How to turn real-time data streams into business value?

We generate digital data every time we tweet, search for a hotel on Travelocity, send an email, tap on a smartphone icon, walk into our badge-enabled offices, or drive through a tollbooth. These actions create real-time digital data streams (DDSes) that companies can use to create new products and services, improve their value to existing customers, and optimize internal operations. Many organizations have used DDSes to enhance their market positions and gain advantage over competitors.

Companies use DDSes to create value through:

Data generation: By originating the stream of data itself — either deliberately or as a byproduct of other activities — companies could then stream this data to other partners. In turn, the partners could harvest this data and create value-added services leveraging the DDS.

Data aggregation: Companies focus on collecting, aggregating and repurposing streams of real-time data.

Service: Companies use one or more streams to provide services to consumers or to improve service quality.

Efficiency: By using real-time data streams to optimize internal operations or to track business performance, companies can improve efficiency (e.g., waste reduction, better response speed).

Analytics: Companies process real-time data and information to produce analyses and improve visualizations. These results assist in better decision-making and produce superior insight, such as through dashboards and data mining.

Assessing usable data streams

Realizing the potential of DDSes depends on not only envisioning possibilities, but also on assessing potential data streams that may already exist or could be generated or aggregated. How can the firm evaluate which streams to harvest? The first step is to identify potential DDSes based on feasibility: how streamable and complete they are.

Streamability enables firms to assess the viability of harnessing a given class of events or creating a data stream that does not currently exist. Those events most amenable to creation meet the criteria of detectability and measurability. An event is detectable if it exceeds a minimum threshold magnitude to sense it. For example, a sensor coupled to a telescope can only detect a star if it receives enough emitted photons. An event is measurable if it can be accurately quantified. A firm’s quarterly profits have high measurability, while the subjectivity of an individual’s pain level limits its measurability. In addition to evaluating the streamability of potentially valuable data sources, a company should consider whether data sources contain the information needed to describe an event. Such information can be categorized as when, where, who, what, how and why.

Once you have identified potential DDSes based on streamability and completeness, the next step is to assess how much value the firm can extract from the initiative, and consider it in tandem with feasibility. Infeasible DDSes with no perceived potential will make investment unwise. Conversely, high feasibility combined with strong upside potential calls for immediate investment.