Data and Analytics Service Standard
City of Tempe Data and Analytics Service Standard
The Data and Analytics Service Standard sets out guiding principles to help the City of Tempe create, manage, and improve transparent, reliable, and high-quality data and analytics services and products that meet the needs of external and internal users.
Data and Analytics Service Guidelines
1. Publish Data and Analytics Services with a Purpose
Tempe prioritizes data, data services and analytics services that are a high priority to our community, including residents, businesses, and internal users. Prioritizing data and analytics services that have the biggest impact for the city and community allows more efficient uses of city resources to provide data products with a clear purpose that are developed to meet user needs.
Why it’s important
Sharing content for the sole purpose of sharing content makes it harder for people to discover relevant and impactful data products.
The city has limited resources to create data and analytics services. It needs to prioritize its efforts towards services that will have the most impact for the city and its communities.
Having a clear idea of purpose helps to identify relevant users and their needs; services that are built for everyone often end up satisfying no one.
What it means
Service teams should:
Seek to understand how the service contribute to wider city and community priorities (such as informing policy making, supporting operation decisions, or nudging public behavior).
Publicly document the purpose of the service using clear explanations accessible to both technical and non-technical readers.
Document and share how different parts of city government or the wider public sector contribute to and benefit from the service.
Follow the open data publishing and review process.
2. Understand Users and Their Needs
Tempe considers the problems users are trying to solve or the information they are trying to gather when developing data or analytics services. We look beyond just the interaction with the service, by considering the full context of the ways end users may need or want to interact with the data and analytics services based on what they are trying to achieve or understand, and their tools and capabilities.
For data services, we aim to understand both the needs of the direct users of the data service and the end users of the products and services the direct users will build. For analytics services, we provide access to the data service(s) that drive them.
Why it’s important
Understanding as much of the context and problem as possible provides the best chance of meeting user needs in a simple and cost-effective way. Focusing on the problem they’re trying to solve, rather than a specific solution, leads to deeper insights and more open conversations. The real problem might not be the one originally identified, so testing assumptions early and often reduces the risk of building the wrong data and tools.
What it means
The team should learn as much as possible about the problem the user is trying to solve or the information they are trying to gather by:
Research users to better understand their needs (such as developing user personas to guide development).
Focus on the actions or tasks that help users solve their problem or gather relevant data and insights.
Conduct community surveys or focus groups to enhance understanding of a problem.
Ask questions about what data needs to be shared, including level of aggregation/disaggregation, spatial granularity, and timeliness and how it should be shared (such as formats or licensing) considering both user needs and protecting sensitive data (see 8. Create Services that Protect Sensitive Data).
Use web analytics and other available data (such as data requests or analytics services) to increase understanding of the problem and whether/when services are being used.
Regularly review public records requests to identify data services that might meet the current needs to the community and improve efficiency for the city.
Build quick prototypes to test whether solution addresses the problems and review with team and end users or sponsor.
Test a beta version of the service when ready.
3. Multidisciplinary Teams
Tempe’s Data Governance Committee is responsible for creating, implementing, and maintaining the Data and Analytics Service Standard in an ongoing manner to ensure quality data and analytics services for use internally and externally. The group is made up of a diverse group of stakeholders from across the city, including representation from Legal and Diversity as well as end users from City Departments. Additional people with skills or knowledge relevant to a problem may be brought in to support specific projects.
Why it’s important
A diverse set of skills, expertise and experience are important for collaboration, decision making and an inclusive approach to developing data and analytics services and tools. Engaging these stakeholder supports broad awareness of the data and analytics services available, opportunities to identify new focus areas and an ability to respond quickly to what they learn about users and their needs.
Teams may be shaped in part by the problem and what they are doing at that point in the project. A technical resource might not be needed until after the initial problem assessment, for example.
What it means
Services should:
Be built by a multidisciplinary team that includes technical and non-technical experts that will enhance the usefulness and usability of the service to a wide variety of users. Non-technical users could include subject matter experts, legal, external partners or leadership.
Build services that integrate with existing systems or support standard publishing techniques to increase sustainability, data availability and useability.
Build services that provide data or analytics that can support evidence-based decision making.
4. Solve a Whole Problem for Users
Tempe works with other teams, organizations, or external partners when necessary to create services and data products that solve a whole problem for users by creating integrated solutions.
Why it’s important
Services that do not work well with other related services make it hard for users to do what they need to. For example, working out how to combine datasets from different sources or compare data or analysis with differently defined geospatial areas.
