1. As a first step, align, link and involve the data & analytics (D&A) organization with the strategy and operations organization. Then conduct a top-down analysis of the strategic needs. But be sure to follow this with a “bottom-up” review of the specific types and sources of data you will need, so that you can fully understand what it will take to achieve your goals.

2. As a second step, develop a data collection, standardization and cleaning process such as follows:

• How frequently does data need to be updated?
• What data archiving and summarization requirements are there?
• What data is needed where and when? Today and in the future?
• How can you integrate data silos? What data transfers are necessary, and how frequently?

This stage is critical for creating data connection, rather than just data collection – meaning how different data should be joined and integrated to make meaningful conclusions.

3. As a third step, create a data governance framework to answer the tough questions about roles & responsibilities, approvals, and workflows, in order to ensure that the data is well-managed and utilized.

As data is the most under-utilized asset in many companies, improving its utilization will naturally create value.

We believe that a data strategy should be aligned with, and should support, the organization’s strategic drivers, such as increasing revenue and value (growth), reducing cost and complexity (efficiency), and ensuring survival by mitigating risk (risk).

4. Set up an official forum – a “Data Vision Engineering Workshop” – which would meet quarterly or semi-annually in order to identify future business opportunities by answering questions such as:

• What does the collected data tell us in general, and what conclusions can be drawn?
• What does the collected data tell us about customer needs?
• What defensive or offensive actions are needed?
• How can we turn these conclusions into assets?

As always, the output of a good data strategy should be measured. Organizations should be able to quantify the extent to which sales have been increased, and or efficiency has been improved, or risks have been reduced, as the result of better utilization of data.

A simple measurement such as a “Return on Utilized Data” (ROUD) could be employed as follows:

Increased sales + increased efficiency + reduced risk, divided by cost of investment (cost of internal hours + resources + cash spent).

The result would be expressed as a percentage or a ratio.

To summarize, a good data strategy should answer the following questions:

• What data do we need to give us significant advantage?
• How should we utilize the data in order to draw good conclusions?
• How can we govern this as an ongoing process?
• What kinds of competence and skills do we need?