I interviewed Andrea Stroud who discussed APQC’s Recent Analytics in the Supply Chain Research Study.

 

 

 

 

 

 

Today were speaking with Andrea Stroud the Supply Chain  Research Program Manager for APQC. Andrea Can you tell us a little bit about  what you are doing and what are you currently working on regarding APQC recent Analytics in the Supply Chain Research study?

 

Absolutely. Thank you, Dustin. It's always a pleasure to speak with you about our research. APQC is actually a global membership-based organization that focuses on best practices with organizations and bench marking. And what I do is I focus on developing a research program and conducting primary research and published relevant and meaningful content for our members. So we've had some pretty exciting research in the analytics area focusing on analytics in the supply chain.

 

There's been tons of buzz about analytics. It's everywhere, especially with regards to its use in the supply chain. An increasing number of organizations are tapping into the power of analytics to improve performance and identify improvement opportunities. APQC's recent analytics in the supply chain research really looked at organizations that have an analytics department or program and also how analytics is used to help make decisions that impact supply chains.

 

Supply chains are using analytics for a number of activities. So things like creating scoring models for vender and supplier quality cost and stability, detailed demand forecasting at the level of point of sale, and even safety stock level recommendation. This is just to name a few.

 

Even with the manufacturing groups, we see that they're using data to help predict when maintenance is needed for things like refrigeration and air conditioning units. So it's really all very interesting and companies are very excited about it.

 

And do most Supply Chain organization have a formal analytics program or structure?

 

Dustin, that's a great question. From our research that we've conducted, three-fourths of the survey participants for our study report that their organization has a formal analytics program or structure.

 

So typically what we see is about a third have a centralized analytics function where there is like a center of excellence that supports the entire organization. And then we see about another third that have a hybrid of both a centralized and a decentralized function. So they have that center of excellence that supports the entire organization, but then there are these subgroups. For example, procurement may have an analytics group or supply chain planning may have an analytics group that are supporting that particular area as well. And those groups tend to report back up to the centralized analytics group.

 

Then about 15% have a decentralized function. What we notice is the programs and the structures vary depending on the structure of the company and the company size and whatever the needs are. A good example of this really is for oil and gas. Many oil and gas organizations are very large and often have a decentralized organizational structure. These are the organizations that often have that hybrid structure for analytics that includes an analytics center of excellence and a decentralized programs and/or analytics teams throughout the organization.

 

But a problem that is often seen when an organization has a hybrid analytics program or structure,and that's that the groups aren't talking to each other and sharing knowledge about their process. So organizations sometimes end up replicating work in invested-in areas. They don't need to invest in it. If you already have those resources in the organization, you want to make sure that they're being utilized in certain skill-set criteria, training criteria are all being passed down to some of those decentralized groups. So the communication has to happen for it to be successful.

 

Can you tell us what are the challenges or barriers to organization  analytics efforts in the supply chain?

 

Absolutely. In our recent analytics in the supply chain study, we found that supply chain organizations that provided the challenges, that their organizations were experienced. And there were really five top ones that came out. The first one was really about maintaining organizational momentum for analytics activities. These are companies or organizations that have started an analytics program, however to keep people continuing to be interested in it and to find it beneficial, organizations have to keep making sure that they are communicating the benefits of the program and all their resources that are available from it and its uses. So that's the first barrier is, again, that momentum for analytics activities.

 

The second barrier we found was finding the tactical resources. So the people need it to carry out analytics activities. A lot of organizations look outside. But a lot of times they can look right inside their organization. And there are a lot of people that have certain analytical abilities. They may not necessarily be at the most advanced level, but these are people who you can train who already have a certain amount of subject matter expertise. So finding the tactical resources that are needed, you don't always have to go outside. But that is a major barrier for organizations.

 

The third one is executive buy-in, getting that approval and acceptance for analytics activities. It's a challenge because executives want to see... They want to know what those business performance impacts will be for analytics. And you really get that buy-in through pilot programs. But it's very challenging for organizations initially to get the buy-in. But if you start a pilot in either a decentralized group or on a specific project where you're applying analytics, it really helps to get that executive buy-in.

