A large part of Dow Chemical’s business is producing the building block chemicals that go into an array of products. These chemicals are highly dependent on raw materials derived from natural gas and oil. In recent years the cost of these raw materials has been volatile, and the profit margins are typically thin. SAS had the opportunity to talk with Brian Ames, Vice President of Olefins, Aromatics and Alternatives in Dow’s hydrocarbons business, and Mark Bassett, Vice President for Dow Oxygenated Solvents, a business in the Performance Materials Group, about how they use analytics to hedge against these price swings and ultimately manage their businesses more profitably.
Can you share some of the benefits from using analytics?
AMES: We haven’t put financial metrics on what the benefits are yet, but anecdotally we know that the [analytical] model is able to collect all the information that we think about and put it into a structured approach so that we can get outputs from all of these different variables on a weekly basis. It helps us understand price direction on propylene [a key oil-derived raw material], and when we have a better understanding on the price direction, then it helps us to manage pricing policy with the downstream products that we sell to our customers. This allows us to maximize value without foregoing volume. In addition, we can now make better decisions about when to buy and sell these products on the open market.
How do you build an analytic culture?
AMES: I think you have to be patient because you can’t just slap a model together and then expect it to work. You have to develop the model so that it is reflective of your business. It is important to involve a lot of people because people learn from putting these models together. As time goes on, the confidence in using the model grows.
Dow uses a center of excellence approach, which houses a number of analytic experts together, rather than dispersing them into the business units. What would you say to business users who have this option?
BASSETT: You need to think about what you are really looking for, how realistic it is, and what you are going to do with it. In our case, our center is very focused, very business-aligned and very creative. But you still need a well-defined problem; you need to understand your key drivers and you need to convey that.
Can you give us some more tips on defining the problem?
BASSETT: You have to think about what you are ultimately trying to accomplish. For us, in one particular case, it had to do with the volatile raw material environment. We realized if we could predict what is going to happen, we could manage our business a lot better. So we identified a few big levers: What is going to happen with costs? What’s going to happen with demand? What are the competitive influences?
Sometimes the answers derived from analytics turn conventional wisdom on its head and are not always warmly greeted by the business unit. What was your unit’s reaction to the models on the propylene chain?
AMES: Everyone welcomes better price clarity on these products. The problem is that nobody trusts it until they have the time to work with it, and so we’re going through that process right now. I’m gaining confidence in the model. The model was first rolled out in late 2010, but we’ve upgraded it recently. So we’re still getting ourselves to a point where we can really know that we have the fully optimized model, and so you’re catching me a little bit early, maybe, compared to others in this particular model that we’ve been working on. But that’s fine. I think we’re just on a path now to get to this point where we can start really showing to people how they can improve their confidence, and obviously we’re spending time with each of the downstream businesses to show them the model and help them to get confidence in it too. So we’re just going through that process still.
Find out more about how Dow Chemical transforms its business with analytics.