Digital and Data Principles focus on establishing high-level guidelines for how information flows through your organisation. These principles create a shared language for data across departments, guiding frontline staff in understanding what good data practices look like. Rather than detailed policies, these principles foster effective data use, sharing, and interoperability within and between departments. They address fundamental questions about data collection, quality, access, and the ethical use of resident information.
| Standard | Explanation |
|---|---|
| Understand users and their needs | Look at the full context to understand what the user is trying to achieve. Understand the different costs created for a user of data. These might include learning costs (e.g. understanding what data exists and how to use it), compliance costs (e.g. governance processes that allow access to data) and emotional costs (e.g. the frustration that arises from not being able to access the data they need). Why it’s important Understanding as much of the users’ context as possible gives you the best chance of meeting users’ needs in a simple and cost-effective way. Focusing on the user and the problem they’re trying to solve - rather than a particular solution - often means that you learn unexpected things about their needs. The real problem might not be the one you originally thought needed solving. Testing your assumptions early means you can learn what works sooner. What it means Teams should learn as much as possible about the problem users need them to solve by: - Doing user research to understand what users need - and, where relevant, secondary research and analysis. - Building quick, throwaway prototypes to test their hypotheses. - Using web analytics and other data that is available to get users’ feedback. |
| Enable joined-up, real-time services | Why it’s important Real-time services help the public get the help they need, and understand their rights and responsibilities faster. They reduce the amount of form filling and can speed up decisions. Cities are complex so the totality of needs of a user’s needs are rarely met by a single department. Using data to join up services means more of a user’s needs can be met more of the time. What it means - Services and analysts need real-time access to data - Using common identifiers and standards across datasets |
| Support data-driven decision making | Why it’s important The lack of good quality data can result in slow decision making and a lack of accountability. Data can help decision makers make better decisions that, in turn, lead to better outcomes for the public. Some decisions may be automated or semi-automated if appropriate safeguards are in place. What it means - Decision makers have access to the information they need. - The city has appropriately skilled analysts. - There is an audit trail of the data and software used to make informed decisions. - The process is fully replicable, independent of the technology used. |
| Collect the right data the first time | Why it’s important If a service collects data in formats that are bespoke, proprietary or vary between channels (for example different data is collected via paper forms, web and text channels), it has implications for the operation of the service. It also means the data is harder to use for other purposes. Increasing the quality of data at the source can improve downstream outcomes. What it means - Lowering the cost of services collecting data digitally by operating ‘common components’ for things like web forms and text messaging. - Collecting the same data across channels (analog and digital). - Setting design standards for form design. - Using agreed data standards and identifiers. |
| Make it simple to find the right data | Why it’s important If service teams and analysts don’t know what data the city holds, can’t access it, or have to spend time reformatting it, then the needs of users go unmet. What it means - There is a public catalogue of the foundational datasets that the city holds along with a description of each field. - Users know where the catalogue exists and can access it depending on their needs. |
| Data access, not data sharing | Why it’s important ‘Data sharing’ is, essentially, copying-and-pasting data from one database to another. The data gets duplicated, which creates redundancy. It can easily get out of sync, which creates problems for users and the teams operating services. With ‘data access’, data is stored once and accessed via APIs by different users. What it means - Data is managed as a common public resource. - There are processes, technology and open standards for accessing data. - Services and analysts can request access to data with the reasonable expectation that their request will be granted. |
| Use common identifiers, data standards and design patterns | Why it’s important If data is stored using different identifiers or formats (for example using multiple geospatial standards), data needs to be transformed before it can be used. Using proprietary identifiers and standards can limit the reuse of data. Using multiple design patterns for presenting data means users have to learn different interfaces, which can mean that accessibility requirements are inconsistently met. What it means - Engaging in city-wide processes to agree on common data standards and identifiers. - Making explicit decisions not to link certain datasets. |
