Interview: The Big Trends in Predictive

I caught up with Shekhar Iyer, Global VP of Business Intelligence and Predictive Analytics for SAP, as he was preparing his presentations for the Gartner BI Conference in Sydney next week, and the SAP Innovation Forum in the UK on March 11th.

What’s your background?

I joined SAP about 7 months ago. Prior to that, I worked for SAS where I was the general manager for the  global center of excellence for Information Management, Analytics, Customer and Risk Intelligence. And before that, I worked for Informatica in the San Francisco bay area, helping them with their acquisition and integration strategy among other things.

Unusually, you’re based in Stockholm. How did that happen?

Yes, I left a beautiful view of the Bay Bridge to come here! The reason is that my wife had an incredible work opportunity, and I wanted to experience Sweden. I love it here, especially in summer — and I’ve survived four winters so far.

What is your role at SAP?

I’m responsible globally for the go-to-market strategy, centers of excellence and global sales specialist teams that are driving the growth for the business intelligence and advanced analytics solutions. My team works closely with the regional/market unit sales specialists and the centers of excellence around the world.

What do you see as the big trends in the BI Market?

The biggest trend over the last few years has been the rise and involvement of the business user/analysts. Business people want to get data, play with data, and get meaningful insights. There’s a tremendous desire for self-service. As dependence on IT has reduced, business people are doing things themselves in a more personalized, agile way.

The second big trend is that business intelligence is becoming less about looking backward and more about predicting and optimizing the future. Business people want to be able to go beyond traditional analytics to a broader ‘big data’ foundation for insights – but they still want something that is easy enough to use without having to spend half their lives getting three PhDs. We want to democratize the use of advanced analytics and make it pervasive!

But people have always wanted self-service. What changed?

It’s a more real-time world — a world where working from gut feel and general trends aren’t enough. People need more precise, objective data for decisions. People want to personalize the insight acquisition process, and as the “time to insight” has become compressed, the heavy dependence on IT has created frustration. And IT departments have suffered cuts and costs constraints, so they have less time and resources to help. There’s only so much they could do, and this has lead to pent-up demand. The trend towards better interfaces with simple consumption on mobile devices has spawned the question “why can’t I have that for my enterprise software, too?” New data exploration and visualization tools like SAP Lumira give business people what they’ve been looking for.

Predictive has been the “next big thing” in BI for at least two decades. What has changed? Why are people finally ready now?

The hype of big data is now actually being translated into business benefits. There’s a greater of level of appreciation for insights and data, and a realization that the volume, size and complexity of today’s data just can’t be handled by traditional means. In addition easier tools are now available that give people access to predictive technology without having to overinvest in new skills.

The success of internet companies such as Google, Facebook, and LinkedIn have helped make this approach more real. A lot of their tremendous success is based on leveraging their data: spotting patterns, making predictions, and adapting their business to take advantage in real-time. They’ve set the trend for mainstream businesses and the next way of adopters are seeing the benefits.

What do companies have to change to take advantage of this?

There has to be a two fold strategy:

1. Make the existing precious data scientists productive and give them solutions that give them speed without comprising quality
2. They need to look at traditional business intelligence and resources, and find people with the desire to upskill to advanced analytics. And take a look at what we’ve done with the InfiniteInsight product from KXEN. You don’t need a PhD – somebody with a mathematical background can take advantage of it. Organizations should identify the right people, get them exposure to these new solutions, and apply the techniques to new business use cases.

Any tips on finding those use cases?

Across all industries, anything to do with customers is a good bet: customer intelligence and the ability to upsell and cross-sell effectively. This need has existed for as long as there have been customers, but you now can do it in new ways, in less time, and leverage real-time platforms to enhance the customer experience in new and differentiated ways.

SAS recently became a partner of SAP. How do you see the relationship between the two?

We live in a world of coopetition. SAS is a partner, a great company, with great solutions. But to go back to your question “why hasn’t predictive been more widely adopted?”– SAS is really for data scientists. We also provide tools for data scientists, and support things like the open-source predictive language R in our platform. But recent studies have shown that the current high demand for data scientists is not going to fulfilled any time soon:  research by the McKinsey Global Institute (MGI) forecasts a 50-60% gap between the supply and demand of people with deep analytical talent by 2018.

