New service delivery models: four councils using open data for service redesign

Mon Apr 30, 2018

Open data and public services

New service delivery models – an ODI project – aims to increase understanding of data-enabled service delivery models in local government, and to encourage more open data to be published to improve public services.

The project is part of the ODI’s £6M three-year innovation programme, which has an overall aim to build data infrastructure, improve data literacy, stimulate data innovation and build trust in the use of data.

As part of the project, we awarded funding to four forward-thinking local government organisations to develop open data projects which explore how data could be used to improve public services – making them more efficient, innovative and citizen-focused.

This element of the project builds on the report – Using open data for public services published in February 2018 – where we looked at examples of open data use in public services. Following completion of the four local government projects, we asked the teams to tell us about their experience – the successes and challenges – in the hope other service delivery teams working on open data projects will find their experiences helpful to learn from. There’s a case study from each project and each team will be publishing tools created during the project that anyone can use.

The teams and their challenges

Doncaster Council and UsCreates

Doncaster Council explored the careers information, advice and guidance service for young people across the borough. Currently in Doncaster, careers data is fragmented and difficult to access. The team explored helping young people get better, tailored information about their career options for training, education and employment.

Kent County Council, Kent Energy Efficiency Partnership, and UsCreates

Kent County Council is committed to reducing fuel poverty. Fuel poverty has short and long term detrimental effects on physical and mental health. The team looked at datasets from across the NHS and the wider public sector to more accurately and effectively help those at risk of fuel poverty.

North Lanarkshire Council, Snook and UrbanTide

North Lanarkshire Council is on a mission ‘to improve economic opportunities and outcomes for all’, and now has an ‘open by default’ data policy for non-sensitive data. The council applied this open approach to its business rate data, to better understand the demand for the data and reduce Freedom of Information requests.

London Borough of Waltham Forest, the Audience Agency and Technology Box

Waltham Forest explored how data can be used to increase and widen engagement in arts and culture. The team used wifi access point technology to investigate how to capture data and map user journeys to improve understanding of the borough’s ‘cultural footprint’ whilst acting ethically and engaging the community in decisions about the data collection and use.

Different services, similar experiences

In our report Using open data to deliver public services we looked at eight case studies where open data is helping to deliver public services. The services varied from healthcare to transport to waste management; but all had similarities in terms of organisations, processes, effects and outcomes.

In a similar way, we’ve highlighted the shared experiences of our local government teams to compare the lessons learned through their projects.

Ecosystem conveners

Many councils are responsible for holding data and for the outcome of a public service, but are not directly responsible for delivering that service. These services are delivered by a network of organisations. This was the case for the careers, advice and guidance service in Doncaster City Council and the fuel poverty interventions led by Kent County Council.

In a complex service ecosystem, creating a common understanding of goals and motivations was essential. The Doncaster and Kent councils brought their partners together to understand how they could improve the service using data.

Regular contact supported better organisation collaboration. The teams used techniques such as co-design workshops successfully in the projects to support this approach.

Open innovation

The funds awarded to the local government teams allowed them to explore new ways of working with data, outside of their normal practice.

Waltham Forest Council felt the project inspired colleagues to consider what might be possible. The North Lanarkshire team saw interest from other councils across Scotland who could see the benefits of publishing non-domestic business rate data openly. Doncaster Council felt the project was a ‘vehicle for change’ – opening up conversations about the use of data in the borough.

Collaboration and peer learning is key

Each team was led by a local council or authority with external partners specialising in service design, data, research and technology. Bringing together multidisciplinary teams from a range of organisations helped to build skills, knowledge and perspectives across the team for these and future data projects.

Knowledge of data varied across partners and organisations. All of the teams felt more needed to be done to raise digital and data literacy across organisations so that everyone is able to interact with data and understand the potential benefits and impacts. For example, UrbanTide trained the North Lanarkshire Council team on open data before the project, which was beneficial during the project.

Build on your existing practices

The teams adopted existing tried and tested service design frameworks and open, collaborative ways of working in their projects. It allowed the teams to experiment with data in a familiar environment.

Waltham Forest Council used and adapted its transformation framework for the project. The team discovered new tools, such as the Open Data Institute Data Ethics Canvas, could be embedded in their usual practice.

Improving data infrastructure, availability and access

Kent and Doncaster councils ran a landscape review across the spectrum of closed, shared and open data through interviews with councils and local authorities to form a picture of what data was available.

