Imagine a world where data doesn’t just sit in spreadsheets but drives smarter decisions and sparks innovation. Across the globe, companies are tapping into advanced forecasting and analytics to make this a reality.
From rethinking forecasting as a global materials science leader to refining processes in the transportation sector, these four real-world stories reveal the extraordinary potential of data-driven decision-making with SAS Analytical Forecasting.
Dow: Empowering innovation with statistical demand planning
Dow, a global leader in materials science, empowers innovation across industries. Josh Ackerman, a Data Science Manager at Dow on the Enterprise Data and Analytics team, is passionate about the value of data and forecasting.
Ackerman’s team supports all of Dow’s businesses and functions by delivering diverse analytics applications. "Our goal is to empower every part of the organization with the tools they need to make smarter decisions," he adds.
One of the key challenges the team faced was improving forecast accuracy for its global demand planning initiative. "We had an existing set of forecasts that were created using our demand management software," Ackerman explains. "Those models were univariate in nature, meaning they used just historical data to predict the future, and the results just didn’t look realistic."
Leadership and business teams at Dow were concerned that these forecasts failed to capture the complexities of today’s market conditions, especially after global disruptions like the pandemic and energy crises in Europe.
"They challenged us: Can we bring in market intelligence, other economic factors, to explain some of the variation and allow us to extrapolate more reliably?" he recalls.
Another critical goal was to reduce the manual effort required to adjust forecasts. "Our marketing and commercial teams had to massage the numbers significantly to make them align with what they believed were realistic economic conditions," Ackerman shares. "We wanted to create a process that could embed that intelligence directly into the baseline model so those teams wouldn’t need to make so many manual adjustments."
The results have been promising. "We’ve had results deployed into our production systems for the last few months, and the feedback from our business partners has been really positive," Ackerman notes.
This innovation has significantly reduced the need for manual adjustments, enabling more accurate and realistic demand plans. "Those forecasts are really used to inform the optimal production plans that we run on a monthly basis," Ackerman explains.
The software also plays a critical role in guiding strategic investment decisions, demonstrating its value in shaping Dow's future. "It’s exciting to see how these tools influence not just day-to-day operations but also long-term strategies," Ackerman reflects.
Carnival: Predicting costs with advanced forecasting
Since 1972, Carnival Cruise Line has been helping guests sail the world. Junior Peña, Manager of Financial Analytics, has led a major shift in forecasting for the cruise line.
With over a decade of analytics experience, Peña replaced outdated, Excel-based methods with a unified platform powered by SAS Forecasting. “The root of the project focused on food and beverage,” Peña explains. “We have lots of areas here at Carnival that we like to predict for, but the main focus was food and beverage.”
"We’ve moved from scattered processes to a system that works for everything – whether it’s forecasting fish, potatoes, pencils, or t-shirts," he explains. This shift dramatically improved scalability. Now, adding new products no longer means starting over.
Another major accomplishment was shifting from a historical analytical approach to a forward-looking system. “In the past, a lot of our analytics were only looking backward... But now, the question is about figuring out how we can predict our costs better.
This is where SAS came in and has been a central tool in helping us get to a better place so we're not just looking back, says Peña. “We're getting a better forecast in place to lower costs ultimately and hit the bottom line.”
Using SAS forecasting also encourages deeper insights, helping teams ask better questions and test assumptions. " That's the part that I really enjoy,” says Pena. “Now, we’re getting the right questions to the right teams for them to really think of forecasting and modeling in a different way."
Hershey: Improving accuracy with holistic forecasts
Known for its iconic chocolate bars, Hershey has grown into a family of more than 70 beloved household brands. Rick Dimon, Senior Data Scientist at Hershey, has reimagined the company’s approach to forecasting. "I’ve been ‘making moments of goodness’ at Hershey for three years, and it’s been an amazing journey," he says.
Dimon’s team oversees forecasts for Hershey’s global volume, which guides supply chain, financial and commercial planning. Recently, the team launched a revamped forecasting and planning process with a new Integrated Dynamic Planning (IDP) platform.
“I think subliminally, the goal of every forecast change is a much more accurate forecast. But at Hershey, we were trying to develop a much more holistic view of forecasting,” Dimon shares. Before this change, siloed demand planning caused inefficiencies. “When I started at Hershey, we had a pretty strong statistical forecasting foundation, and most of our data was there and generally right,” Dimon continues. “But our demand planning group was siloed, and we were forecasting every single individual lane from our internal distribution centers to each individual retailer's warehouse. So, the first step was getting our forecast out of the weeds.”
Dimon and his team led the charge to transition to a national forecasting model in place of individual and siloed forecasts. "Moving to a national forecast removes most of the noise and allows us to be much more accurate," he explains. The new IDP group aligned planners around shared data, shifting the focus from debating numbers to discussing inputs and assumptions.
Rick’s team also frequently updates models to ensure current and accurate information. "We rerun our models on a weekly basis. Now, demand planners aren't changing their forecast every single week, but we're providing them with the best possible starting place every single week," he adds. These advancements have led to smarter strategies and more collaboration across the organization.
Penske: Simplifying complex models to improve monthly planning
Penske is a leader in automotive and truck retailing, transportation logistics and truck rentals. Kwaku Baa, Director of Customer Experience and Analytics at Penske Transportation Solutions, has transformed the company’s revenue forecasting. "My team primarily develops models to explain customer behavior, but today I’ll focus on one built with our finance team," Kwaku shares.
Previously, Penske’s finance team relied on simple Excel models that lacked accuracy and scalability. These models were also univariate in nature, so they only considered past events. However, forecasting based on past events alone didn’t provide confidence in the output or statistical rigor.
Our finance colleagues talk about how our models have really helped them to do their business planning from month to month because they are able to get very accurate forecasts for the next two years.
Kwaku Baa, Penske
To solve this, Kwaku’s team adopted SAS Forecasting Studio to develop time series forecasting to show how individual products affect overall forecasts. “Those who have used SAS Forecasting Studio know that it gives you the benefit of being able to create well-structured hierarchical forecasting models,” explains Baa. “You can narrow your set onto one product, but that product will also fold into the family tree so that you can actually understand how each of these individual products affects the overall forecasting of a higher or a parent product.”
The results have been transformative. "Once we implemented these, the feedback has been very great,” Baa says. “Our finance colleagues talk about how our models have really helped them to do their business planning from month to month because they are able to get very accurate forecasts for the next two years.” Updated monthly and reviewed biannually, the models enable better decisions and long-term strategies. "After three years, we’ve achieved remarkable success," he concludes.
Learn more from Dow, Hershey, Carnival and Penske
From Dow to Hershey, Carnival, and Penske, these stories show how advanced forecasting and analytics are reshaping industries. Josh, Junior, Rick and Kwaku dive deeper into their forecasting journeys in an on-demand webinar – sharing tips and tricks for other companies to use on their forecasting journeys.
During this session, you’ll:
- Understand the diverse applications of SAS forecasting solutions across industries.
- Discover best practices for implementing forecasting tools in your organization.
- Learn how to enhance forecast accuracy and drive better business decisions.
- Explore real-life case studies that highlight the impact of data-driven forecasting.
