KDnuggets Home » News » 2014 » Sep » Opinions, Interviews, Reports » Exclusive Interview: Imran Siddiqi, SAP on the Strategic Value Proposition of Big Data ( 16:n16 )

Exclusive Interview: Imran Siddiqi, SAP on the Strategic Value Proposition of Big Data


We discuss the impact of Big Data advancements on business strategy, value proposition of Big Data, importance of partnerships, key risks and mitigation strategy, how to win sustained patronage for Big Data projects and more.



Imran SiddiqiImran P. Siddiqi is Senior Principal at SAP, where he advises executives on business value creation from technology innovation. He leads strategy engagements with executives, helping them use Advanced Analytics to gain a sustainable competitive advantage in their industry. In addition he is a thought leader and evangelist for big data analytics, writing practical guides for CMOs, CFOs, and the nascent Chief Analytics Officer roles.

Prior, Mr. Siddiqi was at CEB, a best practices research and analysis firm, where he held multiple roles including Senior Director of Strategic Marketing, Chief of Staff to the CEO, and Senior Director of Research. Prior to CEB, he held roles in strategy consulting at Bain & Company and at Kaiser Associates; he started his career in FP&A at Engro Corp.

Mr. Siddiqi is passionate about teaching and education, and is active in the community through nonprofit work, including serving on the Board of Directors of Ingenuity Prep, a charter school in Washington DC.

Here is first part of my interview with him:

Anmol Rajpurohit: Q1. How is the increasing interest in Big Data as well as increasing ability to harness Big Data impacting business strategy? Currently, what is the most important value proposition of Big Data to businesses - cost savings or additional revenue or something else?

Imran P. Siddiqi: While I’m a data scientist-in-training, I’m also a Strategy Consultant overseeing this across multiple industries, so this is always of interest to me.

ValueThe answer varies by Industry and also by organization culture. Some industries have realized that there is a lot of catching up to do – for example in Consumer Products their own industry association the GMA put out a report that shows there’s a lot more that could be done to leverage Big Data in that industry. On the other hand, in Pharma, Capital Markets and Public Sector, Big Data has long had for a significant role to play. At SAP, we have very deep expertise and understanding of each of over 25 different industries, so we are able to work within the industry maturity to help our customers take advantage of these technologies and capabilities.

Although both the Harvard Business Review and McKinsey & Company put out reports a couple of years ago that brought Big Data and Advanced Analytics into the conversation at the CEO and Board level, the response of most organizations has been mixed – in some, the CIO took the mantle and approached this as an IT problem, while in others the Chief Marketing Officer or the Chief Supply Chain Officer have taken it on as game-changers. This is why we wrote our “Getting Started with Big Data” white papers, to fill in the gap between the strategic, business-changing vision that CEOs and Board have, versus what IT and other functions are delivering.

In terms of value proposition to the business – cost savings versus revenue enhancement – ultimately it comes down to what the strategic goals are and how big data can help the achievement of those. One of the most interesting things I’ve seen is the non-tangible benefits that can come as “free gift with purchase” if you will.

For example I’ve got one client that is using big data and predictive analytics for preventive maintenance in their manufacturing plants – but while the cost reduction in their annual plant maintenance spend is in the millions, they are even more excited about the impact on quality and ability to meet delivery times for their customers.

AR: Q2. As a Big Data vendor, what is more important - focusing on one's specific product or delivering an end-to-end solution to the customer through partnerships?

IPS: We are unique in that for us, it doesn’t have to be an either / or question, especially if you really define “end-to-end” correctly. We pride ourselves on being able to truly provide end-to-end solutions, especially through our SAP HANA platform. Partnership

In addition we have one of the largest partner ecosystems in the world, from very deep, niche industry players to those that serve broader markets. Your readers may know there are over 100+ partners that are building extensions on SAP applications. What they might find even more interesting is that there are 1,500+ startups in the SAP Startup Focus program, which are completely focused on big data, predictive analytics and real time data decision solutions built on the SAP HANA platform.

