Wednesday, June 18, 2014

Big data meets the supply chain — SAP’s Supplier InfoNet and Ariba Network combine to predict supplier risk

The next BriefingsDirect case study interview explores how improved visibility analytics and predictive responses are improving supply-chain management. We’ll now learn how SAP’s Supplier InfoNet, coupled with the Ariba Network, allows for new levels of transparency in predictive analytics that reduce risk in supplier relationships.

Listen to the podcast. Find it on iTunes. Read a full transcript or download a copy. Sponsor: Ariba, an SAP company.

BriefingsDirect had an opportunity to uncover more about about how the intelligent supply chain is evolving at the recent 2014 Ariba LIVE Conference in Las Vegas when we spoke to David Charpie, Vice President of Supplier InfoNet at SAP, and Sundar Kamakshisundaram, Senior Director of Solutions Marketing at Ariba, an SAP company. The discussion is moderated by me, Dana Gardner, Principal Analyst at Interarbor Solutions.

Here are some excerpts:

Gardner: We’ve brought two things together here, SAP’s Supplier InfoNet and Ariba Network. What is it about these two that gives us the ability to analyze or predict, and therefore reduce, risk?

Charpie: To be able to predict and understand risk, you have to have two major components together. One of them is actually understanding this multi-tiered supply chain. Who is doing business with whom, all the way down the line, from the customer to the raw material in a manufacturing sense? To do that you need to be able to bring together a very large graph, if you will, of how all these companies are inter-linked.

Charpie
And that is ultimately what the Ariba Network brings to bear. With over 1.5 million companies that are inter-linked and transacting with each other, we can really see what those supply chains look like.

The second piece of it is to bring together, as Sundar talked about, lots of information of all kinds to be able to understand what’s happening at any point within that map. The kinds of information you need to understand are sometimes as simple as who is the company, what do they make, where are they located, what kind of political, geopolitical issues are they dealing with?
What we find is that suppliers don’t behave the same for everybody.

The more complex issues are things around precisely what exact product are they making with what kind of requirements, in terms of performance, and how they’re actually doing that on a customer-by-customer basis. What we find is that suppliers don’t behave the same for everybody.

So InfoNet and the network have come together to bring those two perspectives, all the data about how companies perform and what they are about with this interconnectedness of how companies work with each other. That really brings us to the full breadth of being able to address this issue about risk.

Gardner: Sundar, we have a depth of transactional history. We have data, we have relationships, and now we’re applying that to how supply chains actually behave and operate. How does this translate into actual information? How does the data go from your systems to someone who is trying to manage their business process?

Kamakshisundaram
Kamakshisundaram: A very good question. If you take a step back and understand the different data points you need to analyze to predict risk, they fall into two different buckets. The first bucket is around the financial metrics that you typically get from any of the big content providers you have in place. We can understand how the supplier is performing, based on current data, and exactly what they’re doing financially, if they’re a public company.

The second aspect, through the help of Ariba Network or Supplier InfoNet, is the ability to understand the operational and the transactional relationship a supplier has in place to predict how the supplier is going to behave six-to-eight months from now.

For example, you may be a large retailer or a consumer packaged goods (CPG) organization, maybe working with a very large trucking company. This particular trucking company may be doing really well and they may have great historical financial information, which basically puts them in a very good shape.

Financial viability

But if only one-third of the business is from retail and CPG and the remaining two-thirds comes from some of the challenging industries, all of a sudden, operational and financial viability of the transportation supply chain may not look good. Though the carrier's historical financials may be in good shape, you can’t really predict how the supplier is going to have working capital management in terms of cash available for them to run the business and maintain the operation in a sustainable manner.

How does Ariba, Ariba Network, and InfoNet help? By taking all the information across this multitude of variables, not only in a financial metrics, but also the operational metrics, and modeling the supply chain.

You don’t limit yourself with the first tier or second tier, but go all the way to the multi-tier supply chain and also the interactions that some of these suppliers may have with their customers. It will help you understand whether this particular supplier will be able to supply the right product and get you the right product to your docks at the right time.

Without having this inter-correlation of network data well laid out in a multi-tier supply chain, it would have been almost impossible to predict what is going to happen in this particular supply-chain example.

Gardner: What sort of trends or competitive pressures are making companies seek better ways to identify, acquire, or manage information and data to have a better handle on their supply chains?