UrbanTide needed access to non-domestic business rate data held by the North Lanarkshire Council for the project. As a standard data sharing agreement, the process was time consuming and could be simplified if arrangements like this were available in the future.

Doncaster Council discovered data was often held by one person and wasn’t available or did not exist. The team manually created a new dataset detailing the careers information, advice and guidance offered in each school or college to fill this gap and meet their need.

Next steps

We’ll be continuing our work on using data for public service delivery this year. You can see some of the tools we’re working on now – like the Data Ethics Canvas and the Mapping data ecosystems tool – to support teams working on their own projects.

We want your feedback

We would love to hear about your experience of using data for public service delivery and any feedback you have on the tools we’ve created. Contact us here.

Doncaster Council and UsCreates
Kent County Council and UsCreates
London Borough of Waltham Forest, Audience Agency and The Technology Box
North Lanarkshire Council, Snook and Urban Tide

Doncaster Council and UsCreates

What was the challenge?

In Doncaster, careers advice and guidance is fragmented for students who are exploring options after completing their General Certificate of Secondary Education (GCSE) qualifications. Young people largely rely on in-school support from a careers advisor. By talking to young people, Doncaster Council found that in-school careers advice often comes too late for many students, and that employers are not always effectively involved. This means advice on employment and training opportunities for young people aged 16 to 18 is often inadequate.

Although Doncaster Council isn’t responsible for providing careers advice to young people – as with other local authorities in the UK – it is accountable for the outcomes, specifically, for the borough-wide figures of young people not in education, employment, or training (NEET).

This project aimed to democratise careers information, advice and guidance in Doncaster for young people – making information about post-16 career options more freely available and accessible. Grounded in what young people said they would find most useful, the team at Doncaster Council and UsCreates had the chance to create something that could meet students’ needs and potentially reduce the number of young people (aged 16–18) in the NEET category.

How is the team addressing the problem?

The Doncaster team is tackling the problem using a service design approach (an approach for designing customer-centred services) – adapting it to include an ‘open data scan’ – a landscape view of relevant available datasets. Team members began by carrying out user research to identify the careers advice needs of students aged 11 through to 18. They also assessed the availability of careers data from a range of sources, including the Department for Education, local schools and colleges, and from the statutory information provided by the local authority, available via ConnectU2.

The insights from the research led to the creation of six ‘challenge briefs’, which linked together the user research, open data sets and the organisational needs, to gain a clearer sense of the problem. An example of one of these challenges is: ‘How can we use open data to make more learners aware of and engaged in work-based provision and work experience?’.

At a co-design workshop, learners themselves, alongside partners from schools, training providers, The Careers and Enterprise Company, explored ideas to address the challenges. Two solutions were taken forward and developed into prototypes. The first prototype focused on addressing information gaps that may affect decisions for students. The second focused on the experiences of parents and careers advisors when giving advice.

Prototype 1: In this prototype, the dataset of careers information, advice and guidance was combined with data on education and training institutions and available apprenticeships to show students the range of options available. Using this prototype, students could also see the distance and time involved in travelling to a place of work or learning. An additional feature would be to allow students to filter local courses and apprenticeships based on their interests and strengths. In the future, it is hoped this service would support young people to make more informed choices about their education, employment or training pathway.

Prototype 2: The team used data showing school and college scores for young people in the NEET category, tracking it over time to identify changes showing where education, employment or training destinations had not been maintained and therefore had possibly not been the right choice for the learner. The team also developed a new open dataset detailing the careers information, advice and guidance offered in each school or college. In the future it is hoped that publishing this data and making the analysis available will help parents decide on a school for their child, encouraging competition between institutions and incentivising them to improve their careers information, advice and guidance service.

Lessons learned

- Exercise your role as an ecosystem convener The project enabled Doncaster Council to convene partners and to create impact in an area beyond its core service provision. It has mobilised partners around the challenge and given focus and momentum to future work. User research and co-design activities enabled agencies to work together on service delivery.
- Learn from your peers The multidisciplinary team of service designers, local government innovators, policy experts and data scientists brought different perspectives and ways of working. The team learned new skills from each other, building capacity for further open data projects across the council.
- Build data literacy skills across the wider team The biggest challenge was supporting the council and partners to understand the benefits of publishing and using data. Useful data shared within the team was not open, but could have been. This created time pressures on the team who had to manually convert information from a PDF that could be fed into the prototypes. The team challenged these perceptions throughout the project, by championing open data publishing, but recognise there is more to do in promoting the benefits of open data.