 

And then the fourth barrier that we often see is having the right technology, the tools, the infrastructure in place for analytics to actually happen. Really, part of that comes down to making sure that you communicate with your IT groups and you lay down specifications for the data, for how the data is going to be collected, how you'll receive the data back once it's collected. So that communication piece is very important, and having the right technology and tools is essential.

 

The fifth and final barrier, this came up in our research that financial resources were needed to carry out analytics activities. However, the research also showed that organizations over the past three years have actually been increasing their budget. I think the thing that comes up here with the barrier is that even though you're increasing the budget, you have to show that return on investment in order to continue to get financial resources that are needed for analytics and to keep that analytics program continuing on.

 

What are the most common type of analytics being used within the supply chain?

 

Dustin, supply chain organizations conduct and use three different types of analytics. So there is descriptive, predictive, and prescriptive analytics. So the descriptive analytics really combines your business intelligence with existing data to provide a vision of what's currently happening in the organization. So you typically see things like your mean, median, mode, frequency distributions, or discrete data points, percentile rankings, that sort of thing.

 

Supply chain organizations have reported... This is the most common type of analytics that's being used in the different areas of supply chain, and especially for benchmarking projects. Even though this is useful as an indicator of an organization's current performance in certain process areas, the descriptive benchmarking does not provide information on why performance is what it is and how it can improve. You really find out those things from the predictive analytics as well as the prescriptive analytics.

 

Predictive analytics, however, uses historical data and various algorithms to predict outcomes of various scenarios to help anticipate future events and predict trends. It really uses things like forecasting and statistical models to help to form analytics to judge and provide recommendation about what could occur.

 

When we look at prescriptive analytics, which not as many supply chain organizations are using, unfortunately, except for in the supply chain planning area. We do see a lot more of the prescriptive analytics. But in that area, it uses optimization or embedded decision rules to find out what should happen in a certain situation. So this form of analytics is really the most advanced because it uses insights that are actually gleaned from the prescriptive analytics that have occurred, to recommend business decisions or actions that are likely to produce a specific result.

 

And what enables analytics and an organization?

 

Well, Dustin, that's another great question. APQC's research has shown that there are really four factors that help enable analytics. So the first enabler of analytics is having the necessary data available, and the data has to be a high degree of quality.

 

So using quality data will help ensure that you get a quality analysis that would allow for more accurate interpretation of the data. So without quality data, the analysis is flawed. So receiving data in a timely fashion for timely decision-making is very important. Many people at organizations have to go through so many hurdles to even obtain the data, and then once they get the data, it could potentially not be in a format that they want, or there is something wrong with the data. So there is a lot of question in organizations about the validity of the data that they're using for analysis. But an enabler of a good analytics program is having that accurate and timely data.

 

Another, the second enabler, is having the ability to interpret results then clearly visualize and communicate those results. So you have to have not only the ability to look at the data, to make sense of the data, but if you make sense of it, but you can't communicate it back to the rest of the organization, then it's not very helpful, so you have to really make an effort to communicate that back. And a lot of times, that's actually done through visualization tools or dashboards. A lot of organizations use that to communicate the data and information back so that supply chain managers can make decisions off of it.

 

Then the third — I had talked about technology, tools, and infrastructure earlier, but that's our third enabler. Having that in place is essential, and it requires, as I mentioned before, good communication between the business and IT groups regarding analytics, needs, data needs. It really helps to put the criteria of the data in place.

 

The fourth enabler is around executive buy-in. So analytics can't be supported without executive buy-in. Organizations with effective analytics programs — those are our best-practice organizations — they don't just have executive buy-in, but they have executives who are actively communicating the importance of analytics use within the organization. So it goes back to what I mentioned earlier about keeping that communication of the success and what's going on with the analytics so that everyone in the organization is aware, and they are knowledgeable, and they know what they can use and what they can do with the data.