| Protect the safety, privacy and wellbeing of residents | Why it’s important How data is collected and used can have implications for the safety, privacy and wellbeing of residents. The full range of risks may not be apparent to analysts or service teams and may change over time. What it means - Assessing risks throughout the development process using gender, equity and social inclusion principles. - Providing feedback systems and ‘early warning systems’ where concerns can be reported and acted on. |
| Be transparent about how data is used | Why it’s important Data is inherently opaque. It should be clear to decisions makers what data they are using to make decisions. It should be clear to residents how data about them is used. What it means - Citing the sources of data used alongside analyses and visualisations. - Publishing information about what datasets different services use. - Publishing the software code for analyses and automated decisions. - Explaining to users of services how data about them will be used at the point of use. - Understanding that the city is the custodian of all data that ultimately belongs to residents. |
| Custodianship and Accountability are Clear | Why it’s important If data doesn’t have a custodian who can prioritise the needs of all users of the data, then the full value of data is not realised. What it means - Each foundational dataset has a clear custodian with an obligation to manage the data as a common resource - Publishing the name of the department with custodianship of each foundational dataset |
| Make sure it works for everybody | Why it’s important Services, reports and analyses (“Outputs”) that use data must work for everyone who needs to use them including differently-abled persons and marginalised groups. Outputs that are inclusive, and accessible are better for everyone. What it means - Meet accessibility standards, including both online and offline parts. - Allow data to be collected and presented in multiple languages. - Allow data to be collected and presented in multiple formats. - Carry out research with participants who represent the potential audience for the service, including people with access needs. |
| Work in multidisciplinary teams | Why it’s important You’ll need a team made up of people with a diverse mix of skills and expertise. It’s important that people who are involved in decision making are part of the team, so they’re accountable to the team - and the team as a whole can respond quickly to what they learn about users and their needs. What it means - A team that’s appropriate to achieve the desired outcomes. - The work is the primary focus of the team. - Include people on the team with expertise in how services are delivered across all the relevant offline channels, and the back-end systems the service will need to integrate with. - Provide the team with access to the specialist expertise it needs (for example legal, policy or industry-specific analysis - from inside or outside the organisation). - If the team is working with contractors and outside suppliers, make sure it’s on a sustainable basis. |
| Use agile and open ways of working | Why it’s important Using agile methods means getting outputs in front of real users as soon as possible. Based on real users feedback, we can observe and generate data on how they use it before iterating the output based on what we’ve learnt. Because you’re not specifying everything up front before you’ve developed an understanding of what users need, agile methods reduce the risk of delivering something that does not meet the needs of users and/or reduces the administrative burden for them. What it means - Not specifying everything up-front. Learn while doing. - Being clear about the type of value you are trying to create (this is normally some combination of meeting the needs of users, reducing administrative burdens, meeting a policy intent, and capacity building). - Working in one or fortnightly sprints where the aim is to deliver real-world value at the end of each sprint. - Prioritising work based on what you are learning. - Having regular showcases that senior stakeholders attend. |
| Define what success looks like and publish performance data | Why it’s important Digital and data work can be seen as ‘innovative’ on its own terms without having any real-world impact. To avoid this, the desired outcomes for data work should be documented (for example, increased uptake of services, less water leaks, etc). Those outcomes may change as you learn more through iterative development and user research. What it means - Documenting what the desired outcomes are for data-related work. - Restating outcomes and vision alongside analyses. |
| Choose the right tools and technology | Why it’s important Forcing service teams and analysts to use a limited set of tools confines their ability to meet the needs of users (e.g. if the only tool you have is a dashboard, then the answer to every problem becomes a dashboard). What it means - Teams should be free to choose the most appropriate tools for the task at hand. |
| Work in the open | Why it’s important Working in the open is a form of governance. Because modern agile ways of working are organised around iterative delivery and incremental change, governance processes need to be too. It also means that people working on data projects in different parts of the city and beyond can find each other without having to navigate management hierarchies. What it means - Publishing weeknotes. - Holding open ‘showcases’ with senior stakeholders that show recent changes. - Publishing software code and example screenshots (within the confines of POPIA e.g. with any personal information removed). |