The public sector is rapidly evolving, is your people analytics strategy fit for purpose and can it meet the increasing demands of a modern public sector?

In this blog, we will highlight the unique challenges that public sector stakeholders face when implementing a people analytics strategy. In light of those challenges, we will then outline how to best design and implement a modern people analytics strategy in the public service.

When it comes to people analytics, the public sector faces a number of unique challenges; 

• The public sector is the largest and most complex workforce of any employer in Australia. A workforce that bridges everything from white collar professionals to front line staff and every police officer, teacher and social worker in between.
• Public sector workforces are geographically dispersed with operations across multiple capital cities in the case of the Commonwealth Government, or a mix of city and regional staff in the case of both state and federal governments.
• The public service operates a multitude of HR systems acquired over a long time, leading to challenges of data access and interoperability.
• Important public service HR data may also be held in manual non-automated spreadsheets prone to error and security risk.
• A complex industrial relations and entitlements framework, details of which are generally held in different datasets. 
• Constant machinery of government (MoG) changes demand both organisational and technological agility by public servants to keep delivering key services (as well as the delivery of ongoing and accurate HR reporting).
• The public sector faces increased competition for talent, both within the public service and externally with the private sector.  
• Citizen and political pressure for new services and methods of government service provision is at an all time high - so not only are your critical stakeholders your customers, they are your voters as well.
• Cyber security and accessibility issues that are unique to the public sector.
• This all comes under the pressure of constant cost constraints that require bureaucracies to do more with limited budgets.

As a result - understanding and best utilising limited human capital resources is crucial for the public sector at both a state and federal level. 

Now that we have isolated the unique people analytics challenges of the public sector, how do HR professionals within the public service begin the process of implementing a people analytics strategy?

1.  Data Orchestration

“Bringing all of your HR data together.”

The first stage of any successful people analytics programme is data orchestration, without having access to all of your relevant people data feeds in one place, it is almost impossible to develop a universal perspective of your workforce.

Having a unified analytical environment is critical as it allows HR to;

• Develop a single source of truth for the data you hold on employees.
• Cross reference employee data within and between departments to adequately benchmark and compare workforces to drive team-level, department-level and public service wide insights.
• Establish targeted interventions and not one-size-fits-all solutions.  
• For example, a contact centre is going to have very different metric results than your corporate groups like Finance or Legal.
• Blend data between systems to uncover previously hidden insights.
• Uncover issues such as underpayments that develop when different systems don’t communicate. Using people analytics to mitigate instances of underpayment is covered extensively in this blog.
• Provide a clean and organised HR data foundation from which to generate predictive insights. 
• Have the capacity to export modelled data to an enterprise data warehouse or another analytical environment (PowerBI, Tableau etc).
• Allow HR via people analytics to support the Enterprise data mesh - covered in more detail in this blog post. 

People data orchestration in the public sector is complicated by the reliance on legacy systems, as well as the constant changes in structure driven by machinery of government reforms.

Successful data orchestration can only be achieved through an intimate knowledge of the source HR systems and a demonstrated capacity to extract information from those systems and then model that information in a unified environment. 

This takes significant technology knowledge, such as bespoke API integrations for cloud based systems and proven experience working with on premise systems. It also requires subject matter expertise in the nuances of HR data. It can not be easily implemented without the right partners.

Ideally, the end solution should be a fully flexible open analytics infrastructure to future proof the public sector and allow for the ingestion of data from new people data systems as they arise (such as new LMS or pulse survey products) while also facilitating the migration of data from legacy systems to more modern cloud based platforms.

2.  Data Governance

“Establishing the framework to manage your data.” 

Now that all of your data is in one place, it is important you develop a robust framework for how to manage that data - in our view this has two parts - data definition and data access.

Data Definition 

Having consolidated multiple sources of data in one environment, the next step is metric definition, which is critical to being able to convert the disparate data sets that you have assembled into coherent, understandable language. It is all well and good to have your data in one place, but if you have 5 different definitions of what an FTE means from the five different systems you are aggregating then the benefits you receive from your data orchestration phase will be marginal. 

Comprehensive metric definitions with clear explanations are needed to ensure your data is properly orchestrated and organisation-wide stakeholders have confidence that data is standardised and can be trusted.

Data Access

HR data is some of the most complex and sensitive a government holds, so existing HR data management practices based on spreadsheets that can be easily distributed to non-approved stakeholders both inside and outside of your organisation are no longer fit for purpose. 

Since your people analytics data is coming from multiple systems you need to provide an overarching security framework that controls who gets access to what information and why. This framework must based on logical rules, aligned to broader departmental privacy policies and flexible enough to accommodate organisational change and to scale to your entire department or agency regardless of its size.

Critically, there needs to be a high level of automation and scalability to use role based security as a mechanism for safely sharing data to decision makers. Today’s spreadsheet based world relies on limiting data sharing, which also limits effective data driven decision making.

Finally, these role based security access frameworks need to be scalable so each new user or change in structure doesn’t require days of manual work from your team to ensure both access and compliance.