 

Do you have any practical tips for organization with analytics program?

 

I do. APQC's recent study, as I mentioned, had showed that the majority of participants were reporting in increasing investment on analytics. And even though there was that increase over the past three years of funding for analytics projects and activities, however, when I talked to different organizations and looked at our recent research, it's clear that even though the budgets have increased, many organizations may not be seeing the return on investment that they were hoping for.So I do want to cover some tips that organizations can use to help them see the performance and value of an analytics program.

 

Really, analytics programs will want to make sure that the goals and objectives of their analytics program align with organizational objectives. It is through this alignment that they can ensure that you are collecting the right data and analyzing it in a way that will help supply chain professionals make business decisions.

 

It's also important for organizations to examine their map supply chain processes and look at where analytics fits into their supply chain process. This helps an organization identify any gaps in the process or areas of need. So one thing that we find with a lot of organizations is they haven't mapped their supply chain process, whether it be something like procure-to-pay process or source-to-pay. If those processes aren't mapped, you really can't see where you can include analytics to help support your efforts. So you really have to have those mapped.

 

One other area, another tip, for organizations is having the right resources in place. And this really relies on three realms of expertise to make it successful. So you have to have domain experts who can really define the problem and understand what needs to be solved in the supply chain area.Second, you need the analytics experts who know the limitation and possibilities of analytics. Finally, you need data management experts who know where to get the data and what the data means.

 

Going back to the analytics expert, when we think about that person, we also should consider what level of analytics is needed and being conducted at the organization. Some organizations have invested in data scientists and statisticians but are only reporting descriptive statistics that a data analyst could technically report. So you end up spending a lot more money than is necessary for the program or for what you're doing currently in the program.

 

Organizations have to think about the number of analytics people that are needed to be on a more advanced level of expertise versus lower-advanced level data analyst or statisticians.

 

Organizations can also take someone on a very advanced level and bring other analysts or statisticians within the organization up to a higher level through mentoring. Again, if you hire a statistician or a data scientist, and you have others within the organization who have some knowledge of analytics and some analysis capabilities, you can have that person really train and develop those people to conduct more sophisticated analysis. And I highly recommend that for organizations, because we've seen that to be very effective at best-practice companies.

 

The final and very important point that I want to make is that for supply chain leaders to understand the impact of analytics to its organization, it is important to assess the measures that are being used to evaluate the program and make sure that the right ones have gone chosen.

 

So an organization shouldn't just focus on business performance measures such as revenue, cost, customer [inaudible 00:16:57] and cycle time, even though that is what a lot of our executives like you to focus on. They are important to look at, but that's not the only thing you should be looking at. You also want to look at things like behavioral change measures. These measures typically help an organization monitor the adoption rates or changes in norms and practices within an organization. They also monitor an organization's use of analytics outputs to help support the decision making.

 

Behavioral measures that are often overlooked are the number and types of actions taken, based on analytics. So you use analytics to [inaudible 00:17:45]. Well, you need to record that and note that, because that's really going to help you determine the success of your program.

 

Looking at utilization and consumption or downloads of analytics outputs, managers download in reports based on analysis. That all should be documented and tracked. The number of service requests for analytics projects, that's another one to track and look at, because that tells you, one, are people utilizing the services that are offered through analytics, and also are they finding any benefit. You'd definitely want to track those things.

 

And the final, the number of employees requesting analytics training, whether it be formal or informal, as well as the outcome of that training, that is essential, because if you have more people interested in training for analytics, you will have more people doing some basic analysis and even learning how to do more sophisticated analysis. It really expresses the need and interest within the organization to have that training and to track that information.

 

Thank you for having me, Dustin. It's always a pleasure, and I look forward to sharing more information with you on our research in the future.

 

About Andrea Stroud

 

 

 

Andrea Stroud.jpg

 

Andrea Stroud

 

Research Program Manager APQC

 

LinkedIn Profile