ODI Data & Public Services Toolkit

This toolkit is designed to help people designing and delivering public services

Data and Public Services Toolkit (case study)

Thu Dec 26, 2019
$download_content = get_field('download_content');

Discover how we developed the Data and Public Services Toolkit which aims to address the challenges public authorities face when trying to use data

Summary

The Data and Public Services Toolkit aims to help public authorities make better use of data. It addresses challenges such as: how to make a business case for data use; how to use data ethically and fairly; and how to identify and utilise relevant datasets.

The toolkit was developed as an output of the New Service Delivery Models project, part of the ODI’s three-year research and development programme, funded by Innovate UK.

UK public services are under pressure to deliver efficiency savings along with improved outcomes for citizens. As part of the R&D programme, we aimed to explore how public service delivery interconnects with the private and charitable sectors and to advise how better data use can make services more effective and efficient.

The toolkit includes the Data and Public Services Business Case Canvas; the Data Ethics Canvas; the Data Ecosystem Mapping tool; How to design to scale checklist; and Data and Public Services case studies.

The aim was for the toolkit to be a practical, useful – and beautifully-designed – package, to be used collaboratively by all those involved in public service delivery, not just people with technical skills.

The toolkit aims to help change how local authorities perceive and work with data. At the ODI, we advocate for better and more ethical data use, and the toolkit provides a practical hands-on tool to help embed and enable good working practices. It aims to equip local government and service designers with the support, tools and capacity they need to adopt the changes we advocate for at the ODI.

As well as being open for anyone to pick up and use, the toolkit also provides a useful set of resources that we use in our consultancy and advisory work, making those engagements more effective and helping to bring additional funding into the ODI.

Key facts and figures

Our role

The ODI team working on the toolkit included researchers, writers, content specialists, designers and project managers. We also engaged with ODI partners, ODI members and regional offices including ODI Cardiff to get expert input to help shape the toolkit.

We based some of the requirements for the toolkit on the research undertaken in the first year of the New Service Delivery Models project. This involved extensive desk and user research, gathering case studies and working with local authorities to identify the barriers and opportunities to working with open data.

The research found that open data can be used in the public sector to help deliver services in three key ways: to improve access to public services; to support more efficient service delivery chains; to ensure more informed policy development.

We worked with ODI Devon to produce some research into the local authority user needs and requirements. This led to the development of learning personas in early 2018, which helped inform the design and the scope of the tools, ensuring that we considered the viewpoints of the different personas, including service delivery managers, budget holders and data users.

We used existing and new case studies to inform the toolkit, showing how open and shared data has been used to improve public services. As well as researching new case studies as material for the toolkit itself, we drew lessons from the services developed by four local councils as part of an ODI stimulus fund.

The design process involved various internal reviews, looking at the tools already available, the use of language and design, and what new tools were required. The review revealed several overlaps between our existing tools and the requirements for the toolkit. We therefore made a decision to iterate on existing tools, for example the Data Ethic Canvas, as well as developing new tools, such as guidance around developing a business case.

We prototyped different versions of the toolkit, ranging from online only, to a slide deck pointing to existing tools, to a mock-up based on the original Data Ethics Canvas. Following feedback from service designers Snook – which identified that there was a very high barrier to entry and that we needed to group our tools together in a simple, user-friendly way – we decided to produce a hard-copy, eye-catching version of the toolkit, alongside a digital version.

In November 2018, we ran three workshops across the country, in Glasgow, Cardiff and London, to test early versions of the toolkit. The workshops included participants from: service designers including UsCreates, Snook and UrbanTide; local authorities including Manchester City Council, Essex County Council, Glasgow City Council, Caerphilly County Borough Council, Bristol City Council, and Neath Port Talbot Council; and other organisations including Data Cymru.

The workshops helped gather face-to-face feedback, enabling us to find out how local authorities could actually use the tools in their day-to-day work. It allowed us to hone the content and focus on user needs, and to get advice on everything from language and format, through to delivery and required support.

The toolkit then went through various stages of iteration, involving policy, research and design colleagues at the ODI. This helped us ensure we had captured requirements from key internal stakeholders, as well as ensuring we hit the requirements outlined in our user research and desk research.

We ran a joined-up marketing and communications campaign to disseminate of the R&D work ODI has done about public services, and equip people in the public sector to understand why and how data can be used to deliver better services. This included an audience-specific landing page, talking-head video, and a lead-capture form.

We worked with external designers and our in-house Digital Production Editor to develop a strong visual identity for the toolkit, which gave each tool its own identity, while also providing a consistent look and feel showing the tools were part of the same ‘family’.

The printed toolkit is presented in a package format, containing the five outputs (leaflet, canvas or fold-out tools). The online versions are ‘print at home’ black-and-white versions, that are designed for low-ink use on standard printers.