Kamakshisundaram: The pressures are multifaceted. To start with, many organizations are faced with globalization pressure. Finding the right suppliers who can actually supply both the product and service at the right time is a second challenge. And the third challenge many companies grapple with right now is the ability to balance savings and cost reductions with risk mitigation.

These two opposing variables have to be in check in order to drive sustainable savings from the bottom line. These challenges, coupled with the supply-chain disruptions, are making it difficult not only to find suppliers, but also to get the right product at the right time.

Gardner: When we talk about risk in a supply-chain environment what are we really talking about? Risk can be a number of things in a number of different directions.

Many variables

Kamakshisundaram: Risk, at a very high level, is composed of many different variables. Many of us understand that risk is a function of, number one, the supply. If you don’t have the right supplier, if you don’t have the right product at the right time, you have risk.

And, there is the complexity involved in finding the suppliers to address needs in different parts of the world. You may have a supplier in North America, but if you really want to expand your market share in the Far East, especially in China, you need to have the right supply chain to do that.

Companies traditionally have looked at historical information to predict risk. And this is no longer enough because more and more, supply chains are becoming complex. Supply chains are affected by the number of globalized variables including the ability to have suppliers in different parts of the world, and also other challenges which will make risk more difficult to predict in the long run.

Gardner: Where do you see the pressures to change or improve how supply-chain issues are dealt with, and how do you also define the risks that are something to avoid in supply-chain management?

Charpie: When we think about risk we’re really thinking about it from two dimensions. One of them is environmental risk. That is, what are all the factors outside of the company that are impacting performance?

That can be as varied as wars, on one hand, right down to natural disasters and other political types of events that can also cause them to be disrupted in terms of managing their supply base and keeping the kind of cost structure they are looking for.

The other kind are more inherent operational types of risks. These are the things like on-time performance risk, as Sundar was referring to. What do we have in terms of quality? What do we have in terms of product and deliverables, and do they meet the needs of the customer?

As we look at these two kinds of risks, we’ve seen increasing amounts of disruption, because we’re in a time where the supply chains are getting much longer, leaner, and more complex to manage. As a result of that, you’re seeing that over 40 percent of interruptions right now are caused by interruptions in the supply chain downstream, tiers two, tier three, all the way to tier N.

So now we need a different way of managing suppliers than we had in the past. Just working with them and talking to them about how they do things and what they do isn’t enough. We need to understand how they’re actually managing their suppliers, and so on, down the line.

Predicting risk
These are models that behave more like the human brain than like some of the statistical math we learned when we were back in high school

Gardner: So, David, it sounds to me as algorithmic or as if a score card is there to generate this analysis. Is that the right way to look at this, or is it just making the data available for other people to reach conclusions that then allows them to reduce their risk?

Charpie: There absolutely is an algorithmic component to this. In fact, what we do in Supplier InfoNet and with the Ariba Network is to run machine-learning models. These are models that behave more like the human brain than like some of the statistical math we learned when we were back in high school and college.

What it looks for is patterns of behavior, and as Sundar said, we’re looking at how a company has performed in the past with all of their customers. How is that changing? What other variables are changing at the same time or what kinds of events are going on that may be influencing them?

We talked about environmental risk a bit ago. We capture information from about 160,000 newswire sources on a daily basis and, on an automated basis, are able to extract what that article is about, who it’s about, and what the impact on supply chain could be.

By integrating that with the transactional history of the Ariba Network and by integrating that with all the linkage on who does business with whom, we can start to see a pattern of behavior. That pattern of behavior can then help us understand what’s likely to happen moving forward.

To make it a little more concrete, let’s take Sundar’s example of a company having financial trouble. If I take a company, for example, under $100 million, what we have found is that if we see a company that begins to deliver late, within three months of that begins to have quality problems, and within two months or less begins to have cash-flow problems and can’t pay their bills on time, we may be seeing the beginning of a company that’s about to have a financial disaster.

Interestingly, what we find is for the pattern that really means something, after those three events. If they begin paying their bills on time all of a sudden, that’s the worst indicator there possibly could be. It’s very counterintuitive, but the models tell us that when that happens, we’re on the verge of someone who will go bankrupt within two to three months of that time frame.

Delivery model

Gardner: Now I can see why this wasn’t something readily available until fairly recently. We needed to have a cloud infrastructure delivery model. We needed to have the data available and accessible. And then we needed to have a big data capability to drive real-time analysis across multiple tiers on a global scale.

