I interviewed Dr. Jozo Acksteiner who discussed Geographic Analytics.

 

 

It’s nice to speak with you today, Jozo, and I’m looking forward today to hearing your views on the topic of Geographic Analytics. Can you first provide a brief background of yourself?

 

Sure, hello, my name is Jozo Acksteiner. I’m Founder & Managing Director of GeoLyx.com, a software firm in the area of Geographic Analytics, and I’m also an Adjunct Associate Professor at National University of Singapore. In total, I have about 20 years in supply chain management and business analytics experience. I previously worked for companies like TNT, DHL, and in consulting for Booz Allen Hamilton.

 

Thank you. What is Geographic Analytics?

 

In essence, it’s a method to visualize data in order to analyze it on the map and to make supply chain decisions more effectively. It’s a big data topic, and for analytics projects with that method, you can increase the effectiveness of making supply chain decisions by more than 50 percent.

 

Why is that? We believe that a lot of the big data power that’s around for supply chain problems cannot really be unleashed because there are some obstacles: Gathering the data; cleaning the data; understanding the data; deriving the results, and interpreting those. Supply chain management problems often have some underlying framing conditions that are very difficult to model with a computer system. E.g. regulatory conditions or even physical infrastructure conditions.

 

For example, you want to find the best location for your distribution center in Europe. There are very good software solutions to help you: You provide data about all your inbound flows, all your outbound flows, all the cost information, what types of goods you’re shipping, and then you plug this into a software and it comes back with a result.

 

When you talk to experts who’ve been using such tools extensively, they often see challenges with these tools. First of all, it takes an awful lot of time to come to get all this data, and to prepare it for the software, before you can actually crunch the data. Secondly, the solution coming out of this black box is sometimes just not feasible.

 

An example: You want to find the best location for a distribution center in Europe. Imagine all flows are equally distributed over Europe and you want to find the one location, such a tool will probably come up with the center of Europe.

 

In this example, that could be Switzerland. Now, do you want to have a distribution center in Switzerland? Switzerland is the only country in Central Europe that has borders, free flow of goods is restricted. So, really, from a logistics perspective, it’s not the most advantageous place. After all this crunching of data, Switzerland would be ruled out because it wasn’t a feasible solution in the first place.

 

This is where Geographic Analytics can help. Instead of trying to get all the data together first, actually, you turn the approach around. You start with reducing the solution space. This means you try to reduce the number of feasible options first.

 

Step 1: Map some basic information and then discuss with subject-matter experts about potential solutions, excluding infeasible solutions. For example, in this case, Switzerland you want to rule out as an option. At the same time you carve out the ideas, building the intuition behind the problem. This is what we call computer-supported intuition. With this you often find approaches that you previously were not even thinking about.

 

Step 2: You deep dive into the analytics. In this way, you do not need to gather the data for infeasible solutions, and you are much more effective with your analysis. As a nice side effect, you’re much better aligned with the experts with whom you’re working because they’re included in the process from the start, and it’s not a black-box solution.

 

Interesting – so you have any other application examples than in Supply Chain Networks?

 

Yes. Generally, Geographic Analytics can be used for geographically related strategic decision-making in supply chain management. Another example is on customer segmentation – to find out where your customers are, and how to service them most effectively.

 

Even outside of Supply Chain Management we see a lot of application for Geographic Analytics. Take for example Marketing analytics – determining where you want to focus your marketing campaigns for example. Visualization of such data on maps and analyzing them on maps can be very valuable.

 

Can you explain some of the other benefits of Geographic Analytics?

 

One re-occurring topic in data analytics is the issue of incorrect or incomplete data. Such data can skew your analysis to the extent that it leads to wrong conclusions. Therefore, data cleaning & understanding is a very important and lengthy process within Data Analytics. We say that the data needs to be transformed from “thin” data – data that you cannot use – into to “thick” data – data that you understand and that you can use for analysis.

 

Spotting data errors in tables is very difficult – here is where the geographic map visualization can help. Let me give you an example: It might be useful to map your service network relative to your customers to determine if your services centers are located at the right spots. Often optimizer software finding the best service center for a customer do this by analyzing zip codes. If the zip code is incorrect, the software does not work correctly.

 

Maybe your software is set up to work with 5-digit numeric zip codes, which works well in the US. But in China you have 6-digit zip codes, and the software may fail as your 6-digit zip code might get truncated.

E.g. a zip code of 320000 gets truncated to 32000, which then would be interpreted in the software as 032000 – which could be at a complete different location.

 

In a sea of data it will be very difficult to spot such problem. But when you see such data on a map and you see a service center far away located from the customer – you immediately see there is something wrong.

 

In summary, the “time to insight” with Geographic Analytics approaches is just much faster than with traditional analytics approaches, particularly in Supply Chain management.

 

This is really exciting. Thank you for sharing today on this great topic.

 

Thanks very much; it was a pleasure.

 

 

About Dr. Jozo Acksteiner

 

 

 

32a2db3.jpg

 

Jozo Acksteiner

 

Manager Supply Chain - Business Analytics - Strategy, Adjunct Associate Professor, Founder

 

LinkedIn Profile