+44 203 137 2533
The CDO View
Creating real value and differentiating the data lake.
It is key to create a vision for your organisation, putting data at the centre of the decision-making machinery, whether automated or human-driven. The data lake layer is a critical part of the data fabric, being fed by all of the operational systems and external data sources, it needs to not only enrich the data, but also provide a range of services from batch analytical access through to real time API’s.
It is within this layer that the key capabilities such as entity resolution, network building, search, etc. should be available to all parts of your organisation. Above this layer, a set of decision making models can be provided, covering functions from fraud detection to marketing.
To achieve operational benefit from the analytics it is essential to have an action layer, where outcomes are either executed directly through automated decision engines, or users can interact directly with the recommendations and associated data.
Data scientists should also have access to further explore the insight, support strategic decision making and identify new opportunities to use analytics.
Steps to achieving a CDO strategy
Although many senior management teams intuitively believe that there is significant value to becoming a data-driven organisation, it is not always clear how to make this happen in practice. There are a few critical steps to ensuring success:
Vision: ensure your organisation understands what could be achieved and why a good CDO strategy can be transformational—focus on concrete outcomes;
Design: it is crucial to shift to an architecture that is layered horizontally. Too often there are silo solutions where the data is vertically integrated into a specific application, each of which require duplicated data and dedicated niche skills, with no flexibility to easily add new models and capability;
Team: establish a shared resource of data science skills that can assist different business functions in getting the most out of the organisation’s data;
Transformation plan: very few organisations offer a green field opportunity, you must assume a migration. Ensure the CDO approach is embedded within change programmes and that it becomes the “data fabric” though which all data is plumbed. Measure this progress through KPI’s for management across the organisation;
Quick wins: these are critical to the success of the programme. Choose compelling opportunities with a proven track record where analytics can provide benefit, for example within the fraud and financial crime domain;
Communicate and sell: the CDO programme must be communicated in a compelling way internally to maintain momentum. Ensure all the departments understand how data is providing significant benefit.
Exploit: once established, before long the data lake will contain most of your data. If there is embedded capability for network analytics, new models will be relatively easy to deploy and you can quickly create new business applications – make this an ongoing program.