So here we are, Ariba LIVE 2014. What are we going to hear when can people start to actually use this? Where are we on the timeline for delivery in this really compelling value?

Kamakshisundaram: Both Supplier InfoNet and Ariba Network are available today for customers, so that they can continue to leverage these solutions. With the help of SAP’s innovation team, we’re planning to bring in additional solutions that not only help customers look at real-time risk modeling, but also more of predicted analytical capability show.
They can identify the suppliers they want to track to as many as the entire supply base.

Charpie: In terms of the business benefits in what we are offering, the features that really bring to life this notion of integrating the Ariba Network with InfoNet are, first and foremost, an ability to push alerts to our customers on a proactive basis to let them know when something is happening within their supply chain and could be impacting them in any way whatsoever.

That is, they can set their own levels. They can set what interests them. They can identify the suppliers they want to track to as many as the entire supply base. We will track those on an automated basis and give them updates to keep them abreast of what’s happening.

Second, we’re also going to give them the ability to monitor the entire supply base, from a heat-map perspective, to strategically see the hot pockets -- by industry, by spend, or by geography -- that they need to pay particular attention to.

Third, we’re also going to bring to them this automated capability to look at these 160,000 newswire sources and tell them the newswires that they need to pay attention to, so they can determine what kind of actions can they take from those, based on the activity that they see.

We’re also going to bring those predictions to them. We have the ability now to look at and predict performance and disruption and deliver those also as alerts, as well as deeper analytics. By leveraging the power of HANA, we’re able to bring real-time analysis to the customer.

They have those tools today, and so it’d be creating a totally personalized experience, where they can look at big data, look at it the way they want to, look at it the way that they believe risk should be measured and monitored, and be able to use that information right there and then for themselves.

Sharing environment

Last, they also have the ability to do this in an environment where they can share with each other, with their suppliers, and with others in the network, if they choose. What I mean by that is the model that we have used within Supplier InfoNet is very much like you see in Facebook.

When you have a supplier and you would like to see more of their supply base you request that you can see that, much like friending someone on Facebook. They will open up that portion -- some, little, none -- of their supply base that they would like you to be able to have access to. Once you have that, you can get alerts on them, you can manage them, and you can get input on them as well.

So there’s an ability for the community to work together, and that’s really the key piece that we see in the future, and it’s going to continue to expand and grow as we take InfoNet and the Network out to the market.
Focusing on a certain industry and having the suppliers only in that particular industry will give you only a portion of that information to understand and predict risk.

Kamakshisundaram: If you take a step back, you can see why companies haven’t been able to do something like this in the past. There were analytical models available. There were tools and technologies available, but in order to build a model that will help customers identify a multi-tier supply chain risk, you need a community of suppliers who are able to participate and provide information which will continue to help understand where the risk points are.

As David mentioned, where is your heat map? What does it say? And also, point to how you not only collect the information, but what kind of mitigating processes you have to put in place to mitigate those risks.

In certain industries, we see certain trends, whether it’s automotive or aerospace. A lot of the suppliers that are critical in these industries are cross-industry. Focusing on a certain industry and having the suppliers only in that particular industry will give you only a portion of that information to understand and predict risk.

And this is where a community where participants actively share information and insights for the greater good helps. And this is exactly what we’re trying to do with the Ariba Network and Supplier InfoNet.

Gardner: I’m trying to help our listeners solidify their thinking of how this would work in a practical sense in the real world. David, do you have any use-case scenarios that come to mind that would demonstrate the impact and the importance and reinforce this notion that you can’t do this without the community involvement?

Case study

Charpie: Let’s start with a case study. I’m going to talk about one of our customers that is a relatively small electronics distributor.

They signed on to use InfoNet and the Ariba Network to better understand what was happening down the multiple tiers of their supply chain. They wanted to make sure that they could deliver to their ultimate customers, a set of aerospace and defense contractors. They knew what they needed, when they needed it, and the quality that was required.

To manage that and find out what was going to happen, they loaded up Supplier InfoNet, began to get the alerts, and began to react to them. They found very quickly that they were able to find savings in three different areas that ultimately they could pass on to their customers through lower prices.

One of them was that they were able to reduce the amount of time their folks would spend just firefighting the risks that would come up when they didn’t have information ahead of time. That saved about 20 percent on an annual basis.
They needed an independent third party doing it, and SAP and Ariba are a trusted source for doing that.

