Back to Hub

Knowing your supply and demand before it happens is now possible with demand sensing

Forecasting customer demand for goods and services is not simply the territory of the supply chain manager, although it is a crucial operation for them. It is also the foundation for the intelligence on which business-critical assumptions are made by other stakeholders regarding budgets, profit margins, cash flow, capital expenditure, risk assessment, capacity planning and resource allocation.

Traditional demand forecasting attempts to give an organization a prediction on which strategic plans can be made, and clearly that prediction must be as accurate as possible. But there are new ways to achieve more accuracy.

Estimating the demand for goods and services is a science that is becoming more precise with demand sensing. Demand sensing is a more effective approach than traditional forecasting methods because it relies on more up-to-the-minute information for real-time accuracy.

Determining where to position your stock for distribution and sales, what to manufacture, what you buy, at what time, and at what price and quality is changing. Where companies traditionally have relied heavily on historical sales data and regular purchasing cycle information, the need for more precision is being driven by new expectations and rapid market changes.

What’s changing and why?

Effective procurement relies on supplier and spend data. A firm with sales data history of about 18 to 24 months can obtain a demand pattern that can generate projected demand for a product category under normal market conditions.

The problem?

Today’s market and supply chain conditions are not by nature “normal,” or steady. They fluctuate and are at the mercy of economic, climatic, geographical and political influences.

Organizations cannot anticipate every nuance of environmental impact on your sources under traditional forecasting methods, whether that be on the goods, services or suppliers. However, while traditional forecasting has not been 100% accurate, the use of technology today is allowing organizations to make ground-breaking changes when it comes to accuracy.

The introduction of machine learning and machine algorithms in the demand-prediction space are producing data that is more far-reaching, reliable and real-time than ever before. Now, organizations can calculate and quantify exactly how much they need to source without incurring waste.

But reducing waste is not the only benefit: understanding how, where and when to realign the supply source in the face of an anticipated impacting event is a real and competitive advantage.

Before demand-sensing techniques were employing AI and machine learning, a user would choose a single algorithm that could give them the best prediction, but now, powerful machines can run dozens of algorithms in parallel, allowing the system to pick the one best suited for a particular geography, customer, channel or any combination that can run into the thousands.

This power means we can see significant enhancements to forecasts that were not possible even just a few years ago. And this is where we see the biggest change in demand prediction — accuracy.

Demand sensing still needs historical data however on which to base assumptions, such as events and trends at the point of service (POS). What has changed, is the ability to foresee the impact on future supply, from the likes of weather events and economic indicators, and on future demand, from the likes of social media and change in public sentiment toward a product or process.

Public demand can change rapidly, and businesses must be able to respond with equal agility. All of this, obtained from demand sensing technology, is giving certain technology vendors a more holistic view of the market to project real-time and longer-term results tied together with lessons from historical data.

How demand sensing helps

Whatever the disruption, an accurate demand forecasting capability that is integrated into one source-to-pay platform and uses the most up-to-date technologies, can help manage supply and demand.

For example, if your system could know that a hurricane would destroy crops in a specific region that you source from, you can better plan to stock up early, get ahead of price hikes, or seek alternative supplies in other regions, or quickly pivot and develop new plans to push other products out to the market. If the path of the hurricane shifts, you would know to align your suppliers to meet the rising demands. These are significant real-life events and you need to make sure you have the ability to align suppliers with shifting demand patterns. This is the crux of demand sensing and planning — for all moving parts.

Demand sensing not only makes sure our grocery shelves stay stocked and your customers are kept happy, but it can alert you to shifts in market sentiment, such as in retail, where consumer behavior may change. Then you can respond responsibly and quickly.

This is the power that is differentiating providers in the market, pure-play demand planning as part of an integrated supply and demand solution with the ability and flexibility to draw data from market events and integrate them into its forecasting to produce both short- and long-term accurate predictions.

The power of the cloud

Demand sensing that drives buying behavior is available today from many sourcing and procurement solutions, which are now capable of broadly collaborating on order pipelines, product lifecycle management and marketing efforts across an enterprise. But accuracy is key to effective supply and demand management and is not realistic in non-AI-driven solutions, as there are so many variables.

The vendors at the forefront of demand sensing technology are taking advantage of the power of cloud computing. Those who simply host their technology in the cloud do not have the same advantage as those who utilize its computing power to be dynamic and agile. This is an important distinction because a competitive feature of today’s business is having the agility to react better and faster.

The need to run demand sensing increases and decreases with market activity, so the technology you use must be capable of producing varying lengths of outlook and levels of granularity.

For example, a forecast of up to two years will have, and will need, less granularity, but as the timeline for the event approaches, so the level of granularity must increase. Granularity of forecasting should change from months, to weeks, to days. This will in turn change according to the product hierarchy on which you base predictions from category level to product group down to SKU level. Cloud computing has the ability to access more data and respond with speed dynamically and accurately.

For procurement and supply chain organizations looking to align their supply and demand through more finely tuned forecasting, selecting the right technology is key, one that uses the most up-to-date processes and can accommodate these variations.

Making the difference

To stay ahead of the influences on supply and demand, and to be able to align the two as market trends shift, success relies on technology that is modern and integrated in an end-to-end unified platform. Today, the technology must also be flexible. You need to be able to take structured and mountains of unstructured data and transform that into usable and reliable information. Platforms that have this capability seamlessly hard-wired in as the result of organic development give the most stable and reliable experience — as opposed to technology being acquired or bolted-on.

Producing the right products and making them available in the right place at the right time is possible only when we are prepared for uncertainty. Whether from the climate, market trends, public opinion, or political and economic drivers, the ability to procure accurately and cost effectively requires the ability to sense things before they make a positive or negative impact.

Eliminating waste, driving sustainability and responding to public sentiment are the goal. This is the premise for today’s demand sensing, but it must come from one trusted source of truth.