No doubt – HR is not exactly a data savvy function: Between 80% and 100% of CEOs require strategic decision support informed by people data, yet less than 20% of them receive such information from the HR function. That message is well understood by HR: In our recent research, European Heads of HR Analytics report that they would like to significantly increase their analytics maturity. So Analytics is constantly climbing up the ladder of HR priorities: In a five-level analytics maturity model, the average company aims to improve their analytics maturity from level 1.9 to 4.3 within the next three years. Moreover, companies tell us that Workforce Analytics is the centerpiece of their digital HR agenda. Three out of four large companies declare Workforce Analytics as a top priority for 2016 – with the other two digital agenda columns (‘consumerization of HR’ and ‘new organization of work’) reaching that 75%-level of importance by 2019.
Two Components Leading to Better People Decisions
Much has been written about workforce analytics, a lot less has been implemented. Yet the business case for advanced workforce analytics is clear: If set up correctly, mature analytics companies achieve better talent outcomes and produce higher business impact. Instead of waiting for all data and technology infrastructure to be in place, they don’t wait for full integration of data and technology before enabling better people decisions through analytics. Here is how:
Component 1: Critical Analytics Use Cases
One aspect that requires attention when growing analytics maturity of an HR function: It’s credibility. Today, very few business leaders believe that the HR function is effective at supporting decisions with facts and figures. Hence to achieve the ultimate goal of workforce analytics – improvement of people-related business decisions – HR needs to gain acceptance of their analytics work. Some of the best practices of companies that have successfully increased HR’s credibility include the building of very-easy-to-use scenario tools for business managers to play around with and see the effects of their assumptions on talent outcomes. Others have given ownership of significant analytics projects to business senior leaders, ultimately benefitting from great analytics results. This aspect alone requires a significant change management effort. Hence companies ahead of the analytics curve have started their analytics change journey with ‘use cases’ that were purely focusing on quick and tangible business outcomes. Even without the data and skill infrastructure in place, these companies got started. They identified critical business needs and started an analytics use case with high visibility and impact in the business. If the required data and skill infrastructure was not available, these companies built their data lake ‘manually’ and hired the required skills from other functions or from the outside. The aim is to build business credibility with decision support where the business needs it the most – in truly critical people related questions.
Moreover, analytics use cases do not only grow business credibility. They also deliver two very important aspects of the future layout of analytics in an organization. First, within the limited scope of an analytics use case it is easier to find basic agreements with major stakeholders of workforce analytics, e.g. data security, data protection and works councils. Advanced companies use their analytics use cases to build up these stakeholder groups’ knowledge of the advantages and the potential risks of workforce analytics. And they draft basic agreements such as ‘foundations of workforce data protection’. Second, the use cases allow for a professional ‘co-creation’ of workforce analytics, together with the end-customer of analytics, typically the line manager or the senior executive. Using a ‘digital customer journey’ approach helps advanced analytics companies to understand the real end-customer impact of analytics. The approach is simple and proven for more than 10 years by consumer marketing:
1. Build (or adapt) an HR customer model that clearly defines the HR end-customer for workforce analytics services.
2. Create a fictive ‘persona’ to represent the end-customer for a specific analytics use case.
3. Break the analytics use case down into several touch-points of the persona with the analytics use case (the ‘stages’).
4. Describe the persona’s current experience at each of the stages – either by real end-customers or by using the customer knowledge of the business HR function.
5. At each critical stage, compare the persona’s current experience to the future expectation for an ‘ideal’ analytics service.
6.Derive the real analytics impact on the end-user of an analytics service from the gap between the persona’s current experience and the future experience.
Side note: In our current research run, TI People analyzes digital customer journeys of HR and co-creates a ‘persona tool’ together with some 50 companies in Europe and the U.S. This tool will help companies design their ‘HR services for a digital world’ – strictly focused on end-user impact. The above 6-step approach will be part of that tool.
Let’s wrap up phase 1: Successful companies build workforce analytics use cases with a high business impact to achieve business credibility, basic agreements with stakeholder groups and a realistic picture of the end-customer impact of workforce analytics. They do it even if they do not have the most sophisticated analytics infrastructure in place – simply to not lose time to produce business impact.
