Breaking down big data for efficient delivery of service

Big data is nothing new, only that recent years have given it prominence and focus pegged on the need to achieve competitive advantage and efficient service.

At its very basic state volume is not definitive of inherent value and more or bigger is not necessarily better. There are other characteristics of data that must be part of its orchestra to create harmony.

Data velocity looks at how fast new data is acquired and processed, an important factor as it impacts the procedures set up to aggregate, sanitise and normalise the data. This affects the decision execution depending on the use of the dataset.

Take telemetry for example, if a use case of the derived data from different sources is to provide real-time traffic data, a 15 minute cycle between data capture to commercial utilisation is not favourable and corrective measures on system design or platform process must be initiated.

Verifiability should be looked at closely. As the sources of data grow, it takes only “one” wrong variable to throw off an entire business intelligence operation.

Data is of value provided it is used before its “sell by date”. Decisions driven by data that is aged can be fatal especially in contexts where data currency is key. In healthcare services or energy provision when thinking nuclear power the issue data validity carries a lot of weight.

In a sensor and trigger driven world, data comes in all shapes, sizes and formats. Platforms must be adaptive. Variety births insights which can cause a tricky situation of privacy violations when hitherto siloed and protected datasets are processed to present a personally identifying profile.

Not everyone needs to plug into big data and that decision is best informed by a concrete goal, driven by a desired output.

The national health ICT master plan would be to reduce the number of hospital visits for non-chronic conditions in all level 3 healthcare facilities across the country by 30 per cent over 48 months in order to realize cost savings amounting to X, that can be channelled to other development projects.

This consideration helps form the game plan as far as data velocity, verifiability, validity and variety are concerned.

Njihia is CEO of Symbiotic

Twitter - @mbuguanjihia