AR: Q3. What are the major common risks that should definitely be included in the planning process for Big Data projects? Any recommendations for mitigating those risks?

Risk MitigationIPS:
• Form follows function: Start with Business outcomes, not with Technology
• Engage the business directly
• Determine which capabilities are core for you to build and which can you outsource
• Plan to get a couple of wins quick, while building up grander “Big Hit” use cases that may have longer lead times before they deliver big value
• Prioritize on 2 dimensions: Business Value and Feasibility
• Take a Pilot, Test, Build and Scale approach
• Don’t forget political support!

AR: Q4. What are the key factors in ensuring sustained patronage to Big Data initiatives in large companies? What are the most important metrics to evaluate the success of Big Data projects?

IPS:

Sponsorship
Patronage is key here. Remember, despite all the press about it, Big Data is not a well-understood field. Most people still don’t know how it actually works, so you have to kind of show them the way.

You need to develop a portfolio of use cases, some of them longer term aspirational ones (those will often be tied to revenue enhancements or business model innovations) but some definitely need to be shorter term “quick wins,” because it’s through delivering on those that you get the credibility, buy-in and sponsorship you need from internal stakeholders and P&L owners.

But we have to be careful – picking use cases is not easy. How do you know what is too ambitious versus not ambitious enough; how do you avoid doing tons of “me too” use cases versus just the right ones needed to get sustained patronage? That’s where our in-house, customer-facing team of over 500 Industry value experts and data scientists can help. Everything we do starts with a maturity model, often within an industry and / or a functional area, so we save time and effort in helping figure out the right portfolio of use cases. We then help quantify the value of each of them and prioritize by juxtaposing quantified business Value against Feasibility.

AR: Q5. How do you contrast BI and Big Data? How can firms assess if they are ready to embark on the Big Data journey?

IPS: I think of the distinction really as being between BI and Advanced Analytics or Data Science, because you can have BI on very large, fast data sets and it’s still reporting, not data science. Both are key to have, and using one over the other depends on the types of QUESTION that you are seeking to answer. Think of it as peeling the layers of an onion, trying to get deeper and more precise.

In many cases BI and reporting is absolutely the most efficient and effective way to get to the answer – for example if you want to know in what regions are sales below plan, and when you know that maybe what products are driving the decline withinBig Data Journey those regions is important also. Well you don’t want to go build a model for that. However if the next question was, “what are the best 3 ways to fix the decline,” that’s where you start to cross over into advanced analytics, because most dashboards can’t tell you that straight off the bat. At best you will be able to get some idea by looking at historical data, putting in your experience and gut feel and then coming up with an answer. So already you are kind of moving into data mining, maybe some light tradeoff analysis. Now let’s peel the onion further and ask, “OK if we deploy approach A – let’s say it’s a promotion on that product targeted at a specific customer segment – then how well is our response going to work?” To answer that, you may need to do segmentation, classification, and so on.

Another way to think about the difference is, how precise do you want or need to be in your answer. Of course as an analyst we always ask the next question and the next one after that and so on, usually until we reach the level of precision we need, or more likely when we run out of data or hypotheses!

The difference between the two is reflected in everything from use cases to tools / architecture / infrastructure, and perhaps most importantly in things like the skill sets you need. Again, our maturity model assessment has all this, in 5 minutes of answering a few questions you can get an instant assessment of where your organization stands. Both BI people and Data science people find it very useful to do the assessment and use that as a way to get on the same page, talk about resources, plans and so on.

Ultimately though, these two disciplines are coming together – I foresee the term “Analytics” will be re-cast in the future to include both the BI side and the data science side, but it will take a while to get there. We are already doing this through the BusinessObjects 4.1 upgrade, and with solutions like SAP InfiniteInsight and SAP Lumira now the non-data scientist BI user can go ahead and take data sets, do guided data mining and some fantastic visualizations; they can share with others because it’s all one single source of the truth, and then run something like SAP InfiniteInsight to start doing guided predictive modeling.

Second and last part of the interview.

Related: