Making people count with people analytics
Mon 3 Feb 2020
How an HR function went on a voyage of taking raw data and deploying predictive analytics within 8 months
People and Human Capital analytics to solve business challenges is gaining momentum. The 2018 Deloitte Global Human Capital Trends report, based on a survey of 11,000 business leaders globally, revealed that 85 percent of respondents considered people analytics important, but only 42 percent were ready to address it.
Figure 1. The Analytics Maturity Model (adapted from Gartner)
In this blog, we share our experience of successfully accelerating the Analytics Maturity Model (Figure 1) for HR at a FTSE 10 company. In under 10 months, we brought them on a journey of integrating and understanding data, deploying self-service analytics and machine learning experiments. This was done with a SaaS data warehouse platform (TrueCue) and an interactive data visualisation service (PowerBI).
TrueCue, a low-code smart data-provisioning platform, helped to bring together and dimensionally model multiple data sources into a single business focussed and performant data warehouse. The unique feature about the low-code approach was the ability to iteratively change and extend the dimensional model.
This enabled the team to focus their time and efforts on understanding the business reporting needs, business logic and KPIs rather than getting caught up on building and deploying the data warehouse.
Milestone: A foundational and extensible People Data Warehouse deployed in 3 months
Building trust in data
The agile approach to data warehouse development and instantaneous data visualisation with PowerBI created a solution that could be stress tested on the fly. The client team had the opportunity to verify the data as the solution was being developed and observe how the feedback was being incorporated into the data model and cascading through to the data visualisations. This resulted in our stakeholders building strong trust in the data and business logic that was incorporated into the data-warehouse.
Milestone: Data tested and validated within 3 months of the project star
The What and the Why
Over 30 dashboards were built in an iterative sprint-based approach (multiple 2-week sprints). These were shared with over 130 users globally including the Senior HR Leadership team.
The analysis covered:
- Workforce Planning; analysis of the workforce composition and alignment with organisation design principles
- Talent Acquisition; a detailed breakdown of recruitment activities and their success
- Learning and Development; a holistic view of L&D activities and return on investment
- Talent Management; analysis of talent movement, turnover and performance
- Diversity; strength of senior management and gender diversity
- Succession Planning; Identification of critical roles and help select and visualise succession
- Reward; Deep understanding of compensation modelling, policy decisions and benchmarking against the industry
Milestone: Dashboards developed, tested and deployed in 6 months
Laying the groundwork for prediction
Data was now validated and centrally accessed through a data-warehouse. The strong level of trust in the data process helped make another leap in the analytics journey, onward into predictive analytics.
Using one of the curated datasets in the data warehouse, one of our in-house data scientists was able to build and deliver a fully-functioning, dynamic and scalable model predicting company turnover in under a month (Balanced Random Forest model). This is now being productionised in the Microsoft Azure environment demonstrating the speed to value of having a fit-for-purpose analytics technology stack.
Milestone: Proof of concept predictive flight risk model developed and tested in 8 months
Mingyang Tham – Analytics Consultant, Concentra and Darshan Baskaran – Management Consultant
TrueCue by Concentra will be exclusively hosting the VIP Lounge at Big Data and AI World on March 11th & 12th. Register your free place.
 Source: 2018 Deloitte Global Human Capital Trends: The rise of the social enterprise
 Data from the HRIS system was integrated with data from other corporate systems and external data sources such as social media data and industry benchmarks of remuneration