Second, they also found that they were able to reduce the amount of inventory obsolescence by almost 15 percent on an annual basis as a result of that.

And third, they found that they were avoiding shortages that historically cut their revenues by about 5 percent due to the fact that previously they couldn’t deliver on product that was demanded often on short notice. With the InfoNet all of these benefits were realized for them and became practical to achieve.

Their own perspective on this, relative to the second part of your question, was they couldn’t do this on their own and that no one else could. As they like to say, I certainly wouldn’t share my supply base with my competitor. The idea is that we can take those in aggregate, anonymize them, and make sure the information is cleansed in such a way that no one can know who the contributing folks are.

The fact that they ultimately have control of what people see and what they don’t allows them to have an environment where they feel like they can trust it and act on it, and ultimately, they can. As a result, they’re able to take advantage of that in a way that no one could on their own.

We’ve even had a few of the aerospace and defense folks who tried to build this on their own. All of them ultimately came back because they said they couldn’t get the benchmark data and the aggregate community data. They needed an independent third party doing it, and SAP and Ariba are a trusted source for doing that.

Gardner: For those folks here at Ariba LIVE who are familiar with one or other of these services and programs or maybe not using either one, how do they start? They’re saying, “This is a very compelling value in the supply chain, taking advantage of these big-data capabilities, recognizing that third party role that we can’t do on our own.” How do they get going on this?

Two paths

Charpie: There are two paths you can take. One of them is that you can certainly call us. We would be more than happy to sit down and go through this and look at what your opportunities are by examining your supply base with you.

Second, is to look at this a bit on your own and be reflective. We often take customers through a process, where sit down and look at supply risk and disruption they’ve have had in the past, and based on that, categorize those into the types of disruptions they’ve seen. What is based on quality? What is based on sub-tier issues? What is based on environmental things like natural disasters? Then, we group them.

Then we say, let’s reflect on if you had known these problems were going to happen, as Sundar said three, six, eight months ahead, could you have done something that would have impacted the business, saved money, driven more revenue, whatever the outcome may be?

If the answer to those questions is yes, then we’ll take those particular cases where the impact is understood and where an early warning system would have made a difference financially. We’ll analyze what that really looks like and what the data tells us. And if we can find a pattern within that data, then we know going in that you're going to be successful with the Network and with InfoNet before you ever start.
We would be more than happy to sit down and go through this and look at what your opportunities are by examining your supply base with you.

Gardner: This also strikes me as something that doesn’t fall necessarily into a traditional bucket, as to who would go after these services and gain value from them. That is to say, this goes beyond procurement and just operations, and it enters well into governance, risk, and compliance (GRC).

Who should be looking at this in a large organization or how many different types of groups or constituencies in a large organization should be thinking about this unique service?

Kamakshisundaram: We have found that it depends on the vertical and the industry. Typically, it all starts with the procurement, trying to understand, making sure they can assure supply, that they can get the right suppliers.

Very quickly, procurement also continues to work with supply chain. So you have procurement, supply chain, and depending on how the organization is set up, you also have finance involved, because you need all these three areas to come together.

This is one of the projects where you need complete collaboration and trust within the internal procurement organization, supply chain/operations organization, and finance organization.

As David mentioned, when we talk to aerospace, as well as automotive or even heavy industrial or machinery companies, some of these organizations already are working together. If you really think about how product development is done, procurement participates at the start of the black-box process, where they actually are part and parcel of the process. You also have finance involved.

Assurance of supply

To really understand and manage risk in your supply chain, especially for components that go into your end-level product, which makes up significant revenue for your organization, Supplier Management continues all the way through, even after you actually have assurance of supply.

The second type of customers we have worked with are in the business services/financial/insurance companies, where the whole notion around compliance and risk falls under a chief risk officer or under the risk management umbrella within the financial organization.

Again, here in this particular case, it's not just the finance organization that's responsible for predicting, monitoring, and managing risk. In fact, finance organizations work collaboratively with the procurement organization to understand who their key suppliers are, collect all the information required to accurately model and predict risk, so that they can execute and mitigate risk.

Listen to the podcast. Find it on iTunes. Read a full transcript or download a copy. Sponsor: Ariba, an SAP company.

You may also be interested in:

No comments:

Post a Comment