Please describe your organization.
Andrew Fast: Elder Research is a consulting firm – a professional services firm. Companies hire us to provide high-end predictive analytics and data science solutions within their business context. In that, we are software-neutral, vendor-neutral. We work and are partners with all the major statistical platforms. Often, we end up working with the platform and the situation that is best for our client. What we're about is delivering value from data, typically through software systems, to help them realize that value and put it into practice.
Daniel Bailey: For full disclosure, we are a SAS partner, and have been a SAS partner for almost 10 years now.
What is your position and responsibilities?
A: I'm Andrew Fast. I'm the chief scientist and my role is to be a technical leader across the company, but particularly focused on the commercial data science consulting systems.
D: My name is Daniel Bailey, and I am the director of commercial analytics for all Elder Research. I lead the various data science projects in delivering solutions for our commercial clients.
What business challenge were you trying to address with SAS Enterprise Guide?
D: I prepared a case study with regard to a client that we used SAS Enterprise Guide to help with their challenges. We've had several clients use SAS software throughout our history.
A: Enterprise Guide is fit for a number of our clients, both in government and in commercial spaces. Most of our users of Enterprise Guide are people who are newer to analytics, in a sense, not having gone deeply into the deep predictive analytics yet. Typically, they move up to Enterprise Miner when they're looking at higher-level analytics. We've used Enterprise Guide in a lot of different places, for different customers, from typical insurance to regulatory oversight, to marketing and sales kind of approaches.
D: With that, I can tell you a little bit more specifically about a business challenge one of our clients was facing. One of our clients is a large national medical insurance company who's known for being an industry leader in innovating and moving with the new, changing best practices with relation to the health care field. They were looking to create analytic capabilities that help them comply with the Affordable Care Act and everything that's going on now with what it means to be an accountable care organization, looking to be able to tie the payment and care-delivery models together and tie that to provider reimbursement.
As such, they wanted to create a performance-based health care scoring model that tied the payment and care-delivery models together. The challenge was great. They had a lot of data, millions of members, hundreds of thousands of providers, administering more than 50 different networks across the United States.
This model, being performance-based, needed to look at the different longitudinal data stretching during five years of provider history, and then they really needed to make sure, being in the highly regulated healthcare industry, that the solution was repeatable, defensible, and statistically sound. As such, the SAS Enterprise Guide tool provided the capabilities that they needed to tackle this problem. That was the challenge.
Please describe the scope of their involvement in detail.
D: To meet this challenge, we leveraged the SAS Enterprise Guide platform to create an automated custom reporting and operational capability, really leveraging the built-in GUI [graphical user interface] that it provides as well as the extension code notes that give a more robust computer programmer the opportunity to extend the functionality of the platform beyond the GUI.
This allowed the organization to integrate these performance-based scores with other operational systems and provide key workflow reports that were used for determining provider reviews. SAS Enterprise Guide made a lot of user reporting related to their SAS report capability and their SAS web report easy. We needed regular scoring, we needed to leverage a lot of data using the OLAP [online analytical processing] capabilities and the database connection capabilities while providing integration, exposing certain tables, views and integration into other operational systems and supporting the regular reporting to the different business units.
The decision to use SAS had a lot to do with the IT [information technology] support and compliance in a complex regulatory environment like health care.
Could you provide a sense of the size of this initiative in financial terms?
D; We've prepared some dollar amounts, but without getting prior approval with the client, I will not mention them.
Could you share any statistics or metrics from this engagement?
D: With regard to results, the capability that we developed for them with analytic solution has been in production for multiple years now. Of course, we've had a couple different iterations and improved over time, but it truly led to significant health care savings while ensuring better outcomes. The capability has become an integral part of multiple business units' workflows, such as network build out, provider quality management and other business units. The software package has had quite a significant impact on the organization.
How did SAS Enterprise Guide support team perform?
D: The support team has always been outstanding. When we send an e-mail, help comes back quickly. If they couldn't follow up quickly with an email, they'll get on the phone with you and walk you through it, step by step, to understand whatever there is and resolve it.
Sometimes, it's as simple as in you put in the code, and they see the issue with the code and send you a revised set of code to make it work. Their support team has always done well, in my experience.
What distinguishes SAS Enterprise Guide from other solutions?
D: SAS Enterprise Guide has some nice graphing capabilities and flexibility with their graphing, such as the ability to upload their PDFs as well as being in Acrobat so you can get more of a dynamic graph instead of just a static graph. The graphing capabilities were nice.
Is there anything SAS could improve or function differently?
D: It's a GUI platform, but it allows and has the full SAS ecosystem behind it. Allowing an experienced SAS programmer to extend the GUI capabilities, but SAS is a coding language. You have to be able to code in SAS to leverage it. As such, I think with some of the GUI options for algorithms could be increased. A non-programmer only gets what they have put in interface, but you actually have a lot more at your disposal. I think they could add some more options for the different statistical algorithms available to you. They're there, but you have to write SAS code to get to them, as opposed to being able to access them through the user platform.
We have five additional questions. For each of these, we ask that you rate SAS Software on a scale of one to five, with five being the best score. What would you give the software for desired functionality of the features available?
D: Four. I think they give a lot of functionality. I don't give a five because they could provide users even more, but not quite as much through the GUI.
What would you give the software for ease of use and implementation?
D: Four. They have a nice workflow – pretty easy to use and intuitive. We didn't install it for our client, but traditionally the setup of SAS has been relatively hard. When we used the implementation it wasn't exactly easy. However, they've made a lot of strides in the last year or so, simplifying their deployment and implementation.
For support and responsiveness of the team or resources available?
D: Five. I've always felt like their support has been tremendous.
Overall, satisfaction with the platform?
How likely are you to recommend the software to a colleague or similar business?
D: Four. SAS's platform is good for complex regulatory environments. They do a lot to make sure that you're in compliance with the environment you're working in. Specifically with this tool being GUI, it made a lot of sense since we were working with not just analytics professionals within the company, but we also had business analysts that were using it. Often, these kind of users didn't have as much of a statistical background. GUI made it easier for business analysts to interface with the tools.