Component 2: Sustainable Analytics Delivery
We have worked with several companies that have successfully accomplished component 1 and have run critical use cases. These companies have achieved a required but insufficient analytics level. It would simply be too expensive for these companies to roll their initial analytics use cases out and build new ones – one-by-one, just as they did with their initial use cases. With all the manual and bespoke work to identify the business problem for the use case, to collect data from various systems, to build analytics models to derive insights from data, to discuss data protection etc., to produce presentable results and to communicate these to the business the cost per analytics service is way too high for sustainable corporate-wide business impact. Hence these companies need a second component – a sustainable delivery model for analytics. After our initial research with some 50 companies in Europe and in the U.S., and after an analysis of some corporate use case projects of large, global companies, we built a framework for an analytics delivery model:
The first pillar of a sustainable delivery model for analytics is the ‘Analytics Data Lake’, i.e. the (virtual) set of data sources to be used in the entirety of analytics services. Many of these data sources will be external, like profile databases (LinkedIn, etc.) or talent intelligence databases (CEB TalentNeuron, etc.). Others will be owned by functions outside HR, like SharePoint databases, ‘owned’ by IT or even by each single business unit that has set up ‘their’ SharePoint. In many companies, up to now HR has not had the mandate to manage or co-own so many and diverse data sources, this in itself being a stretch. Moreover, many data sources of an advanced analytics data lake will be of unstructured nature – each single data point will not necessarily be insightful, but patterns within these data points will contain valuable information for workforce analytics. Hence, the data lake will most probably not be a static one but rather a flexible data model comprised of access routes (if not direct interfaces) to structured and unstructured data sources inside and outside the company.
The second pillar is the design of ‘Analytics-as-a-Service’. Here, the above mentioned customer centricity and ‘persona’ approach applies in most companies. We have sliced and diced analytics services into bespoke vs. standard for three layers of services:
– The data layer, which is the entirety of data sets that the HR function either provides as a standard or adds in upon customer request (bespoke integration).
– The analytics layer, which is the set of standard analytics models the HR function provides to the business plus specific bespoke model built upon business request.
– The presentation layer which is the set of visualization and access tools that HR offers to the business plus – as needed – the advisory services to deliver the analytics insights.
In addition to these layers, ‘analytics-as-a-service’ will only be complete with other advisory services, i.e. to support the business HR function to deeply understand a business problem and with training for various audiences (first and foremost business HR and the business).
The third pillar – Analytics Delivery – is comprised of the various instances for a sustainable delivery model: We will see a corporate CoE role in analytics as well as a local business HR role. Moreover, some analytics services will be delivered through dashboards or inter-active end-user platforms rather than HR professionals, some of them supported by machine learning features. The set of analytics services (second pillar, see above) provided to the business and the maturity of the data, technology and skill ‘infrastructure’ of an HR function will determine the ideal combination of these instances. In the end, the aim will be to deliver all analytics services required by the business (i.e. with a positive ROI) at the lowest possible cost. The two most critical questions for an optimized analytics ROI will be (1) the end-user adoption of the analytics services and (2) the investment in machine intelligence and automation vs. (business) HR training and skill acquisition. The best companies have not built one sustainable analytics delivery model but a journey or roadmap to growing impact and optimized ROI of analytics.
Moreover, this third pillar will also describe the roles of the HR function to deliver the services: As HR rarely had the mandate to build sophisticated analytic infrastructure, HR function usually lacks the required data analysis skills. After having spoken to some 30 thought leaders of workforce analytics, we have built an end-to-end role model for analytics:
– Business Challenger: Business challenging and influencing skills, combined with ownership of analytics policies, service delivery model and projects
– HR Domain Expert: Deep business and HR skills to analyze business needs, paired with basic analytic skills to understand limitations of statistic models
– Data Scientist: Classic planning, reporting and analytics skills plus advanced big data analytics skills
– Programmer: Advanced database design and programming skills to retrieve data from various sources and make it available for data scientists
Finally, the analytics delivery model will describe (1) which data sets to be used (2) by which services (3) delivered by which instances to (4) improve business decisions, typically in the areas of selection, engagement, attrition, succession and cost of work. All of these will then be framed by a governance for Workforce analytics, typically owned by a CoE.
Conclusion: Don’t wait!
With Analytics being around as a hot topic for two years now, actually little has happened so far. We have seen some icons (like IBM) that have invested highly in analytics internally (selling analytics services and technology to their customers as well). We have seen many companies experimenting with advanced modeling and external and unstructured data sources. But compared to the hype, the reality is actually somewhat slow. With some of the above thoughts applied, we believe that Analytics is here now. And that companies should not excuse themselves with missing data and technology infrastructure. There is no doubt that analytics is the centerpiece of an HR agenda in a digital world and the key capability of the HR function in the future, earning the HR function nothing less than its right to exist. So why wait? Start improving the analytics maturity now.
Learn more about How The World of Work Is Changing.