Integrating data and analytics services with related services, such as information pages on a website or transactional services, can make data and analytics easier to find. However, we should be careful not to build big, complicated services that are not intuitive to use because they try to do too much. It can be hard for data users to understand what data means just by looking at a table or documentation. Analytics services provide a human-friendly interface that helps data users understand the information it contains.
Using data and analytics effectively is not just a matter of having access to it. Users need to understand its context, so they know how much to trust it. They need tools and skills to process data and technical support when things go wrong. They are often able to make better use of data when they have peer support.
What it means
Service teams should:
Scope services appropriately to meet user needs so that they are not too narrow or too broad.
Be aware of the service and able to explain how the service will join up with other things such as tools or applications/datasets.
Use existing project scoping processes to assign responsibility for input, creation, testing, etc.
Build analytics services on top of data services, using the same APIs that are available to data service users.
Work with other teams and organizations where that’s necessary to solve a whole problem for users, including engaging with how the data the service uses is collected and stewarded, so we can influence it to improve data quality and its utility to the service users.
Avoid excluding any groups within the audience they are intended to serve (e.g., accessibility standards, digital skills, difficulty level) and carry out research with participants who represent the potential audience for the service, including people with access needs.
5. Inspire Creative Uses of Data with Internal and External Data Community
Tempe encourages evidence-based decision making and creative uses of data to meet community needs or solve a problem. We are working to build a community of data users that includes those inside and outside of government, with the goal of creating new users and increasing demand for current and new services.
Why it’s important
Most people don’t know how to use data and evidence to inform their decisions, so they might not use the service even if it would be useful to them, and improve the city, simply because they don’t know what is possible, aren’t aware that the data are available or could be requested, don’t know to look for it or aren’t sure how to use what is available.
Seeing how data can be used helps to encourage developers and analysts to use it to generate new insights and new services that support city workers as well as those who live, work and visit the city. Often those inspiring uses will come from the community of other users.
Encouraging the development of new products with data can lead to new or expanding businesses which create jobs and economic growth.
What it means
Service teams should:
Draw on internal and external expertise to build a library of insights and stories that inspire and support future data users.
Actively engage the city data community with the service, such as through workshops, hackathons, or innovation prizes.
Enable the discovery of the service and the data through good search engine optimization and clear data descriptions and tags.
6. Make the Service Consistent and Simple to Use
Tempe strives to build services that are simple, intuitive, and comprehensible without oversimplifying things. We test services and analytics tools with users to make sure it works for them.
Why it’s important
People expect services to just work, and city data and analytics services should be no exception.
Services that don’t work can impact trust as well as increase costs and resource use. Making things more complicated than they need to be undermines trust and may reduce usability. Simple tools are often easier to develop which can increase efficiency and usability. Some data and analytics needs can be satisfied with a spreadsheet.
What it means
Service teams should:
Make sure the service helps the user to do the thing they need to do as simply as possible - so that people succeed first time, with the minimum of help.
Test all the parts of the service that the user interacts with, including parts using third-party platforms.
Design the service to work online with a range of devices and analytics tools.
Services should also provide users with a consistent experience from start to finish, for example: making it possible to script the retrieval of data from data services using APIs and file naming conventions.
7. Make the Service Reliable
Minimize service downtime and have a plan to deal with it when it does happen. Fill data gaps and improve data quality over time.
Why it’s important
Users expect to be able to use online services 24 hours a day, 365 days a year and data users expect the data we provide to be accurate, complete, and consistent.
Digital services built by third parties may depend on our data services; downtime of our service affects their end users.
Many users have limited choice over how and when they access analytics services. If a service is unavailable or slow, it can mean those users aren’t able to get the help they need.
What it means
Service teams should:
Maximize data quality in the data used by the service; where it is lacking, publish metadata (documentation) about known issues.
Provide mechanisms for making corrections to data; this can help improve data quality over time for everyone.
Document and maintain architectures to understand the interdependencies of maintenance as well as technology or data structure changes. Develop and follow a change management protocol to review changes with relevant teams.
Carry out quality assurance testing regularly, both on the service and on the data that underpins it.
Test the service in an environment that’s as similar to live as possible and the data with visualization tools that help to highlight errors.
Have appropriate monitoring in place, together with a proportionate, sustainable plan to respond to problems identified by monitoring (given the impact of problems on users and on government).
Have appropriate mechanisms in place to warn users of data services about planned downtime and keep them informed about the status of unplanned service interruptions.
