The Data Office Unpacked Its Importance and Practicality
By Tinia Halfar, Senior Director: Data Strategy, TransUnion
There can be little doubt that today, data is the lifeblood of most, if not all businesses, and thus establishing a data office is more important than ever.
This is particularly true as data is becoming more disparate, due to organisations creating more channels for their consumers to interact with, as well as the impact of social media. Additionally, in a world where competition is only growing fiercer, providing a great or poor customer experience can make or break a business. Gone are the days where businesses can divide their customers into five or ten segments and treat each segment in a predefined way, it is now essential that an organisation personalises their offerings according to each customer’s requirements.
Furthermore, establishing a data office is essential to improving internal operations. Since you can’t manage what you can’t measure, it is critical that efficiency is optimised along every step of the value chain. Having a data office can assist in all of these concerns, but moving towards establishing one begins with understanding what it actually entails.
The Data Office Defined
Simply put, a data office essentially manages, enhances and publishes data and information assets to the business community. Admittedly, data has always been important, but now businesses are maturing and are thus ready to use data more frequently as a business asset.
The first step is having a business strategy and identifying one’s primary objective. This can be further divided into three possible outcomes: cost leadership, differentiation or a focus strategy.
With a cost leadership strategy, a business is likely aiming to become the lowest cost producer, competing on price and targeting a broad market, while competitive advantage is obtained by driving down costs. These companies require core competencies to improve their internal operations as well as effective forecasting and planning to ensure supply chain activities are optimised.
A differentiation strategy, meanwhile, calls for bespoke products and services which are valued by consumers. In this case, customer and marketing analytics, innovation and creative product development are paramount.
A focus strategy concentrates on a limited segment of the market, where it is believed that the customer will be better served by focusing entirely on it. These companies typically enjoy a high degree of loyalty, with core competencies that include real-time, personalised analytics so as to enable treating each consumer uniquely.
Know Your Vision
Once a business has identified the strategy that best suits its goal, the second step is to define one’s data vision. This too requires a good amount of consideration. Its structure and focus depends on the business objectives that need to be achieved.
There are four distinct business objectives for which a data office would be established. The first of these is a compliance data office, where a data repository is required in order to comply with regulation and to intermittently submit reporting to governing agencies, a concern which is front of mind for credit bureau, banks and insurers. The second objective is turning to one’s data to perform analytics, and then using the insight gleaned to make better decisions, a topic that is often explored in greater depth when considering big data and data lakes.
In a similar vein, another data office objective could be to improve customer experiences. In this regard, we are seeing an increasing number of services companies investing in predictive analytics to personalise how consumers engage with their brand. The last data office objective is monetising data assets through creating new products and services so as to generate new revenue streams.
Assess Competencies and Capabilities
The third step in the process of establishing a data office entails assessing one’s data capabilities. Admittedly, there are many proficiencies that need to be in place in order so that raw data can be transformed and packaged into insight and information for the end user to digest. More specifically, these can be grouped into three categories, including building and design, management and publishing.
Building and design entails planning your delivery, and takes into account architecture, integration and development, while data management, or managing your delivery entails data warehouse and data quality management and issues such as business rules, performance, privacy and governance as well as security and quality assessment. Publishing, or exposing data to the end user, includes skills relating to data visualisation, data science, predictive analytics and business intelligence.
This then leads to the fourth step, that of a maturity assessment. Here is where organisations, understanding the objective of their data office, need to honestly determine the maturity for each data capability required. They are then able to ascertain what level of maturity needs to be attained in order to realise their business objective, identify the gap between the two and then prioritise which of the competencies are most important.
This can be done by rating each data capability between zero (non-existent) and five (optimised). Once the key priorities have been delineated, it is much easier to define what the resource requirements are, decide whether to buy or build them and put a change management programme in place accordingly.
The Home Stretch
With all the groundwork laid, organisations are then ready for the final step, that of choosing their delivery model. Here they can opt for a localised delivery model, a hub and spoke model, a full centralisation or centre of excellence. The benefits vary from one model to the next.
A localised model, which is a local instance with line of business delivery and control, for example, boasts a data movement performance advantage due to local connectivity, as well as faster integration with local systems. The hub model similarly runs off a local instance, but with selected components being delivered from a centralised location. This offers the advantage of centralised control and oversight, and thus increased efficiency of administrative functions.
A fully centralised data office, as the description suggests, sees all data capabilities being completely controlled and delivered from a centralised location, with the benefit being that of total central control and management oversight, as well as improved security.
Last but not least, the centre of excellence model has the benefits of autonomy as gleaned from a local instance, but with the benefits of having centralised controls and standards in place as well.
The only caveat that may factor into an organisation’s decision is that failing to at least try and centralise and integrate data from various sources, will result in one’s data only becoming more disparate, which in some regards defeats the purpose of establishing a data office in the first place.
The Way Forward
Clearly, establishing a data office is a complex process, but it is also a necessary one. However, when one considers that data can be sold on its own, and thus open up new revenue streams, this aspect alone makes venturing into establishing a data office well worthwhile. Indeed, selling data, data analytics and providing information for good is a business that TransUnion has been involved in for well over a century which only makes its sustainable value that much more apparent, even before digitisation saw data expand exponentially.