Creating clearer and actionable error messages

One of my first projects as a content designer at Meta FinTech involved updating error messages related to failed payments. Failed payments were a frustrating problem for users and the business.

Timeline

April 2022

Problem

The current payment error messages were vague and didn’t provide users with specific actions they could take to resolve the issue. This led users to contact support.

Team structure

  • 1 Content designer (me)

  • 2 Software engineers

  • 1 Data scientist

  • Legal team

  • Privacy team

  • Security team

My objectives

  • Help users resolve their own issues by making the error content clearer and more action-oriented

  • Align with requirements from the legal, privacy, and security teams

The process

Data and research analysis

Since I was new to the Meta FinTech team, I needed to gain context on why the payment issues occurred and how they could be resolved. I set up time to meet with the data scientist to better understand how often each error occurred and if users could potentially take action to resolve the errors themselves. Based on content design best practices, my hypothesis was that it’s beneficial to provide more specific information about why an error occurred. The data scientist confirmed that specific error messages would be most beneficial to the user.

I also worked closely with the engineers to understand where the error messages occurred. Since they occurred across many billing flows, the error messages needed to be scalable.

Explorations

Generic vs. specific

I explored different iterations for making the messages generic and specific. Specific messages provided the user with more insight into why an error occurred. Generic error messages could scale across more error types and would be faster to implement and update in the future.

Content tones

My iterations also explored using different content tones, including sympathetic, informative, and helpful tones. Using a sympathetic tone could make the error messages feel more human and understanding, however it wasn’t appropriate to use a sympathetic tone if the error wasn’t due to an issue with Meta’s systems. Using an informative tone led to more straightforward and purposeful content, while a helpful tone could help the user feel more supported.

Iterations

I shared my error message content in a design crit feedback session. The other designers agreed that the most impactful error messages were specific and used an informative tone. However, the legal, privacy, and security teams shared some concerns around specific messaging. I worked with these teams to identify which messages could be specific without posing a risk. I also iterated on the content to create messaging that was generic enough to align with privacy, security, and legal concerns, but specific enough to provide helpful context to users.

Error message standards

After this project, I became the go-to person for error messages on the Billing team. I identified the opportunity to create standards for the error messages and create a resource that would help us reuse the existing content. I worked across teams to understand how they created error messages and shared my standards with other teams.

Results

A data experiment revealed that the error messages were statistically more likely to prevent future payment issues. Due to the success of this work, my team planned to implement my error message strategy across other flows within billing.

Challenges

Complexity

The billing space is very complex, especially for someone who is new to the team. It was only my second week on the job, so I appreciated the opportunity to dive into this new space. I’m very detail-oriented, which led me to ask deep questions about both the specific error messages and the larger experience. It’s very important to be aware how the messages fit into the broader system because the content can surface in many different products.

Legal, privacy, and security requirements

The legal, privacy, and security teams helped me identify use cases that could negatively impact a user, which was helpful. However, some of their feedback conflicted with content design best practices. This required me to constantly iterate on my content to balance legal requirements with providing a great user experience. I worked collaboratively with these teams, which helped me efficiently address any feedback.