on a local government website. Is it more effective to search for jobs prioritizing location or by job category/requirements? We know, of course, that “both are important.” But what is more important?

For any online experience, agencies must prioritize requirements and set them into a hierarchical flow to achieve the best user experience. Otherwise, everything is set at the same level of importance in a design, and then the user’s eye won’t know where to look (or what to do) first, second, third, and so on.

Remember, every user is begging, “Don’t make me think” (taken from a book by Steve Krug). Every step must be effortless and clearly laid out—one at a time—with minimal distractions.

Example of button test from “Don’t Make Me Think: A Common Sense Approach to Web Usability” by Steve Krug

In order to test for and determine the priorities of government job seekers, come up with a question and hypothesis based on your research on the target market. What kind of people apply for government jobs? What do they care about most when deciding to apply to a job?

Start with your hypotheses of those questions—what you think might be true– and test the to see whether a hypothesis (A) of a different hypothesis, or counter (B), is more true.

In the job search example, your approach might look like this:

Hypothesis (A): If you think job seekers are looking for convenience in jobs, such as working nearby their home, test a map view of the search results. The user can add their address to see how far each job option would be, how much travel is required, and other details such as salary and commitment required.

Counter (B): Job seekers care more about job details and requirements than location. Perhaps they don’t even know what job they want. In that case, test a job match of the search results. Allow people to enter their skills and requirements in an easy-to-use interface, and show a recommended list of the jobs that match their needs.

Then, you split-test these two variations of approaches. Half of the users see one variation, and the other half sees variation B. See which interface converts better. One measure of conversion is number of job applications received. Another can be length of time spent on the page.

This is just one high-level example out of endless ways you can test what’s best for your citizen, but the principles hold throughout. Once you have results dictating what’s best for your specific citizens, you can design a more personalized experience that helps them.


Adobe Target gives agencies the power to quickly experiment with various personalized experiences to figure out the most effective experience for their specific audience. It even allows you to decide what you count as effective—whether that’s conversion rate or revenue.

Target allows you to easily change and alter pages or pieces of a page in order to see which design works best. It specifies the top five most predictable variables to show why one test is performing better than the other. You can then build upon your theories and segments by simply changing elements of your page. That allows you to improve every stage of the user’s journey.

The goal is to find out what works for your audience, discover deeper audience segments, and uncover creative opportunities yourself. Target also offers geo-targeting features, automated personalization of content and design, and gives optimized creative recommendations based on your user’s behavior and data.

Using Target, Marriott was able to increase enrollments to Marriott Rewards by 50 percent. The hotel captured all their customer data within Adobe, built the brand via social media, and improved the overall digital experience. In the end, Adobe helped Marriott prioritize and personalize their goals, including what to push to what customer at the individual stages of their particular journey.

These same discoveries can be made using Target with government sites today, which empowers agencies to deliver an experience to citizens that is leaps and bounds ahead of what they offer today. The only way to know where to invest your money and labor, is to analyze and test intelligently with your team, and strike the magical combination of copy and creative that cause citizens to respond.

Test and improve with your team using Adobe Target.

Topics: Government