As the Virtual SAS® Global Forum 2020 is running live online, we have a guest blog post today from Udo Sglavo, Vice President of Analytics R&D at SAS. Udo is a long time colleague and co-editor (with me and Len Tashman) of Business Forecasting: Practical Problems and Solutions. In this post he examines how the Principles of Analytics, as articulated by SAS EVP, COO and CTO Oliver Schabenberger, apply to forecasting.
Guest Blogger Udo Sglavo: Do the Principles of Analytics Apply to Forecasting?
“As the industry has broadened from statistics and analytics to big data and artificial intelligence, some things have remained constant. I call these foundational truths the Principles of Analytics. They inform our approach to data and analytics, and they manifest themselves in our products and services.”
Executive Vice President, Chief Operating Officer, and Chief Technology Officer
At SAS, the principles of analytics have significantly influenced our approach to data and analytics as a company in general. If we consider forecasting as an activity that fits well under the broader term of analytics, can we claim that these principles apply to it as well? My answer is yes, and I would like to guide you through these principles and explain how they apply to modern forecasting based on analytical thinking.
But first, let’s list the principles of analytics:
- Analytics follows the data, analytics everywhere.
- Analytics is more than algorithms.
- Democratization of analytics; analytics for everyone.
Principle 1: Analytics follows the data, analytics everywhere
Data are a resource. If you are not analyzing it, it is an unused resource. At SAS, we often say, “Data without analytics is value not yet realized.”
The first principle of analytics is about bringing the right analytics technology to the right place at the right time. Forecasters knew big data before it got famous. Early on they faced challenges turning transactional data into formats suitable for analytical methods. In today’s times of smart edge devices, network routers, machines, health care equipment, cars, phones, and more, we have to reconsider the traditional approach of moving all data to a central database. Instead, we can take advantage of the computational power of these devices, and such a “hybrid” approach will be at the forefront of forecasting systems.
As companies recognize the advantages of cloud computing and cloud storage, forecasting ecosystems have to support cloud-native environments natively. An emphasis on data integration, data quality, data privacy, and data security will be at the center of these systems. To provide a seamless experience and help organizations accelerate their cloud transformation initiatives, SAS and Microsoft are working together to ensure that SAS products and solutions can be successfully deployed and run effectively on Azure.
Principle 2: Analytics is more than algorithms
You should pay great attention to the quality, robustness, and performance of your algorithms. But the value of analytics is not in the features and functions of the algorithm – not anymore. The value is in solving data-driven business problems.
If your forecasting team is mostly concerned with competing on which particular algorithm works best for all your data at hand, you have a problem. Yes, algorithms are at the core of forecasting efforts, but they are only a means to an end: to make the correct decision at the right time. Organizations need to adjust the choice of algorithm to the amount of data, the frequency of data, and the forecasting horizon. And of course, this should happen as automatically as possible.
If you are not yet considering forecasting as a process – which starts with accessing and managing data, all the way to feeding downstream systems used for planning and budgeting – you might as well stop using algorithms in the first place.
Solving data-driven business problems can provide the advantage your organization is aiming for, but how can you gain that advantage? Probably not with algorithms alone, especially when they are a commodity. The organization will gain an advantage with enterprise-grade analytics processes that are scalable, flexible, integrated, governed, and operational. These characteristics are just as important as the algorithms themselves.
Principle 3: Democratization of analytics; analytics for everyone
Digital transformation is an ongoing challenge that almost all organizations face. Data and analytics now play a strategic role in digital transformation. But you will not benefit from its impact unless data and analytics can scale beyond the data science team.
In an ideal world, your forecasting processes are fully automated and fit seamlessly into your business processes. Similar to not having to worry about shifting gears in an automatic engine, you should not have to worry about number crunching. However, you may not be able to switch so easily to this approach, as reality kicks in fairly quickly: new products, assortment changes, unforeseen data issues, hardware problems, you name it. The concept of Forecast Value Added is an excellent tool for focusing where to put your efforts. But still, you need to enable analytics skills at all tiers of the organization, especially in those areas that have more domain knowledge that can be applied to analytics.
Making data and analytics available to everyone is crucial for successful forecasting efforts. We can refer to this as “the democratization of analytics,” and it manifests itself in many ways:
- Availability of visualization tools to everyone for ad-hoc analysis.
- Augmented analytics to support users through natural language processing and automation.
- Automation of data management and modeling.
- Analytics and AI as supporting technology.
- Educational programs that broaden analytic skills.
In a world where everyone has data, it’s what you do with that data that matters.
How can your organization differentiate with analytics? One way is to use analytics to identify what data has the most value. For example, what if a data stream is just noise and features no signal at all? This may reveal a flaw in an organizational process, or a gap in the kind of data needed.
When an automatic forecasting engine tells you that a naïve model gives the best forecast, why is this the case? Are our demand patterns truly “almost random,” meaning too complex for a model to detect any pattern?
Maybe you discover that you don’t keep sufficient historical data, or other types of potentially useful data, because of IT restrictions. You can ask what’s the value of keeping additional data? What further analysis could we do, and what better decisions could we make? Organizations should consider additional data feeds, such as social media streams or macro-economic indicators.
Most important, the organization needs to keep asking: Where can we improve with analytics? What markets can we disrupt? Where can we automate and support performance breakthroughs?
The principles of analytics manifest themselves in many ways:
- Analytics applied to areas of the business where it will have the most impact.
- Data and analytics strategies that expand the successful use of analytics projects throughout the organization.
- A culture dedicated to digital transformation and analytical thinking.
- New business opportunities from monetizing data and disrupting existing systems with analytics.
Why do these principles matter? Because even as analytics evolves and industries transform, these principles stay the same. They provide an internal compass that can inform an organization’s approach to data and analytics and fuel its successful digital transformation.