Design and Manage Data for Sharing

• When collecting data, make sure that unnecessary personal information is not captured. Only ever capture personal information if it is directly needed for customer service and if the customer has given consent. 
• Make sure all your datasets have an owner who is responsible for data management and who can make decisions about who can access and use them. 
• Help people understand the lineage, context, background and reporting associated with your data so that they can trust the data, and can use it appropriately. 
• Look for opportunities to share current or real time data. Publishing live, real-time feeds can be incredibly valuable for some types of data. Make sure you use dates or timestamps to help users identify the age and relevance of your data. 

Prioritise high value data for sharing and release

All agencies should prioritise the sharing of data that: 

• Describes the performance or use of government services – this can help other agencies with planning and service provision and helps the public to make informed choices. Releasing this data supports accountability, transparency and better planning. 
• Is used regularly by a number of business areas or agencies – The data may support joined-up services for a better customer experience. It may reduce duplication and administrative costs. Consider making data available through an API. 
• Provides a foundation for decisions – If data establishes a basis for planning and evaluation by other agencies, individuals or organisations, there is likely to be value in the ongoing release or regular updates to the data. For example, population, household and dwelling projections. Similarly, if data can be layered or combined with many other types of data to generate insights, it is likely to have high value. 
• Represents a substantial investment of public resources – Data collected or produced during a research or evaluation project will often have high value for other agencies and the community. For example, data from public surveys, cost-benefit analyses, or environmental impact assessments can be valuable to share. 
• Supports transparency and accountability – Releasing data can promote open discussion of public affairs or contribute to a positive and informed debate on issues of public importance.

• Make sure any datasets that contain personal or sensitive information are identified and well protected. 
• When sharing data, make sure personal and sensitive information is protected. 

Practice data minimisation

Unless strict safeguards have been applied, personally identifying information cannot be released or shared. Depending on your business processes you should: 

• Ensure individuals only provide the minimum amount of personal information needed for a specific transaction or interaction 
• Enable the public to interact with your institution anonymously or pseudo-anonymously 
• Generalise data being collected (for example, ask for an age range rather than specific age) 
• Provide clear notices at data collection points when the provision of personal information is not required (for example, in community engagement or feedback processes) 
• Ensure contracts with third party service providers require them to minimise the personally identifying data they are collecting and managing. 

• Engage with your customers around data collection, management and use and make sure full consent is obtained. 
• Talk to the community, research sector and industry about the type of data they want your agency to release, and how they want this data released. 
• Make sure your is data useable by providing good metadata that uses plain English. 
• Provide data dictionaries and other documentation that explains your data. People in other parts of your organisation or other parts of government aren’t going to understand your codes, abbreviations or specific terms are. 
• For geospatial datasets, document which geographic coordinate system, map projection or datum the data is in. If the geospatial data was transformed from another datum, the transformation method should also be noted. 

To build data that helps deliver customer value and that establishes trust with the community in government use of data, it’s important to engage with customers. Customer engagement about data should: 

• Build trust and collaboration 
• Inform how government collects and uses data 
• Inspire and empower consumers to use government data 
• Grow awareness of the use, benefits and opportunities of government data use 
• Identify priorities and data demands for different community audiences 
• Target specific challenges and attempt to identify appropriate solutions 

Customers feedback should also drive government data release priorities, strategic data priorities, data use, real-time release, data visualisation and innovative data partnerships. 

• Drive data sharing across government, to help improve service outcomes and government decision making. 
• Make sure your data is machine readable, in a format that makes it easy to process and use. Making data available through APIs automates the production of up to data machine-readable data. 
• Use common standards for your data wherever possible, so that it’s easy for your agency and others to know exactly what your data refers to, and how to compare your data to other datasets. 
• Apply data quality statements to all your data, so that people know how your data was created, what it can help them to understand, and any limitations it may have. Use the Data Quality Reporting Tool to easily generate one of these statements for your data, and publish the statement generated alongside your data listing. 
• Look for opportunities to leverage data from Data.NSW in your own business environment 
• Work with Government Information (Public Access) Act teams, and analyse common customer service enquiries, to identify data that can be routinely released.  
• Examine information already published on your agency website to determine if the data published in PDF in annual reports, budget papers, reports, grant funding etc can also be released as machine readable data.

• Be open to feedback and listen to users of your data, who may have advice on how to improve your data collection and management. 
• Look for opportunities to share or release more data. This could involve requirements in all contracts for service providers to share or release the data they collect, ensuring all contracts state that any data is owned by the government and must be returned to government custody at the conclusion of the contract, or making more data available in accessible, machine readable form. 
• Corporate reluctance to share and release data is often driven by quality concerns. Documented data governance frameworks, data quality requirements and regular monitoring of data processes can build maturity and create a culture where all staff are aware of their data responsibilities and how their work contributes to customer outcomes. 
• Once your data is available, keep it updated. If your data changes often, the best way to ensure it’s refreshed is to put automated processes in place to publish regular updates. Consider publishing your data via an API to make access to data automatic. 

Be open and transparent with the community and release as much open data as possible. 

Communicate clearly and frequently with the community about how your agency is using data, and how data is contributing to better customer services and community outcomes.