.png){kind=small}
Our Approach to Scaling UX
Achieving scale is one of the holy grails of any business, but scaling craft has historically been next to impossible. User Experience is a foundational aspect of itswhatsnext, not just as a service we offer as a business, but it’s forms the bedrock of all subsequent decisions we make.
We are often dealing with the tension between the business critical nature of “Great UX”, and the hours required to achieve it, and the increasing demands of additional clients or projects.
Knowing scale comes from automation, we redesigned our approach to UX and integrated a custom GPT model that allows us to speed up data ingestion, interpretation and analysis.
The outcome:
Consolidated sprint head hours and timeframes
Improved thematic analysis
Improved ideation
Our process outlined below:
Classic Sprint
Day 1: Understand
Morning:
Kickoff meeting: Set the stage, establish long-term goals, and begin formulating a focus for the sprint week.
Expert interviews: Engage with internal and external experts to gather insights and input on the problem space.
Afternoon:
Define the challenge: Synthesise the morning's insights into a clear problem statement.
Map out the problem: Create a user journey map that outlines the steps the user takes to interact with the product or service.
Day 2: Diverge
Morning:
Inspirational talks: Look at how other companies have solved similar problems.
Sketching: Engage in exercises designed to generate a broad variety of ideas and solutions, such as "Crazy 8s."
Afternoon:
Solution sketching: Individually sketch detailed solutions, focusing on one particular aspect or the entire user journey.
Day 3: Decide
Morning:
Present sketches: Each team member presents their sketches without revealing the author.
Voting: Use dot voting or another decision-making process to choose the most promising ideas.
Afternoon:
Storyboard: Turn the winning sketches into a detailed storyboard that serves as a blueprint for the prototype.
Day 4: Prototype
All day:
Build a prototype: Create a "fake" product that is realistic enough to test with real users. This can be done using various tools and methods, depending on the complexity of the product and the resources available.
Prepare for user testing: Develop interview guides and plan the logistics of the test day.
Day 5: Test
All day:
Conduct user tests: Interview and observe users as they interact with the prototype. Typically, five users are enough to uncover most major usability issues.
Gather feedback: Collect and document observations, looking for patterns and insights that can inform the next steps.
iwnGPT Sprint
Day 1: Understand, Diverge, and Decide
Morning:
Kickoff and Goal Setting: Briefly outline the sprint goals and expectations. Unlike the classic sprint, this session is shorter, focused on aligning the team's understanding of the objectives.
Data Ingestion and Analysis with iwnGPT: Input all available user data, market research, and previous feedback into iwnGPT model. The AI then analyses this data to identify key user needs, pain points, and opportunities, providing a comprehensive report that would traditionally take days to compile.
Midday:
Expert Input and AI Integration: Incorporate insights from internal or external experts directly into the GPT model to refine its output. This step ensures that the AI's findings are balanced with human expertise and intuition.
Idea Generation with iwnGPT: Use the AI to generate a wide range of solution ideas based on the analysis. The team can prompt the GPT with specific challenges or allow it to propose solutions autonomously.
Afternoon:
Idea Review and Selection: The team reviews the AI-generated ideas alongside any human-generated ideas. This collaborative process may involve dot voting or another prioritisation method to select the most promising concepts.
Storyboarding with AI Assistance: Develop a detailed storyboard for the selected concept(s). iwnGPT can assist by drafting user stories or suggesting key frames for the storyboard, accelerating the process.
Day 2: Prototype
All day:
Rapid Prototyping: With the storyboard as a guide, create a high-fidelity prototype that's ready for user testing. iwnGPT can contribute by generating copy, suggesting design elements, or even producing code snippets for interactive features.
Prepare for Testing: Finalise the prototype and plan the testing session, including drafting interview questions or tasks for participants. The AI can help by predicting potential user questions or concerns based on its analysis.
Day 3: Test
All day:
Conduct User Tests: Perform user testing sessions as planned. The team observes and records user interactions and feedback.
Feedback Analysis and Next Steps: Use iwnGPT to quickly analyse feedback from the testing sessions, identifying patterns and insights. The team then discusses these findings to decide on immediate iterations or longer-term improvements.
Why it works, and what we use it for:
Sprint with iwnGPT - Benefits:
Efficiency: Reduces the initial phase of the sprint from three days to one, allowing teams to move faster from understanding to prototyping.
Depth of Insight: The AI's ability to process and analyse large data sets can uncover insights that might be overlooked in a manual process.
Creativity: Combining AI-generated ideas with human creativity can lead to more innovative and diverse design solutions.
Iterative Speed: With AI assistance, teams can quickly iterate on designs based on user feedback, enhancing the product's user experience more rapidly.
Sprint with iwnGPT - Use Cases:
Creative variation testing: understanding creative efficacy
eCommerce testing: UX, user journey optimisation
Service utilisation optimisation: Service bottle necks and drop offs
Loyalty program utilisation:
New product and feature ideation:
Conversion rate optimisation
Get in touch with us to see how we can drop the cost of your UX by integrating iwnGPT into your workflows