Planning for supply chain risk assessment and mitigation

carloscastilla - Fotolia

Next-gen technology helps mitigate supply chain risks

Supply chain risks can be very disruptive for a company, but technologies like IoT, big data analytics, machine learning and AI can help uncover and mitigate them.

The supply chain is almost like the circulatory system of a company; if all is working well, products are sourced, then produced and shipped in a continuous flow.

Just as the body's circulatory system faces risks from factors that can interrupt the flow of blood, a company's supply chain faces a constant menace of risks that can sometimes interrupt the flow of goods and even threaten the health of the company itself. Discovering and mitigating these supply chain risks should be on the mind of every company, according to a panel at the recent SAP Ariba Live user and partner conference.

Like problems with the body's circulatory system, some supply chain risks are more visible than others, but almost all are difficult to diagnose and treat. This is changing now, however, as new technologies, like internet of things (IoT) sensors, machine learning and big data, are being introduced into the supply chain.

The most important vehicles for such improvements can be systems like the Ariba Network, according to Padmini Ranganathan, SAP Ariba vice president of products and innovation. The Ariba Network contains a great deal of contextual data about business suppliers that can help determine supply chain risks if used properly.

Uncovering hidden risks

"The challenge is not always about the challenge of risk itself, but the challenge of finding that risk," Ranganathan said. "This is because it's hidden inside different processes and systems, and it's fragmented information and data."

The first step to rooting out supply chain risks is to resolve data quality issues and consolidate the data that can help determine risks, Ranganathan explained. The second step is to introduce organizational structures that create transparency and simplicity that enable employees to take the right steps when supply chain risks are suspected or discovered.

"This is not just having the data and doing the right thing with the data or having the right reports, it's about having it at the right place, the right time and with the right person, so you can actually take the right steps," Ranganathan said. "You're not taking long risk assessments when it's a supplier that's already doing well for you, but you want to make sure that you are there to catch it if that supplier slips up even just once."

Technology is a very powerful tool that can bring the two elements together, Ranganathan said. This includes big data and analytics with tremendous capacity to match patterns, and machine learning, which can uncover troubling information about a company that could lead to supply chain risks.

"You can actually find out when there's a board member or senior management leader that's caught in something -- a senior leader going down does impact the performance of the company. So if you can see that upfront, the types of things that you can do to help manage it are going to look different, and technology can help with that," Ranganathan said.

Dealing with the increasing amount of data

The amount of data that companies have to deal with to monitor their supply chains is enormous, and it's getting bigger and more complex all the time. Information is coming in from sources of all kinds -- the media, databases of watchlists or sanctions, for example -- and companies need to clean this data and connect it in context to the buyers and suppliers in their network.

To help with this, SAP Ariba has partnered with OutsideIQ, a Toronto-based company whose DDIQ application performs automated due diligence, reducing the time and resources needed to analyze all that data. SAP Ariba is now using DDIQ as part of its SAP Ariba Supplier Risk service, which monitors suppliers in the Ariba Network to help minimize risk, reduce damage to a company's reputation and meet compliance regulations.

"The challenge now is finding the data, because there's so much out there. But it's also cleaning the data to make sure that it's the right data," Dan Adamson, CEO of OutsideIQ said. "Machines are great at combing through vast amounts of data, but they're very bad at the context part. So our job is bringing in that data, applying the first layer of context to that to make sure that it's the right entity, that it's a risk that you would actually care about, and then how to deal with it is another layer of context."

Context is king

Context is vital, Adamson said, because not all suppliers are equal. Some key suppliers, for example, might come from a heavily regulated industry, and they may need to be treated differently than the supplier that provides your paper clips. OutsideIQ's technology helps to "filter out the noise and bring the right events to bear," Adamson said.

This helps when the risks are present, but not cut and dry. And reputational risk is often quite fuzzy, Adamson said. It's clear if a corporate director is charged with fraud, for example, but less so if they have given possibly misleading statements to the board.

"That then depends on the rest of the sentence and the context, and there's a degree of fuzziness there," he said. "This is a difficult area, and we score a lot of this behind the scenes, and we'll say how sure we are about something with that degree of fuzziness."

More than just visibility into the supply chain

Supply chain risks involve issues that go far beyond visibility into where your product is, or even company reputation, according to James Edward Johnson, director of global supply chain risk management and analytics for The Nielsen Company, a global marketing and data research firm. Nielsen is an early adopter of SAP Ariba Supplier Risk, which Johnson believes can help make the supply chain a force for good in the world by focusing on doing business with ethical and responsible companies.

"How do you make sure that, when you negotiate deals, your push for price isn't merely favoring those people who are going to cut corners, abuse their workers, enslave people, rip up the environment, dump chemicals in the rivers and lakes?" Johnson asked. "If you are only ever pushing for price, you might be making those problems worse, and there will be no incentive in the marketplace for things to get resolved other than government regulation, which is spotty at best, especially in many parts of the world."

Johnson agreed that the sheer amount of suppliers and data make this a difficult problem to solve. Technology, like SAP Ariba Supplier Risk, provides one way to address the problem.

"You just can't do the proper due diligence you need across thousands of suppliers. Just looking up something as simple as the corporate filings at the secretary of state's office is a ridiculous amount of work," he said. "And that's not even going to get you the information you really need to find out what's going on. You need to see information that's coming out from news sources all around the world."

Using technology for peripheral vision

Risk is about seeing the unseen, Johnson said, and to do this, peripheral vision can be more important than central vision. Financial reports are a good leading indicator of supply chain risks of all kinds because, once companies fail to hit their quarterly numbers, they may start to take shortcuts.

"You have to start looking at things that might provoke a company to start engaging in bad behaviors," Johnson said. "If I start to see those ancillary or related risks that I think might be a leading indicator, it will force us to be a lot more aggressive in contract negotiation."

For example, if a supplier starts to show some weakness in integrity, Johnson explained, it might not stop a company from doing business with them, but it may lead to tougher contract negotiations with controls put into place and documented follow-up procedures.

Technology like machine learning can improve your ability to determine reputation risks because it strips out selection biases that may influence reputation scoring, Johnson said.

"Once we move to machine learning, [there still may be] fuzzy answers, but now we can start to say whether or not a comparison is truly comparable," he said. "We can strip out the recency bias, we can strip out some of the selection bias, or just, at least, identify it and say more objectively that this is happening; it's in our data, it's a little biased, but it's biased in this direction by about this much. We should have science-based answers, we should have the data and we should know how well we know what we say we know. "

Next Steps

Global contract manufacturer Jabil developed a supply chain visibility system to mitigate disruptions

Here are some ways to introduce visibility into the supply chain

What are the key factors to include when developing a supply chain risk management plan? 

Dig Deeper on ERP and machine learning