| The Economics of Data Analytics | ||
| Prev | Next | |
Data Analytics uses multiple resources from IT, Finance, and Operations. Every new implementation is therefore prone to some of the issues typical to these fields and some more due to interactions between people not used to work together.
Average IT failure rate: 5%-15% (a loss of $50-$150 billion per year in the USA and €142 billion in the European Union as of 2004).
Harvard Business Review: for 1,471 IT projects, average overrun 27%; one in six projects had overruns of 200% in cost and 70% schedule.
PricewaterhouseCoopers: for 10,640 projects from 200 companies in 30 countries across various industries only 2.5% of the companies successfully completed 100% of their projects.
Because of their layered structure Data Analytic projects can fail at multiple levels making them more prone to overruns than regular IT projects (55% vs 10%).
The mingling structure of a Data Analytic project brings a tremendous potential value. However, the value still needs to be discovered, understood, and managed correctly to reach its full potential instead of coming to bite its user.
Data and algorithms are not commodities. Their value has at best a weak link with volume or other traditional measurements. Without structure and strategy data and algorithms remain empty promises.
Valuation of Data Analytic projects is derived from the traditional financial valuation methods with special care due to the heuristic nature of predictions based on mostly unstructured data.
As in many other investments costs are guaranteed but benefits are only for the well informed. Valuation is the tool that helps you see beyond this conundrum.
