Data science improves organizational function, but it isn't always cheap. Scientific data analysis enables organizations to develop more internally efficient business systems. They can more easily reach potential stakeholders and engage them. Data informed organizations can prepare forward, more accurately refine and refocus strategies, and more effectively manage risks. The problem is data science itself is an investment heavy human resource, and those costs are passed on to the employer. Trainees need a strong background in probability and statistics, and expertise in a speciality subject, like political science, medicine, engineering, or economics. Finally, to round it off, data science as a practice is a computer science intensive activity. Data scientist need proficiency and expertise in complex computer applications to get the job done.

The path is rigorous and time intensive. The result, in a global information boom with rapidly growing demand for skilled data analysts, is a shortage of deep analytical talent, and rising costs of training and retaining analytical departments. Small organizations are not typically positioned to hire or train data scientists and supporting analytical teams. Public organizations lose talent quickly to the private sector. To add to the investment challenge, new technologies, software, and data formats are developing rapidly. Training and retaining in-house people to acquire expertise in the variety of applications and languages now on the market, doesn't ensure those applications won't become obsolete in the near future. The risk is that investment in analytical resources doesn't come with any reasonable assurance long-term demands will be met. Fortunately, there are options, and organizations are starting to use them.

High costs and high risks are prompting more and more organizations to outsource data management, data science, and analytics to a growing number of data start-ups. The market is small, compared to the accounting or law markets, but nonetheless it's large enough to sustain an already growing number of stand-alone data-focused businesses. These start-ups accumulate expertise in analytics and data science, and distribute it across a wide number of organizations and sectors. Firms taking advantage of the stand-alone data science market more efficiently meet project budgets and timelines, and limit investment risk.