Opportunities in Data
Making the optimal business decision, whether at a strategic or an operational level requires information – and that information needs to be timely, accurate and at the right level. The raw data that will be used to generate that information can live in different systems, different functions and even different companies (i.e. outside the organization). Today, new insights will deliver competitive advantage – and those insights will be delivered out of data as the raw material.
Data is not only an abundant raw material, but it is one that can be captured or generated in new ways. The insights gained from data, especially combining the data inside the organization with external data such as geo-spatial or public domain data can lead to entirely new lines of business. The asset is potentially valuable and game changing, however, it is easy to get sidetracked and make huge investments in data infrastructure or analytics that do not pay off.
Data Strategy will provide the organization with a means of managing data as an asset, whether it is governance of the data resource itself, identifying the highest value opportunities to pursue, building the data supply chain or championing a data mindset across the organization.
Recognition of data as an important resource will require certain processes to be established to guarantee data quality, compliance with all local legislation and establishing organizational policy regarding archiving, sharing and ensuring the integrity across various data silos.
Data governance will likely be implemented through a data standards body at the program level – and will ensure that stakeholders are appropriately involved.
A ‘data product’ is the processed data deliverable that creates business value. Inside the organization these may be reports, dashboards or data feeds that go into processes control or the insights that drive criteria for new avenues of exploration.
Suppliers or Customers can consume data products outside the organization. For example, suppliers of equipment will be supplied with the performance of consumables that will allow them to lower costs or increase usable lifetimes. Customers may be provided with advanced information with regard to supplies that will allow them to make forward buys.
The data could also be used to create entirely new product-markets; many organizations have successfully built new revenue streams based on their data exhaust.
Data Supply Chains
In a physical supply chain, raw materials are processed and turned into consumable products – this demands a reliable supply of quality raw materials supplied in a cost effective manner. To provision data products requires a similar reliable supply of data that is timely, accurate and correctly structured.
Creating an effective data supply chain means connecting the sources of data inside and outside the organization to the points where information is consumed – for decision making (e.g. which are our bottom 10% of customers, are they actually profitable, what would be the effect of removing them) or real time process control (where should I deploy my scarce resource next).
The Data Supply Chain is a virtual system where existing sources of data are harnessed, re-structured as necessary and then plumbed into a network where data is allowed to flow. ETL processes, Data Warehouses and Business Intelligence application form one aspect of the data supply-chain, however, being able to handle Big Data projects and real time processes will require new approaches to the problem.
Insights are used by the organization – by operatives, analysts, managers and executives to make both strategic and operational decisions. The insights can also be delivered to suppliers and customers.
Most large organizations have effective systems that provide reports, analytic tools, dashboards and ultra-high level summaries. Effective use of data visualization will augment those tools and allow decision makers to better understand the insights. People are increasingly familiar with the idea of applications deployed to mobile devices and insights can be delivered in this way.
Risks and Mitigations
There are numerous pitfalls that create misalignments, cost and time overruns or impair the success of data projects and programs.
Stakeholder Buy In: As both the sources of data and the points of consumption are distributed across business units and geographical territories, support of the program in all its manifestations is vital. Missing this creates issues where data is not made available, that insights are not leveraged or obstacles in the supply chain prevent effective use of data.
Rigid Approach: The organization-wide data strategy should be flexible and should accommodate the unique needs and priorities of each stakeholder group and each business function. This will require the ability to work with a broad range of technologies and to serve all business processes.
Appropriate Technology: The technology considerations are heavily influenced by corporate strategy, by vendors and analysts. The current vogue is towards the Big Data stack (cloud storage and processing, NoSQL databases, Hadoop File Servers and processing, Streaming data sources and Algorithms to process unstructured data) – yet only certain business cases really demand this. Therefore the Big Data stack should be deployed where appropriate and more traditional forms or hybrids used where this serves the purpose.
Program Structure: Sizing the project and structuring it in such a way as to achieve regular small successes, each building on the next is important. Being able to use one project to demonstrate a positive return and using that to fund the next is important. However each project should successively contribute to building the long-term capability, hence it is important to have a strategic data architect that will move the program towards an end-state.
Areas of Focus
The Data Strategy will define a portfolio of projects – each with a specific role to deliver results, build infrastructure, gain support or establish governance. Selection and definition of projects is a key responsibility of the Data Program and must be done with full engagement with business and technology stakeholders.