Actively work towards fixing any organizational or contractual issues which make it difficult to maximize availability or data quality (for example, by agreeing a common set of languages, tools, and ways of working for technical staff).
Give plenty of warning to data users before deprecating old versions of the service.
Turn off services and features that are no longer used or supported.
8. Create Services that Protect Sensitive Data
Tempe’s Data Governance Committee and multidisciplinary team evaluates what data the service will be collecting, storing and providing and ensures it is handled legally, ethically and equitably. Using guidance and information provided by the Data Governance Committee and Information Security and Privacy Office, the team works to understand how we classify the data, the organization’s legal responsibilities, and ethical and security risks associated with the service.
Why it’s important
Data and analytics services often hold sensitive data about people, organizations, and the city environment. City government has a legal duty to protect this data and an ethical duty to ensure it is used in ways that advance equity. Failing in that duty would undermine public trust in city government.
What it means
Service teams should:
Actively identify security and privacy threats to the service and potential adverse impacts of providing access to data and information, and have a robust, proportionate approach to securing and governing access to data.
Carefully assess any data acquired from third parties to ensure it has been collected and shared legally, ethically, and equitably.
Engage with a diverse set of stakeholders, particularly those representing communities reflected by the data or affected by actions that are informed by it, to respect the people represented in the data and to understand concerns. City governments must ensure data has equitable impacts. Listening to the concerns and reports from the community, particularly those who are under-represented, can help avoid harms.
Have a plan and budget that allows us to manage security and mitigate harms during the life of the service (for example by responding to new threats and unanticipated consequences, putting controls in place, anonymizing data, and applying security patches to software)
Process users’ and data subjects’ personal information in a way that’s secure and respects their privacy and take similar care over sensitive information about organizations, communities, or the environment.
Where appropriate, aggregate data or remove certain attributes so that datasets cannot be combined to uncover sensitive data.
Follow Tempe’s Sensitive Regulated Data: Permitted and Restricted Use standard (https://gis.tempe.gov/ordinances/sensitive-regulated-data-policy.pdf) and complete the Security and Privacy Worksheet (https://hub.arcgis.com/documents/tempegov::security-and-privacy-worksheet/explore) prior to developing a service.
9. Choose the Right Tools and Technology for User-Centered Services
Tempe focuses on choosing tools and technology that let us create a high-quality service in a cost-effective way. Good tool and technology decisions minimizes service downtime and helps prevent the need to change direction in the future.
Why it’s important
Technology decisions represent a significant investment in time, resources, and money. Choices impact the ability to create, iterate and operate services in a sustainable way. For data services, the technology we choose can determine the data standards and interfaces we provide, and that data users build around.
What it means
When considering data and API standards, technical architecture, choice of programming languages, development toolchain and other technology choices, service teams should:
Use automation (including data collection) where possible.
Understand total cost of ownership of the technology and preserve the ability to make different choices in the future.
Develop a common set of languages, tools, and ways of working for technical staff including standard tools for developing and sharing services or supporting analytics.
10. Define What Success Looks Like and Publish Performance Data
Tempe will develop metrics for data, services, or data products to measure success as well as opportunities for improvement. Tempe will collect and use data from multiple sources both online and offline.
The value of data and analytics stems from how they change people’s decisions and actions, not from numbers of downloads or users. We will iterate and improve metrics and performance data collection practices as we learn more about user needs.
Why it’s important
Defining what good looks like and identifying appropriate metrics means that we’ll know whether the service is solving the problem it’s meant to solve. Collecting the right performance data means we’ll be alerted to potential problems with the service as well as whether a change results in the expected effect.
What it means
Service teams should:
Identify metrics which will indicate how well the service is solving the problem it’s meant to solve, and track performance against them.
Actively seek feedback to identify impacts – positive and negative – that are not being measured and improve metrics where necessary.
Use performance data to make decisions about how to fix problems and improve the service.
Have a sustainable plan to monitor performance and respond to identified problems.
Publish performance metric data for key services.
Next Steps
The Data Service Standard should be consulted when there are:
Updates to services and products.
New services and products that are in queue for creation or in testing.
Older data that is being used in an analytic function for the first time.
Contributing to the Standard
The Data Service Standard is a living document and will be reviewed periodically for updates and as new information or user recommendations are submitted to the Data Governance Committee. All feedback to this Data Service Standard can be submitted to the data@tempe.gov email.
Last updated