Future plans

The longer term ambition is to share the findings with the Doncaster Opportunity Area partnership board. If approved, the team will build the business case for future development. The provision of careers information, advice and guidance was highlighted as a particular challenge for Doncaster.

Exploring how open data can support service redesign provides an opportunity to demonstrate why data should be published and how it can impact on young people’s choices and aspirations.

Kent County Council and UsCreates

What was the challenge?

Across Kent and Medway there are 64,596 households where residents are living in fuel poverty – meaning their income is below the poverty line (taking into account energy costs); and their energy costs are higher than is typical for their household type. Living in a cold home can have both short and long term detrimental effects on wellbeing and physical and mental health.

Working with Kent County Council, the Kent Energy Efficiency Partnership (KEEP) wants to reduce the number of people at risk. Better data and information sharing are among the areas it is exploring to support this. With an estimated 24,300 excess winter deaths in England and Wales in 2015-16, the team was keen to look at how data could better inform its interventions by predicting which areas in Kent were at risk of fuel poverty.

How is the team addressing the problem?

The discovery work focused on three areas:

effective targeting of vulnerable households
learning from past interventions, and
developing ways to estimate the impact of further investment tackling the issue

The team conducted discovery calls with Kent County Council to map available datasets: closed, shared and open. Team members evaluated each dataset to establish if the data could help address the problem. The potential to improve the service would only be achieved by combining datasets from across the spectrum of closed, shared and open data.

In the chosen prototype, the team used the Kent Integrated Dataset (KID), a closed pseudonymised dataset curated by Kent County Council which combines data from 250 health and care organisations across Kent and Medway; and the Wellbeing Acorn dataset, a geodemographic segmentation of the UK’s population – detailing characterics of people and places. Analysis of the Wellbeing Acorn and KID datasets was conducted by the Kent Public Health Observatory team who have permission to access these datasets.

The team looked at segments of the population, along with health and care needs by postcode. A systems modelling approach allowed the team to predict areas where there is increased likelihood of fuel poverty.

Combining closed, shared and open data allowed the team to identify areas at risk at postcode level with an average of 20 residential properties per postcode. This is a significant improvement to the previous analysis which showed geographical areas with approximately 1,600 residents. It has the potential to create a significant improvement to the intervention planning.

Lessons learned

Keep your redesign or intervention focused The fuel poverty intervention is provided by a network of partners with different skills, availability, geographic locations and goals. Developing a refined brief for the project helped to create a shared understanding and to define roles and focus.
Understand the motivations of those who design and deliver the service Regular meetings to understand goals and challenges provided the project team with a clearer set of objectives from funders, strategists, policy officers and frontline workers.
Run data discovery sessions A landscape review of data (closed, shared and open) through interviews with councils and local authorities helped to form a picture of the available data. The people responsible for data processing and data analysis were particularly helpful as well as stewards of datasets that are linked to fuel poverty identifiers, for example data on types of housing.
Keep the user in mind Much of the project focused on strategy and systems which can feel conceptual and disconnected from the end service user. To keep their needs in mind, personas were used throughout the project.
Understand data access and governance Datasets that contain information about people, such as the Kent Integrated Dataset cannot be released openly but are important in helping to better understand fuel poverty. To use this data effectively, systems and governance should be established to allow access to named individuals, enabling effective analysis alongside other datasets.

Plan for the future

The ‘at-risk’ postcodes (from the ACORN tool) will be used by the KEEP group to improve and target fuel poverty intervention work. In the future, Kent County Council hopes to combine further datasets, such as data about gas supplies to develop ‘fuel poverty flags’. These can help service planners to identify residents who are likely to fall into fuel poverty and to commission services that can support better delivery of interventions.

London Borough of Waltham Forest, Audience Agency and The Technology Box

What was the challenge?

Waltham Forest Council has a strong tradition of creativity – it believes that fostering culture is central to improving quality of life in the borough. Inspired by the creative and cultural heritage of the area, the council recognises the social and economic benefits that culture can bring to its 275,800 residents. It believes that it has a key role in increasing cultural opportunities and participation.

In February 2018, Waltham Forest Council was named as London’s first Borough of Culture. The council used this project as an opportunity to explore how data and technology can help to widen and increase participation and improve its service to the community.

