Outside the IT department, data is not an area of business that traditionally received much attention, let alone its own strategy. Usually data comes under the scrutiny of management when a crisis hits:
- Conflicting or erroneous numbers are reported to senior executives or the board of directors.
- A very bad (ie, costly) decision is made based on erroneous data.
- A new software implementation fails to deliver the expected results because of long standing data problems.
An effective data strategy will not only head off these problems but turn data into a valuable business asset. I am using the word “data” in the broadest sense of the term, which encompasses business analytics, business intelligence, cubes, reporting, data visualization, web analytics and of course, big data. To ensure success a plan is required to tie everything together into a coherent strategy.
Here are seven steps to building an effective data strategy:
- Quick Wins – The first step is to get out of crisis mode by dealing with the most pressing data problems, which includes the catalyst that brought the data strategy concept to the forefront. Get a grip on the root of the problem. If it can’t be fixed immediately, then at least have a plan for doing so soon.
- Assessment – Conduct an audit of data sources, reports, dashboards and the related software. Document what is working, where data gaps exist and any other problems, from both a business and a technology perspective.
- Road-map – Define the business problems to solve and the business processes to optimize, then ask some questions: What can be done with the available data? Where do the data gaps exist? Does data quality need to improve and data quantity need to increase? Can the required data be procured from an external source? Can useful data be obtained by conducting experiments (for example, A/B split testing on the website)? Will updated or new software, hardware or a cloud service be needed?
- Execution – Put in place the highest value items identified in the road-map. Change management is an important component, as achieving business objectives via data driven decisions often demands a cultural shift. The implications can be far reaching, affecting staff, customers and suppliers. Executing on the strategy also includes obtaining executive sponsorship and communication of early results.
- Iteration – Executing a data strategy is unlike a traditional IT project. The results are sometimes surprising and require the project team to revisit the initial objectives and scope. The technology is evolving quickly, which will open up possibilities that did not exist when the strategy was created. There needs to be a feedback loop to keep the results relevant to the business.
- Alignment – For long term success, the data strategy must be aligned with business strategy and IT architecture. Executives must continually endorse the strategy and communicate its importance to the business as a whole. IT support is critical, as supporting the data infrastructure is their role, and it cannot be viewed as secondary to other initiatives.
- New Opportunities – An effective data strategy is not about continuing to solve yesterday’s problems. Once some success is achieved, new data driven business opportunities can be explored. Data analysis will uncover new markets, new potential customers and other business processes to optimize or automate. Does the data itself have value for customers and suppliers? Can aggregated and packaged data be the basis of a new product or service?
A crisis does not have to occur before putting a data strategy in place. The importance of data has more credibility than ever, with so many examples of successful companies, such as Amazon and Netflix, using data as a competitive advantage. In 2014, there is no reason for smaller companies to not be doing the same.