This project was commissioned by SLAB (Scottish Legal Aid Board) and was a service design-led project focused on improving access to holistic money and debt advice as part of the digital advice service at Shelter Scotland. I joined the project after the initial research phase to help analyse the research, ideate, and design potential solutions.
The first step in the process was understanding the needs and barriers faced by individuals seeking money and debt advice. The initial research phase of this project was very comprehensive, featuring multiple rounds of interviews with people who had lived experience of homelessness and money problems. We also conducted surveys and desk research to explore how these individuals access money and debt advice and the current barriers they face.
After analysing and affinity mapping the initial data with the team we were able to pull out some strong themes that possessed the potential to become barriers for people seeking advice through Shelter or other services.
We found participants reported that the information provided was often overwhelming and challenging to comprehend. Additionally, negative emotions, such as shame and embarrassment around their financial situation, made it more challenging for them to seek advice from real people, even though speaking to someone was often the most beneficial for their understanding of how to get help.
With the team, I facilitated an ideation workshop to generate ideas for making money and debt advice more accessible, reassuring, and easier to understand. During the session, we generated 7 innovative concepts, including a 1-step answer tool, a guide to money problems, a way to adjust the level of information, and finally, the narration tool. After evaluating these concepts based on their value and effort, we chose the narration tool as it provided the highest perceived value to users.
After scoring the concepts, we decided to validate the narration tool idea to determine if it was worth pursuing. To do this, we recruited 6 participants from our website using Ethnio and conducted remote moderated sessions over Zoom to validate the concept.
The first prototype was very low-fi, created in PowerPoint, and aimed to understand if people would have a better comprehension by listening to advice. We also captured preferences for the type of voice and speed used to narrate content. The participants provided valuable feedback, expressing a dislike for robotic voices and the need for adjustable speed to meet individual preferences.
First prototype
Following concept validation, I created a high-fidelity mock-up of what the narration tool would look like and worked alongside the developer to create the prototype using Azure and Murf.ai.
To allow participants to use the narration tool as it might appear on the page, we created this prototype within a staging environment. The first iteration of the tool used an audio file that would play when a participant interacted with it and provided complete control over the audio, including voice selection and playback speed.
To test the prototype, we conducted guerrilla testing over two days in Dundee and Kirkcaldy central libraries, involving 17 people and focusing mainly on the mobile design.
A significant number of participants found that hearing information helped them better understand it, especially those who identified as neurodiverse, such as having dyslexia or ADHD.
Some participants mentioned finding it useful in case they had forgotten or lost their glasses.One participant with dyslexia shared that she had to convert all her university notes into audio to successfully navigate her course and would have found a tool like this massively beneficial.
During our observations, we noticed that participants would occasionally lose their visual place on the page while using the tool. Interestingly, while participants frequently adjusted the speed of the narration tool to suit their preferences, they showed no interest in changing the voice and tended to stick with the narrator they first heard, regardless of the three different voices that were tested.
After the guerrilla testing, we took the insights we gained and made refinements to the prototype. One important enhancement was the addition of text highlighting, which allowed users to visually follow along with the narration.
In addition, we began exploring ways to scale the tool and ensure its smooth functionality on the live site. This phase involved addressing technical challenges, such as data storage and scalability, to ensure the long-term sustainability of the tool. To achieve this, we used advanced AI software that would read the page and stream the narrated audio files back to the user. This approach had two significant benefits: it prevented the need to host large audio files directly on the site, thereby avoiding potential slowdowns in page speed, and it made the tool more hands-off by eliminating the necessity to regenerate audio files whenever page content was altered.
Before launching the tool, we conducted 5 unmoderated usability tests on both desktop and mobile platforms. To further validate our earlier finding regarding the tool's benefits for neurodiverse individuals, we ensured that at least 50% of the test participants fit this profile.
During this round of testing, we discovered that participants highly appreciated the option to listen to the page, especially since most of them had never encountered it before. Among the various features, the text-highlighting functionality was regarded as the most valuable.
Participants with dyslexia expressed particular satisfaction with the tool, finding it extremely helpful. Additionally, other participants mentioned the convenience it provided for multitasking, such as taking notes while listening.
" It's giving you advice and help and doing it in a way that's really easily understandable. "
Live on Shelter Scotland site
Initially, the Narration Tool will be implemented on the rent arrears page within the money and debt advice section to gather insights from real users and assess performance.
This page was chosen as a location where the tool could offer high perceived value for users, based on the results from earlier testing and conversations with the digital advice team in Scotland.
By deploying the Narration Tool in a controlled environment, we aim to gather valuable insights into user interaction and make any necessary changes or improvements before rolling out the feature across more pages.
The launch of the live page marked an important milestone in the project. We have established key metrics to measure the adoption rate, time spent on the feature, and error rate, which will provide valuable insights into the tool's performance. Additionally, gathering feedback from users through surveys when the tool closes or is paused will help us assess its impact on their overall experience with digital advice. By tracking these metrics, we will be able to gauge user engagement and determine the feasibility and benefits of expanding the tool to more advice areas.
Throughout the project, we have seen successful outcomes in various areas, including research, ideation, prototyping, and user testing. Embracing innovative techniques like guerrilla testing has yielded positive results and enriched our understanding of user needs.
By implementing the narration tool, we have provided individuals seeking financial advice with a more accessible and user-friendly means of accessing oral advice. The tool's clear and understandable content presentation aims to empower users to make informed decisions about their financial situations. This project has effectively showcased the value of service design in enhancing services and improving user experiences.
Overall, the project has been successful in achieving its objectives, but there is room for further improvement, particularly in establishing a comprehensive feedback mechanism to measure the tool's impact and support its ongoing development.