The toolkit was shared via @ODIHQ on Twitter to 57,000 followers in July 2019

The toolkit generated interest on social media, with shares from the Service Design Fringe Festival, Civic Hall Toronto, Code for Canada and Design Humanly who tweeted:

The team has presented the toolkit at a range of events to promote the toolkit. This included the ODI Open data for local government meetup in September 2019 – with attendees including local authorities and service designers; and Local Gov Camp in November 2019, where we ran a training session that generated several tweets including: ‘Nice business case canvas in the @odihq toolkit. #localgovcamp’

We ran an Open Cities Workshop on 9 December 2019 in Bristol, with participants including Bristol City Council, Bristol is Open and Avon & Somerset Police. Feedback included:

  • “Really enjoyable and useful learning experience with subject matter experts”
  • “The data ethics and morals session was highly thought provoking and linked to an ODI Data Ethics Canvas which is a genuinely useful tool”
  • “The tools provided have offered us a way to navigate through the data challenges we face as an organisation and champion data issues amongst colleagues”

The toolkit is being shared and disseminated via the ODI Data Access and Public Services campaign, a cross-project campaign that brings together all ODI offerings for the public service design audience. The dissemination campaign includes events, meetups, press articles and social media engagement.

What was challenging?

Running the initial research and engagement phase separately to the design phase was challenging. We completed the research and engagement with local councils, and then moved on to the design phase. This meant the design phase felt quite separate and was more drawn out than it may have been if it had been ‘baked in’ as part of the user research and engagement. The design process actually took longer than the research. It would be useful to align ideas from both end-user and ODI perspectives around design early on.

We had to invest quite heavily (time and budget) in time with the external designer, with our in-house designer working at the external design studio. While it was incredibly useful and delivered a beautiful product, we should have allocated some of this resource to spend more time working with local authorities themselves.

We wanted to produce a tool for the public service design audience, but that also worked as a core ODI piece of content. To achieve this, there was a lot of ‘design by committee’, with many rounds of feedback on content and design and no clear sign-off process. This possibly led to some dilution as we tried to meet the needs of too many people.

A lack of organisational memory can cause problems. The team for the project brought in experience that was lost when personnel changed. Better documentation and internal systems may have improved this.

For future projects that create tools and toolkits, we should:

  • focus on user needs, ensuring the message is not diluted by competing internal priorities, and that there is good engagement with the key audience(s).
  • build in prototyping from the early stages for user research. It is useful for research participants to see, use and play with early prototypes as an effective way to refine our understanding of user needs.
  • have a very clear time-bound sign-off process, with accountable people at each stage.
  • have more design in-house, and have design more embedded in the research than as a separate phase.
  • ensure good documentation, and clear direction from the outset.

What went well

The final product is a beautiful toolkit – hard-copy and online – that adds a lot of value, particularly for this type of output, given the topic of data is arguably pretty dry and abstract. It succeeds at getting people – who don’t think that it’s their thing at all – to engage with it. It is important to employ use quality design. Presenting rich visual material to people is really positive and effective.

The research stage of the project informed the requirements. These led to the development of the tools, which are now being used across other parts of the ODI – training, business development, workshops for other projects etc.

We were able to build services to support the use of the toolkit. Fifty-eight people booked on ‘Introduction to Data Ethics and the Data Ethics Canvas’ course during the duration of this campaign, and we delivered £19k worth of in-house training to external companies.

The hard copy and online versions worked well together – with the hard copy delivering an eye-catching bright tactile tool, and the online version providing extra context where needed – for example, the online version of case studies is more extensive than the summaries in the hard copy.

It was great getting the team to effectively engage externally. The project team realised how effective it is to get out of the office and speak to people, to really learn about people’s needs.

Sharing messaging and products through campaigns works well. The specific ODI Data Access and Public Services campaign has worked well to develop an audience-focused plan, with specific KPIs and plans for who, how and when to engage with people.

We’ve built strong relationships with the Ministry of Housing, Communities and Local Government, and also strengthened existing relationships with local authorities, such as Bristol City Council. This has had a positive impact on engagement in other work like the Open Cities project. We’ve been able to build on those networks started in projects like this.

For future projects, we should build our capability in design, so it is a built-in element of the research process when we are prototyping ideas. The design element is really important, and possibly seen as unusual/radical in a data product which can create more impact. We should continue to engage with our key users and external experts, ideally face-to-face, and in people’s own environments, to get authentic insights into needs and working methodologies.

What have we learned?

Our tools should be accessible to all audiences. We have received feedback around the lack of accessibility of the Data Ethics Canvas (in terms of screen-reader compatibility), and this also applies to other tools. The Data Ethics Canvas was not developed to meet Web Content Accessibility Guidelines (WCAG). We should ensure all outputs are compliant to AA success accessibility criteria of the WCAG.

The engagement in workshops before and after launching the toolkit was great. We gained a lot by running workshops in different cities, in terms of learning about local authority culture and working practices. Our Open data for local government meetup in September 2019 – with attendees including local authorities and service designers – was a really effective format to share the toolkit, and use the meetup community to help with dissemination.