In last week’s Brief I described how The Institute’s courses, and its Mastery, Certificate and Degree programs would continue without interruption, following our acquisition by Elder Research, Inc.  Now I’d like to talk about how the Institute’s students stand to gain from the expertise and experience of Elder Research.  

A common topic in the analytics press is the high failure rate of data science projects; I’ve seen estimates of 75% and 80%.  A typical scenario is where a model succeeds technically, but either fails during deployment or is never deployed. Elder Research reports a much higher success rate – 65% – but they consider even this unsatisfactory, and are constantly striving to improve their secret sauce.  Some of the ingredients in that sauce include:

1. Attention to building the right analytics team, comprised of visionaries, architects and technical workhorses
2. A flexible mix of Elder Research consultants and client subject matter experts, to incorporate business familiarity and expertise, and to leave the business with the ongoing capability to manage the project
3. Acquiring software and tools after needs assessment, rather than before (which often produces “shelfware”)
4. Guiding the project buy-in process so that all stakeholders expectations are aligned

What does this mean for Statistics.com students?

1. We look forward to building elements of this secret sauce into our courses, to give you an edge in the marketplace.
2. For those seeking to deploy projects, or build an analytics capability in your organization, you can build on your Statistics.com education and consider ERI as a potential consulting/training partner.

You can read more about the Elder Research difference by reviewing their white papers on the cost of model complexity, applications of reliability assessment, what data scientists can learn from software engineers, and more.

If you think you might have a project now that could use some help from Elder Research consultants, or a custom training program, contact Janet Dobbins at jdobbins@statistics.com.