Developing a new data system for Free Schools in the UK
Improved data quality has opened up opportunities to use advanced technology like AI
As part of our partnership with DfE, we’ve been transforming the data system for managing free school projects across the UK – the Free School Store (FSS). Usability, data quality, and accessibility issues had emerged over time. We took the project from discovery to beta.
It’s a great example of how we’ve been transforming legacy systems and improving data accuracy and use. We enabled DfE to fully transition away from the legacy system, saving substantial maintenance costs, and improved data quality has opened up opportunities to use advanced technology like AI.
It was a complex piece of work involving a team of people, and I wanted to share some headlines about the approach we took.
What are free schools?
Free schools are academies that operate independently from local authorities. They have several freedoms including:
- setting their own pay and conditions
- greater control over their budgets
- not needing to follow the national curriculum
They’re typically new institutions built with DfE funding, so it’s important to capture information on applications as well as data on build costs and development.
The FSS was used widely by various stakeholders including delivery officers, team leaders, the School Places Analysis team, and the Grant Management and Funding Unit. It was outdated and due for decommissioning.
Open a free school – journey map
Understanding user needs and system challenges
Through discovery we wanted to identify user needs and pain points and how to address the systemic challenges with FSS. We interviewed over 100 users, which helped us understand the main issues:
- data entry: there were too many fields, many of which were irrelevant to specific tasks
- navigation and usability: without a modern interface, navigation was challenging particularly for users with accessibility needs
- data quality concerns: the lack of field validation meant there were frequent data entry errors
- low confidence and use: given these issues, FSS users often relied on alternative data sources which was inefficient and meant data was being stored in more than one place
Validating our hypothesis and prototyping
We knew from the outset that FSS required a comprehensive overhaul, so we needed a clear roadmap for alpha that targeted high priority issues and would establish a good foundation for the new system rollout.
Our aim was to validate our hypotheses and develop some prototypes that would address the main challenges with targeted improvements. These included:
- simplified data entry and navigation: refining input fields and enhancing workflows to reduce complexity for users
- enhanced accessibility: adhering to GDS standards and prioritising inclusivity to ensure usability for users with diverse needs
- improved data quality: integrating validation mechanisms to enhance data accuracy and build user confidence
We worked with delivery officers, policy advisers and other stakeholders to map out the full workflow and identify critical data points. Then tested 3 prototypes with users focussed on simplifying data input and enhancing information retrieval.
By analysing the existing fields we were able to focus on core data needs, reduce duplication, and make sure data entry was efficient. Automated checks and validation prompts helped us improve data accuracy and reduce the need for back-end corrections.
We built using C# .NET and Azure for compatibility with DfE’s infrastructure and kept a firm eye on sustainability by optimising server usage, our reduction of duplication, and investigating green hosting options.
We also carried out partial accessibility audits and created user profiles for different accessibility needs so we could do some system walkthroughs.
Refining our MVP
In beta we looked to refine our minimum viable product (MVP). With our insights from alpha, we had a clear technical and usability direction. Alongside moving to new technical architecture, we prioritised areas to improve for users:
- streamlined workflow: FSS used 1,152 data fields. We determined that only around 245 of these fields were needed (with about 50 new fields added afterwards)
- improved usability: we addressed the 30+ tasks identified for delivery officers and School Places Analysis team members, while maintaining familiar tasks/milestones to make the transition easier for users
- meeting accessibility standards: we designed the new system to meet accessibility standards and ensure it was inclusive (making it WCAG compliant)
- reliable data quality: the system was designed to make sure information entered by delivery officers would be a reliable data source
Here’s a few of the things we did.
Co-designing the “create project” journey
The process for creating new free school projects and setting up records wasn’t intuitive and lacked comprehensive data validation. We did some co-design work with DfE’s subject matter experts. Then prototyped a new journey with a linear data-entry form to reduce cognitive load and improve accuracy, and tested it with users.
Mapping the new “create project” journey
Overhauling risk ratings
The risk ratings in FSS used a simplistic “high” or “low” system that didn’t reflect policy guidance. It also didn’t capture historical data which limited the ability to track changes. We redesigned the risk ratings journey, implementing the required Red-Amber-Green system and adding a historical view. Developing the design iteratively, informed by user research.
Screenshots of new risk ratings
Reusing solutions across teams
Several DfE services act as project or case management systems similar to the one we were developing. To maintain consistency, we analysed how other teams handled project lists and filtering, then customised these solutions for our unique needs.
Cognitive walkthrough for accessibility
We tested the new design using accessibility profiles. Feedback from users and other project teams highlighted areas for improvement, including logical content arrangement, moving less relevant content to other sections, and clear labelling.
Using lift and shift for minimal changes to familiar tasks
Some user journeys were well established and familiar to users. For low risk tasks, we employed a “lift and shift” approach, replicating successful legacy workflows. We reviewed the guidance with DfE experts, so we were able to transfer these tasks confidently and free up resources for higher priority areas.
Outcomes and feedback
We launched the first phase of the new system – Manage Free School Projects (MFSP) – in June 2024, with a second release in October. The public beta delivered significant improvements in user satisfaction and operational efficiency:
- reduced task time: design changes led to major efficiency gains, for example, reducing “Create Project” journey task time by over 70%
- enhanced risk management: the new risk ratings system enabled better monitoring and risk management with improved data
- high user satisfaction: user feedback was overwhelmingly positive reflecting a clearer and more intuitive user experience. FSS had a baseline System Usability Scale score of 47, while MFSP achieved an impressive score of 91
A great achievement for our blended team.