145071452

Studying this technical article and answering the related questions can count towards your verifiable CPD if you are following the unit route to CPD and the content is relevant to your learning and development needs. One hour of learning equates to one unit of CPD. We'd suggest that you use this as a guide when allocating yourself CPD units.

This article was first published in the May 2019 Ireland edition of Accounting and Business magazine.

Organisations of all sizes, including many accountancy and financial services firms, are collecting huge volumes of data on customers, transactions and digital activities. But extracting actionable insight from this vast data concentration is proving a significant challenge. Many organisations are turning to analytics to extract insights from data to drive actions with positive business outcomes.

Analytics is the discovery, interpretation and communication of significant patterns in data. Derived using mathematics, statistics, predictive modelling and machine learning techniques, these patterns are then used to improve decision-making. Analytics is now deemed an essential capability in gaining competitive advantage in the digital economy. Common business applications of analytics include fraud detection, credit risk analysis, marketing optimisation, web behaviour analysis, dynamic pricing, inventory optimisation, customer segmentation, customer retention and churn analysis.

The analytics operating model selected by an organisation shapes its ability to capture business value from investment in analytics infrastructure and resources. In attempting to build an analytics-driven culture, organisations strive to implement the operating model that best enables their analytics talent and capabilities to thrive.

There are three types of analytics operating model:

  • a centralised analytics environment, with a core unit serving enterprise-wide analytics needs
  • a decentralised organisation of analytics teams dedicated to individual business units
  • a hybrid hub and spoke structure of a centralised strategic analytics unit, with analytics teams embedded in business units.

Centralised

A centralised unit provides analytical expertise, support and output to business teams while setting the roadmap for analytics implementation throughout the organisation. This tends to involve resources working out of a central location on diverse analytics projects, serving functions and business teams with limited access to data, tools or resources. It facilitates enterprise-wide prioritisation of analytics projects.

The centralised model suits large, single-business organisations with a need for analytics applications that cross functional boundaries. It gives analytics talent the opportunity to share ideas on a wide range of analytics projects and use cases, leading to a greater understanding of the company’s strategy, capabilities, lines of business and competitive environment.

The downside is that it can create distance between analytics resources and the business teams, particularly if all analysts are housed in one location – a disconnect that can lead to gaps in understanding of business problems and opportunities. It is vital to have structures to ensure the analytics function is meeting the needs of the business.

Decentralised

In the decentralised model, analytics resources are dispersed across the organisation in different business units with little or no coordination. Business units conduct analytics activities, exercising complete control over their analytics resources and needs, with no mechanisms in place to facilitate collaboration or coordination on an enterprise level.

Dedicated business-embedded analytics resources yield short-term benefits through fast turnaround on analytics requests. Analytics resources in decentralised teams possess considerable domain knowledge and business insight.

The decentralised model suits companies with few analytics resources. It can be an indication that senior management do not place high strategic importance on analytics, and can lead to inefficiencies in the long term. It is, for example, difficult to set enterprise-wide analytical priorities, and to develop and deploy staff effectively. Resources are allocated only to analytics projects in each isolated silo, with no visibility of analytics activities or priorities outside their function or business unit. Without some central coordination, decentralised analytics frequently leads to inconsistencies in the figures and metrics produced by each isolated silo.

Hub and spoke

As organisations move up the analytics maturity scale, a more balanced approach can effectively manage the enterprise-wide analytics resources while meeting the growing business demand for analytics to solve problems and identify opportunities. The hub and spoke operating model has a central ‘hub’ team, focused on enterprise-wide strategic analytics initiatives, linked to ‘spokes’ – analytics teams embedded in individual business units responsible for the design, development and delivery of analytics solutions.

The hub consists of a lightweight, agile and innovative analytics team responsible for developing and executing the enterprise-wide analytics strategy as well as identifying appropriate frameworks, tools and standards to enable the organisation to become analytics-driven. The hub builds relationships with spokes across the organisation, working collaboratively with these analytics teams to provide support and technical guidance while driving the adoption of best practices. The goal of spoke teams is to develop insights that have maximum impact for individual business units.

The hub and spoke model promotes knowledge sharing between analytics teams, reducing effort duplication while enabling greater efficiencies.. The central hub offers analytics talent capability development opportunities as well as the opportunity to specialise in different areas of the business to deepen domain knowledge while building trusted relationships with decision-makers.

Whichever model an organisation adopts or moves towards, the next important step is deciding on the resources and skills required to embed analytics in an organisastion. That will be the subject of the next article.

Jamie Renehan, FCCA, leads Bank of Ireland’s Advanced Analytics team.