So to make good on our promise of democratizing the use of predictive a new more modern self-service type approach is needed. We’re focusing on creating business analyst friendly tools. We want predictive to be more pervasive just as BI has become more pervasive, by making it simpler and more self-service than traditional techniques. For example, we embed predictive in business applications and we’re working on more industry and process use-case accelerators to help avoid the predictive skills shortage and improve time-to-market.

We believe large numbers of customers will use both types of solutions – focused tools for data scientists, but also a new set of solutions for the business people and problems that companies previously had to turn away from because of the lack of advanced analytics resources and skills. It’s like cars – there will always be a use for a stick shift, but having an automatic opens up the driving opportunities to more people with less training.

What are your favorite examples of this trend?

Vodafone Netherlands recently did a video that talked about how they were using the new opportunities. Data mining by telecommunications companies is not new, by any means. But the rapid time to market of the new solutions was really valuable to the business side. They could quickly test different hypotheses, segment customers, and create new products. For example, they created a new data roaming package specifically for Dutch skiers. They were able to quickly pinpoint the best target customers and the right price point. Using previous methods, it could have taken two or three months, by which time the season would have been over!

We’ve also been working with large retailers like Macy’s, to help make sure that customers of their online store can get the right recommendations in a real-time, personalized way. And it’s not just about cross-selling — if they can provide customers with what they want, faster, and show them fewer irrelevant ads, then it improves customer loyalty. They chose to work with SAP because of the short time it took to get a working solution, as well as the ease of use and the quality of the insights. This has been a constant theme from our customers again and again, and that’s why we’re excited…

What do you see looking to the future?

One trend is convergence: how traditional business intelligence, agile visualization, and advanced analytics are coming together. It’s an overused analogy, but it’s a bit like how the iPhone combined a phone, internet access, and music into one simple, easy to use device, and collapsed the inefficiencies.

Next, we need to really embed insight into processes and systems. The easiest to use system is one I don’t need to do anything about. SAP is the leading business process company in the world, and we are embedding things like fraud management, customer intelligence and engagement, real-time product recommendations, call center next-best activity, HR analytics like employee retention, and so one. One great example is the procurement insight we’re introducing to the Ariba network. With trillions of dollars of transactions, a 1% increase in insight for a procurement officer can provide huge value.

By making predictive easy, and embedding into processes, we believe we can help improve both customer-facing and internal processes.

What are you doing on a concrete product level?

First, we’re combining the best of both worlds. We’re combining our predictive solutions like SAP Predictive Analysis that catered to the data scientists with the InfiniteInsight solution to provide simple, easy interfaces for different personas, people with different levels of skills.

Second, we’re embedding industry and process insights into applications.

Third, we’re embedding predictive into the heart of the in-memory SAP HANA platform. We believe this is a huge opportunity to expand the scope of what is possible with predictive analytics. At the same time, we’re continuing our agnostic approach, working with other partners. For example, Teradata recently signed an OEM agreement to include our technology as part of the Apprimo CRM solution.

Final thoughts?

It’s exciting times for us and our customers. SAP is investing significantly in solving big data problems. Predictive is one of the key accelerators that will give the business value in the future, and we can help do this in a simple, easy efficient manner.

Thank you for your time!

 

Upcoming presentations from Shekhar Iyer where you can hear more:

SAP: Is there a New Age of Analytics in the World of Big Data?

Gartner BI Summit, Sydney, 24 February, 2014 (14:15 – 14:45)

Analytics in the new age of big data still hinges off the same basics: faultless data governance, great visualisations, predictive modelling. How do the new trends – industrialised analytics, personal analytics, big data and prescriptive analytics – change the game plan to achieve collective insights across the business? Join SAP GVP Big Data, Business Intelligence and Predictive as he explores the key take-aways from both worlds.

SAP Predictive Analytics – Democratizing Predictive Analytics Usage

SAP Innovation Forum UK, London, 11 March, 2014 (11:45 – 12:10)

We will discuss the SAP strategy around the predictive analytics solution area and how the recent acquisition of KXEN changes the game in the predictive world. We will discuss some customer case studies and how they are using the solution and the business benefits that they are deriving.