Interpreting and using performance data to continuously improve services.
Performance analysis involves measuring and interpreting data to see where and how a service can be improved. Refer to the Data skills group for more detail on data-related skills.
Some relevant roles: data scientists, performance analysts
Managing and developing analytics architecture
Supporting the product or service manager with accurate and actionable data, by managing and developing analytics instrumentation.
Tying analytics to the service
Participating in the development of the vision for data/analytics architecture and workflow across the service or product (including instrumentation, tagging and user management).
Representing the user
Using data and research analysis to represent the user in product and design discussions.
Understanding business processes
Linking to business process
Having an understanding of:
- the business processes behind services you analyse
- how changes to the technology can support and improve the process for users
Defining the data
- knowing what data sources to use, who needs what data and in what format
- understanding how digital data can fit in with agile business processes to drive data-based decision making in the service
Supporting development teams to capture data
Collecting the data
- identifying data collection points
- supporting development teams to instrument transactions and content appropriately, in order to capture user journeys and completion rates
- designing a measurement plan
- defining tagging
- creating and validating digital analytics tags
- setting up custom dimensions and metrics where required
Using data to understand content
Considering all data
- collating and using data to be aware of the online and offline context of the service
- considering all types of data, including offline data such as from a call centre
Understanding the wider context
This involves understanding:
- user language
- search engine optimisation principles
- user journeys through GOV.UK content
- competing and misleading websites
Measuring how your service is performing, so you can make sure that your service continues to meet user needs in a cost-effective and efficient way
Following a process of continual iteration, measurement and analytics in order to monitor and improve information and services. This involves:
- understanding user needs
- deciding what to measure
- installing and configuring platforms
- establishing a baseline
- aggregate data
- analysing and visualising data
Defining key performance indicators
Thinking about other useful key performance indicators (KPIs) specific to your service that will help you measure and improve its performance. Measure these in addition to the 4 core KPIs:
- cost per transaction
- user satisfaction
- completion rate
- digital take-up
Creating data supply
Using analytics packages
Using digital analytics packages such as Google Analytics for implementation, administration, visualisation and analysis to:
- create an account
- set up data views
- create dashboards
- schedule reports
- set up custom reports and custom alerts
- manage user permissions so that the team can reliably and easily access appropriate data
Refining the data
- refining the data output using spreadsheets, pivot tables and other tools such as Google Refine and Tableau
- using these tools to collect, interpret and present data for analysis, enabling user-centric design decisions
- working with digital analytics, financial and user data
Using digital analytics application programming interfaces (APIs) to extract and manipulate data
Providing business insight from analysis
- providing insights from data
- identifying and telling actionable stories and business insights from the data to stakeholders and the team, to inform user-centric design decisions
Using analytics tools
Analysing data collected through analytics tools such as referrals, sessions, page views, exits and bounce rates
Applying tools to user experience
Create events, filters and segments, goals and funnel visualisations using relevant digital analytics tools to identify the areas where users are experiencing problems.
Working with user research data, and combining it with performance analysis insights to better support delivery of products.
- synthesising other data sources (eg call centre data)
- working with user researchers and data scientists to tell richer, more actionable stories based on evidence
Advances in analytics technologies
Identifying opportunities to enhance data collection and analysis.
Keeping up to date with advances in digital analytics tools and data manipulation products.
A/B and multivariate testing
- so the effects of changes to your live service can be tested
- to support iterative improvements
Data security and privacy
Having an understanding of data security and privacy, including concepts such as privacy-preserving data mining, data security, data provenance.
The Service Design Manual includes an overview of the digital performance analyst role, including competencies and skills required.
The Digital Analytics Association is a not-for-profit, volunteer-powered association that offers educational material, community forums and research materials related to digital analytics.
Google offers free online certification and training courses on Google Analytics.
Brandon Foltz has created an online video that introduces simple linear regression.
The STEPS website offers a free online guide to setting up and testing hypotheses.
Web Analytics Wednesday is a free networking event for web analytics professionals in London.
Digital analytics for UK government services shares a list of interesting links aimed at digital analysts working on UK Government services and exemplar projects.
Analytics talk is an example of a blog on digital analytics for business, here with a particular focus on using Google Analytics.
The Service Design Manual includes guidance on measuring cost per transaction.
Analytics Academy is a Google community offering free lessons and tests on analytics.