Chandran Saravana, Senior Director, Advanced Analytics, SAP, joins SAPinsider Studio at Reporting & Analytics 2015 to discuss SAP's predictive analytics suite. Saravana co-presented at the Reporting & Analytics keynote to shed light on predictive scenarios within SAP Cloud for Analytics.
This is an edited transcript of the discussion:
Ken Murphy: Hi, this is Ken Murphy with SAPinsider, and I’m at the SAPinsider Reporting & Analytics 2015 event, and here with me is Chandran Saravana, who is the Senior Director, Advanced Analytics, SAP. Chandran, thanks for being here.
Chandran Saravana, SAP: Thanks for having me Ken.
Ken: You were a co-presenter at the keynote today. I was hoping you could start by addressing the state of predictive analytics today. Why is it so important? Is the volume or velocity of data outpacing the platforms that organizations are using and have today?
Chandran: Basically Big Data is mainstream right now. It’s no longer a buzz word. Every company has a lot of data they are generating as of today, either inside the enterprise or externally to their enterprise. There are a lot of business contacts that exist within the enterprise and outside as well. So the data is pounding, whether it is velocity, veracity – all of those things are there. Big Data poses two things: challenges and opportunities. The opportunities are humongous. You don’t have a luxury of simply avoiding any of this data, you still need to process it and analyze it even in the case that it is not useful. You have to go through the entire cycle in order to take advantage of it. The challenges are skills gap, especially within the enterprise. They invested heavily on their existing skills, they aren’t leveraging predictive analytics as in the other areas of advanced technology, there’s a lot of uptake in the skills that are required. That is the biggest challenge as we see it in the marketplace.
Ken: And obviously you need to solve that challenge to take advantage of those opportunities.
Chandran: Absolutely. In today’s marketplace, competing in the marketplace you have to leverage all type of data and leverage existing skill sets. That requires a type of platform which is appealing to all types of persons within the enterprise. I categorize three types of personnel: Data scientists, highly skilled labor force, and a business analyst. And the business users. Only 0.3% are data scientists, 3% are business analysts and the rest are purely business users who are consuming this information. So the tool must be appealing to all three types of personnel and leverage their existing skills and take advantage of predictive analytics.
Ken: To take a step back, with those opportunities that you mentioned it’s not difficult these days to prove an ROI for bringing predictive into the organization?
Chandran: Absolutely, and if you look at our customer cases everything is about ROI. All the cases it’s highly quantifiable. People will say they implemented Predictive Analytics for a customer, channel analysis, and they got this much ROI. Everything is tied very highly within ROI, it’s purely an ROI gain with respect to Predictive Analytics.
Ken: We heard in the keynote the Business Value Cycle, we heard about SAP Cloud for Analytics and SAP Cloud for Planning, and you took us through the (SAP) vision for using analytics throughout the entire organization in an integrated fashion. Maybe touch on what sets SAP apart from the field in that aspect?
Chandran: In on-premise as of today we sell predictive solutions, BI, EPM, Financials, GRC. The key here is a lot of companies are addressing pinpoint solutions, addressing particular verticals or one particular solution. Whereas the key value proposition for SAP is bringing all the solution together and available in the cloud, and also the ability for the customer who is already invested in on-premise to operate in a hybrid model. And you don’t need to have all the solutions, but you can go very easily across all the solutions. There are several business processes built in workflows, it’s not just about BI and predictive. So if I’m a marketing specialist I have to run a campaign, I will utilize BI, I would utilize planning, I’d utilize GRC – what is my risk of exposing the market budget? All those things are combined, so that gives a huge opportunity for the LOB users and all the business users who can operate very easily in the cloud. And also it brings collaboration across all the silos within the company.
Ken: To solve the challenges, to take advantage of and capture these opportunities, it’s really on a business-by-business case depending on your industry, your needs, and what you need to solve?
Chandran: Exactly, so if you look at the use cases like in marketing, regardless of industry they all have a customer, they have to grow their customer retention, loyalty. These use cases are very generic across all industries. So when you come to the cloud, it’s an opportunity to very easily address across the business whether you are in marketing, supply chain, or sales you have the ability to collaborate and put together in a solution for your business that can address your (particular) business.
Ken: You mentioned data scientists, and I’m curious because Predictive Analtyics seems complex but you hear about SAP and the Run Simple methodology. Is predictive really going to be available to that average end consumer?
Chandran: Legacy tools are pretty much focused on data scientists that requires statistics, a math background or even PhDs. The whole vision of SAP is democratized predictive analytics across the organization, not for only data scientists – for us, data scientists are also important and we have a tool that can address that. But at the same time we want to go and target the business analyst, the business users. It’s not about just building a very good predictive model; building a predictive model or doing data mining is a team sport that requires a collaboration between data scientists, business analysts, and the end consumer of it. It’s not finally building a predictive model, it’s how you take the predictive model and build vertically into an application business processes so that you have an impact on it. I’ll give an example. Let’s say you have a call center agent and somebody calls to the call center. And the call center agent can offer the customer – usually someone calls to complain about something – but he or she is able to make a next-best offer. Now, that has nothing to do with predictive analytics. But with predictive analytics powered underneath it, the end user is able to consume without even knowing about predictive analytics. That is where we are heading.
Ken: Chandran, thanks for joining us to share these insights.
Chandran: Thanks Ken.