How is the team addressing the problem?

Waltham Forest Council explored culture as a service – looking beyond traditional service delivery models to understand the role the council plays through community assets (land and buildings which benefit the community) and its approach to delivering cultural opportunities across the borough.

The project had three main aims:

Gain a richer understanding of engagement with the arts, culture and heritage in Waltham Forest
Develop ideas and approaches to increasing and widening engagement with the arts, culture and heritage
Gain a better understanding of the practical implications and challenges and of designing data-enabled interventions

To achieve these aims, the project had two strands which explored the questions: ‘Who is engaging with the arts, culture and heritage in Waltham Forest?’; and ‘How can engagement be increased in the borough?’.

Strand one (borough wide)

The first strand looked at how Waltham Forest Council could increase and widen engagement in culture by looking at some of the barriers preventing access to culture across Waltham Forest.

To understand current engagement, Waltham Forest Council worked with the Audience Agency to analyse existing open, shared and closed data sources. This included data under an Open Government license (OGL), such as demographic information from the UK census, and shared data, such as modelled consumer behaviours data from Experian, alongside previous Audience Agency studies.

The team developed cultural profiles and plotted the data onto maps of the borough so that engagement could be seen by location. This allowed the team to spot new engagement insights and patterns. Interviews with residents explored some of the barriers to engagement with culture in depth. It also showed that Vestry House Museum – a prized historic building situated in the heart of Walthamstow Village which holds collections and displays of local history and domestic life – is located between areas of both low and relatively high levels of engagement, and this was seen as a potential venue to be used to help expand the culture programme in the borough.

Strand two (community assets)

This strand explored the collection of data through wifi access point technology within Vestry House Museum. The museum celebrates Waltham Forest’s heritage through a range of exhibitions and events. The building has an events space and large garden; it also enjoys a central location in Walthamstow Village, a few streets away from the main high street. However, compared to the William Morris Gallery – a very successful cultural venue in the borough – Vestry House Museum is currently not realising its full potential.

To explore how Vestry House Museum could be expanded and better used, wifi access points were installed in the museum to capture data on visitor journeys, alongside a new wifi landing page. Users of the wifi service were informed how the data would be used and asked to respond to a number of questions; this included questions on their age and the reason for their visit. This information enabled the project team to identify common cultural motivations for this specific venue. It also allowed them to gauge visitor numbers and patterns, and to map out dwell times and visitor journeys, to inform how Waltham Forest Council could create better customer journeys for people who use the space.

With an abundance of data and new information, Waltham Forest Council used ‘ideation’ techniques, specifically one they called ‘smashing light bulbs’. Ideation is a creative process to rapidly generate and develop ideas allowing the team to look at how to improve the service. The council applied this technique to around 30 ideas to widen participation in the borough, and at Vestry House Museum specifically.

What was the impact?

The initial insights drawn from the data collected at Vestry House Museum show that there is a relatively high footfall of passers-by, as it is located in a busy residential location, close to a popular London tube station. The developed prototype ideas used this information to focus on how the borough could increase the conversion rate of people walking past and unique venue visitor numbers. This includes developing new signage on-site and at every stage within common visitor journeys to encourage people to visit.

The discovery work will continue to inform the new Borough of Culture Programme outputs and objectives, which include the redevelopment of Vestry House Museum as a physical space and as an alternative events space. It will also help develop new initiatives across the borough over the next few months, including a scheme enabling young people to volunteer at cultural events, as well as a new cultural cycle and walking map of the borough.

Waltham Forest Council found that the workshop on the Open Data Institute’s (ODI’s) Data Ethics Canvas had significant benefits for the project and plans to use it within the discovery stage of future projects. Crucially, the council believes the project has helped to develop a council-wide approach to implementing ethical and data-enabled service design and delivery.

Lessons learned

Using existing frameworks Using existing design or transformation frameworks supports the agile and open approach needed to innovate with data for service delivery. The team learned how to expand its use of data within the existing Waltham Forest design practice, building up knowledge for future projects. Working in a consortium brought different skills, expertise and perspectives to the project.
Draw it out The project inspired others in the council, and raised the stakes for service innovation and what is technically possible. To articulate the use of data to other stakeholders inside and outside of the council, Waltham Forest Council developed a data journey map.
Consider the ethics of your data use There were significant ethical and legal considerations around the use of data collected from individuals. Waltham Forest worked with their legal teams on the wifi terms and conditions and with the ODI on the ethical use of data. The team used the ODI's Data Ethics Canvas to guide discussions and make decisions around its approach to collecting, analysing and publishing data.
Build data literacy across the organisation For many staff, open data is unfamiliar. Explaining open data was key to engaging people to talk about the ethics involved in data collection and publishing. Throughout the project, the team worked with colleagues from across very different departments: legal, insight and intelligence, business intelligence, culture, digital and ICT. The breadth of interest stems from the realisation that data really can improve Waltham Forest Council’s services.

