Dan Masciopinto
SVP Product and Services, NextHealth Technologies

img_1478-e1455638442273Last month, Eric Grossman, the CEO of NextHealth Technologies (“NHT”), discussed the 4 requirements of a “No Regret Investment” in analytics. Next, we are going to peel the onion and delve deeper into the key technical and strategic considerations of the platform needed to deliver a compelling return for that investment.

At the foundation of many organizations’ attempts to deliver value through analytics is a substantial investment in an enterprise data warehouse (“EDW”). While those investments are often successful in supporting various operational reporting and BI needs, there are often challenges when attempting to drive predictive or prescriptive analytics from those same sources. Very often, new big data/analytics projects lead to the need for the collection of new data sources, refinement of existing data models, and even building new extracts in support of the effort. All of that activity precedes the need for the analytic team to build new models for each new problem and then train and refine the results over time. At the end of the day, these types of efforts often result in “one-off” solutions that don’t provide the scale needed to drive the desired return on investment.

As NHT has evaluated the keys to success in building out an analytics solution, we believe there are three key technological guiding principles to drive ROI:

  1. Extensibility – The process for developing analytic models from existing data sources has to be extensible to a wide variety of business problems without having to repeat the design->extract->model->train process for each new business solution.
  2. Flexibility – In a world of an ever changing market dynamics, regulation and availability of new data sources, solutions need to be able to adapt without requiring a redesign in reaction to every change. New data sources need to be easily incorporated and models need to be adaptable to changing conditions
  3. Cost Effectiveness – Given the substantial investment that most business have already made in their data infrastructure, incremental investment in advanced analytics can give executives pause if there is no clear and measurable path to ROI. Part of that equation is building on top of cost effective, and often open source, solutions.

At NHT, we believe the principles above require a platform, and not simply additional acquisitions for your technical stack. At the heart of our platform is a NoSQL database. The decision to choose NoSQL or RDBMS as a basis for your data platform often inspires semi-religious battles over which is most appropriate. The reality is that NoSQL and RDBMS can be deployed in highly complementary ways that usually extend and enhance an organization’s existing investment in a RDBMS.

The flexible nature of NoSQL platforms is ideally suited to support the ever-changing requirements faced by companies in the health care space. Incorporating new data sources no longer require a complete design -> extract -> load cycle that stifles rapid development. Not to mention that NoSQL is perfectly suited for the longitudinal data structures and aggregations needed to support predictive analytics. These aggregations and longitudinal representations truly enhance and extend, without replicating, the rich data sources often available in an EDW.

Finally, the technical scale and low operational costs associated with a NoSQL platform enable realization of returns on investment that are otherwise a challenge to achieve. The previous blog post of NextHealth’s VP of Engineering, Charles Selvaraj, explores some of the specific drivers of the tremendous cost effectiveness of this choice in platforms. Ultimately, the cost per gigabyte or transaction/second for NoSQL can be many times less than the cost for RDBMS, allowing you to store, process, and maintain far more data at a much lower cost. For example, ConstantContact, a digital marketing company, published that by scaling out with NoSQL vs. IBM DB2, they saved 90% in software costs and implemented the new system in 1/3 the time.

Ultimately, following these three guiding database principles – Extensibility, Flexibility, and Cost Effectiveness – will significantly and positively drive returns on analytics investments. NextHealth continues to prove out these principles in the groundbreaking outcomes we’ve created for our risk-bearing healthcare clients. Instead of waiting years to see returns, our clients see them in a matter of weeks, and they are sustained over time. C-suite executives are demanding speed-to-value for their analytics investments as well as measurable, causal outcomes. Building a platform that delivers on these demands has never been so critical.