3.  Secure People Analytics Distribution

“Delivering people analytics content to your internal stakeholders.”

The next step, once you have consolidated your data and established an appropriate data governance framework, is to present and distribute this data to your internal stakeholders.

This is what we refer to as the distribution phase of your people analytics implementation.  

We established in the last section that for privacy and security reasons, different stakeholders require access to varying levels of information. The distribution phase goes one step further and places access within the prism of what individual stakeholders need in order to successfully do their jobs. 

For example, the information and insights necessary for a Departmental Secretary and a HR business partner to do their jobs are wildly different and therefore should be tailored to their particular needs. So, organisation wide metrics and reports in the case of the Departmental Secretary and team or individual level metrics for the HR BP or line manager. 

This is further complicated by disclosure requirements and reporting unique to the public service. This includes;

• Media requests regarding public servant pay and conditions
• Statutory reporting requirements for annual state of the public service reports 
• Submissions to and appearances before parliamentary committees 
• Disclosure to independent oversight inquiries or agencies

As a result, public sector HR leaders are required to walk a tightrope of both breadth and specificity. So how do we recommend you do this? 

• Offer a baseline of standardised metrics for the whole organisation.
• Tailor that baseline based on role-based access requirements, so stakeholders only see information that is relevant to drive data driven decision making.
• Deliver those insights at scale - the wider the stakeholder group consuming your outputs the better. 
• Ensure those outputs are timely and relevant - daily or weekly updates are recommended.
• Be able to justify your insights and offer access to raw data, calculations and metric definitions. 
• Continually educate your stakeholders about best practice people analytics.
• Increase reporting sophistication based on the people analytics maturity of your stakeholders - simple reporting for entry level stakeholders, more complicated predictive insights for the more advanced. 

To get the most out of your people analytics strategy you need to deliver two things; 

• Role based access to the widest stakeholder group across your department, the wider the group of employees that have access to detailed datasets the easier it will be to deliver data driven decision making. 
• Support your team with a change management programme to grow their analytical capability over the course of time.

4.  Extracting Value from your Data

“Using AI + Data Science to generate predictive insights.”

Now we get to the fun part - using data science to supercharge your analysis and generate predictive insights.

However, to quote the great theologian and people analytics pioneer - Spiderman - “With great power comes great responsibility.”

Most data science work today is performed by a very small number of people using arcane knowledge and coding in technologies like R or Python. It is not scalable and rarely shared.

The use of machine learning capabilities with people data requires a thoughtful approach that considers the following;

• Does your AI explain its decisions?
• Could the decisions your machine learning environment recommends withstand the scrutiny of a parliamentary committee? 
• Do you adhere to ethical AI frameworks and decision making?
• What effort has been made to detect and remove bias? 
• Does harnessing predictive insights require a data scientist or can it be used by everyday stakeholders within your department?
• Will your use of AI adhere to current or future standards, such as those recently proposed by the European Commission? To learn more about the European Commission proposal regarding new rules for AI, click here.

In integrating the use of machine learning into your people analytics programme, you must ensure that models are transparent and can be explained to both your internal and external stakeholders. 

5.  Using People Analytics to Support Public Sector Reform

“Public sector HR driving data-driven decision making.”

A people analytics strategy does not exist in isolation, it is a crucial aspect of any departmental strategy. However, in speaking to our public sector HR colleagues - they often feel that their priorities are sidelined or they don’t have the resources to argue for their importance. A lot of this has to do with the absence of integrated datasets and outputs to justify HR prioritisation and investment. 

We see people analytics and the successful aggregation of disparate data sets as the way that HR can drive their people priorities forward. 

If HR can present an integrated and trusted dataset, that allows comparison and cross validation with data from other verticals including finance, community engagement, procurement and IT. This gives HR the capability to be central to overall decision making and support broader departmental corporate strategies from the ground up. 

We have written extensively about the importance of data driven decision making in HR and using people analytics to support enterprise strategy, this content can be found on our blog here - www.onemodel.co/blog 

Why you should invest in people analytics and what

One Model can do to help.

The framework of a successful public sector people analytics project outlined above is the capability that the One Model platform delivers.  From data orchestration to predictive insights, One Model delivers a complete HR Analytics Capability.

The better you understand your workforce, the more ambitious the reform agendas you can fulfil. One Model is set up to not only orchestrate your data to help the public service understand the challenges of today, but through our proprietary OneAI platform - to help you build the public service of the future. 

One Model’s public sector clients are some of our most innovative and pragmatic, we love working with them.

At One Model, we are constantly engaging with the public sector about best practice people analytics - last year, our Chief Product Officer - Tony Ashton (https://www.linkedin.com/in/tony-ashton/) - himself a former Commonwealth HR public servant appeared on the NSW Public Service Commission’s The Spark podcast to discuss how the public sector can use people data to make better workforce decisions.  That podcast can be found here.

Let’s start a conversation

If you work in a public service department or agency and are interested in learning more about how the One Model solution can help you get the most out of your workforce, my email is patrick.mcgrath@onemodel.co