Plan for the future

Shortly after the project closed, Waltham Forest was chosen to be London’s first Borough of Culture. Waltham Forest Council is looking forward to implementing the use of open data within the discovery stage of cultural projects for this programme and is hoping to use these projects as an example of how open data can inform successful service design or delivery of other service areas in the future.

Alongside this, the project team will work with other services and teams across the council to implement the ODI Data Ethics Canvas into their practice. The canvas will be used to help establish good practice ahead of formal regulation, such as the implementation of the new General Data Protection Regulation in May 2018.

North Lanarkshire Council, Snook and Urban Tide

What was the challenge?

North Lanarkshire Council receives high volumes of Freedom of Information (FOI) requests, and calls and emails to its customer service team about non-domestic business rates. These are taxes charged on buildings being used for non-residential purposes, such as shops, offices, warehouses and factories. The council manually answers these queries, getting the required information from large internal datasets. The council publishes a cut-down version of the dataset monthly to illustrate new ratepayers and outstanding balances. It is a time-consuming and costly process and often the same information is requested multiple times.

With the belief that ‘every FOI request is a service failure’, North Lanarkshire Council's goal was to publish non-personal, non-domestic business rate data so that customers could access the information quickly and easily, with the ultimate goals of reducing the number of requests to the customer service centre, increasing transparency and providing residents with information about their local area.

How is the team addressing the problem?

North Lanarkshire Council is on a digital and cultural transformation journey with a specific aim ‘to improve economic opportunities and outcomes – for all’. Realising the value of data and establishing routine processes and procedures to reuse data is central to enabling this aim.

The team (North Lanarkshire Council working with Snook and Urban Tide) aimed to publish the full non-domestic rates dataset, with personal data removed, for example sole trader data. The existing service does not publish this dataset due to the time and complexity in removing personal data. This will lead to the publication of over 6,500 data entries, allowing for a comprehensive picture of rates being made available as open data. Further de-identification techniques have been tested that could lead to the publication of the full dataset.

The team used the USMART platform to process and create the large open dataset. Using data in a sandbox environment meant that data could be manipulated quickly and de-identification of data could be tested. Putting in place the appropriate data sharing and licence conditions to enable data analysis between organisations was essential, but was a challenging process. The council proactively engaged with the legal departments to create agreements.

The redesigned service will improve the accuracy of the whole non-domestic rates dataset by automatically processing existing open data sources including datasets from: Companies House; the Food Standards Agency Hygiene rating scheme; the Office of the Scottish Charities Regulator (OSCR); the Charities Regulator for England and Wales; and Google Business API. Early findings indicate that these datasets can be used to understand businesses not currently paying non-domestic rates.

Lessons learned

Understanding data access Requesting access to data and creating data sharing agreements takes time and should be built into your project at the start.
Find creative ways to get user feedback There was a limited response from service users who were asked to take part in research which was designed to understand the challenges faced. The team used detailed FOI requests to build a picture of needs.
Use existing service design toolkits Team members used service design templates and techniques that they were familiar with and existing open data resources such as the Scottish Government Open Data Resource Pack. The toolkit developed by the team is designed to be read alongside this resource. They found lots of service design resources and techniques were transferable to open data projects and that reusing materials reduced the number of new tools required.
Build data literacy across the team UrbanTide had trained North Lanarkshire Council on open data practices before the project, which meant the team was more comfortable with the topic and championed data-enabled services across the council.

Plan for the future

The service is expected to reduce the burden of publishing datasets and also provide significant additional information to businesses about the rates landscape across North Lanarkshire.

Significant expansion is also proposed to improve the accuracy of detection of missing non-domestic rates payers and to use machine-learning algorithms to detect payments that are higher or lower than they should be. More accurate collection of rates will help inform policy redevelopment and more efficient service delivery chains that can increase the level of non-domestic rates collected, enabling reinvestment into public service delivery.