Demand Sensing – NextGen Supply Chain Technology

All of you must have heard the famous English premise, “If the past is the best predictor of the future, it follows that the very recent past is the best predictor of the very near future” Haven’t you?

It fits perfectly in the context of forecasting future demand. Top management is forecasting demand by traditional methods like time series by aggregating past trends of various parameters like Product attributes, geography, age group, price segment, etc. This process is carried out by time series methods and other traditional demand forecasting tools. It looks accurate on paper as an aggregation of more parameters increases the accuracy of Forecasting naturally.

https://blogs.worldbank.org/opendata/true-demand-data">True Demandhttps://en.wikipedia.org/wiki/Demand_sensing">However, it is not showing the actual picture of “True Demand” – The demand for the products and services that consumers truly want. It depends on many real-time events which cannot be addressed in the traditional method of Demand Forecasting. Hence, it is required to add data of current events in the Forecasting technique, to forecast the near future demand. This is possible with the process of “Demand Sensing”

Under “Demand Management Trilogy”, I have already published two articles. The first article is on “Demand Planning” wherein you can understand it very comprehensively and the second article is on “Demand Forecasting” which is a buzz word in the Sales and Marketing industry and give insights to top-management for all important business decision making which helps the business to meet future customer demand. Despite both these concepts are still relevant and provide insights to C-level executives to make strategic decisions, both are missing the capability to address the near future demand which is dependent on current and real-time events. To incorporate the capability of anticipating near future demand in the Traditional Forecasting Method, a concept of “Demand Sensing” has evolved and it is the topic of the third and final article of “Demand Management Trilogy”

WHAT IS DEMAND SENSING?

It is a methodology and technology that leverages new mathematical techniques and current real-time information to create an accurate near-future forecast (whatever may be of hours or days depends how the dynamic supply chain is) of demand, based on current realities of the supply chain.

Demand Sensing is fundamentally different in forecasting as it uses a much broader range of demand signals including current data from the supply chain and various mathematical models to create a more accurate forecast that responds to a real-world event such as Market Shifts, Weather Changes, Natural Disasters, Consumer Buying behaviors, etc. Demand sensing uses technology that takes huge amounts of data and recognizes patterns, so supply chains have actionable signals and can make accurate decisions. It helps companies understand consumer behavior and variables by synthesizing big data in real-time.

Demand Sensing is focusing on “Customer-centric” Supply Chain – What does the customer want and when? The prime motive of customer-centric Supply Chains is to deliver high service levels with optimum stock levels.

GOAL OF DEMAND SENSING

The goal of Demand Sensing is to helps planners to make short-term decisions based on what just happened hours or days ago and not what happened last year. It overcomes the latency issues associated with the traditional time-series statistical methods. In Short, Demand Sensing focuses on eliminating supply chain lag by relentlessly reducing the time between events and the response to those events, more specifically, the goal is to reduce the total time elapsed from the emergence of a statistically meaningful mix of demand signals to the planner’s ability to respond intelligently to those signals.

HISTORY OF DEMAND SENSING

Historical data series are by nature disconnected from current events that affect demand in unpredictable ways – a financial downturn or recovery, a spike in energy prices, an outbreak of disease, Regional unrest, or natural disasters. Even changing weather patterns such as cold snaps or heat waves alter consumer demand from historical patterns.

Adding Current Data:

Hence, the inclusion of current data signals from throughout the supply chain and new mathematics to sort through the masses of data and determine what is predictive.

Inclusion of data from retailers, such as promotional data (e.g. items, prices, sales), launch data (specific items to be listed/delisted, ramp or down plans), and inventory data (e.g., stock levels per warehouse, sales per store). By taking into account data from across the value chain (potentially through collaborative supply chain management and planning), manufacturers can smooth spiky order patterns. The benefits of doing so will ripple through the value chain, helping manufacturers to use the cash more effectively and to deliver a higher level of service. Best-in-class manufacturers are also accelerating the frequency of planning cycles to synchronize them with production cycles. Indeed, some manufacturers are using near-real-time data to adjust production.”

PROCESS STEPS TO START DEMAND SENSING:

Demand Sensing requires to follow the below-mentioned fundamental steps in terms of data, people, technology, and skills.

1. Collect last 3 years of accurate data
2. Identify and list down the numerous external factors that could impact your product demand in the future such as climate, economy, population etc.
3. If you are working on 3 years past data, then use the first 2 years data to identify variables or events and its outcomes which may affect the real-time demand for 3^rd year.
4. Bring in the databases of forecasts for the future that was made at that time and check the deviations along with the 3^rd year’s actual sales to measure how accurate the forecast is.
5. Take your historic forecasts and assess how accurate they turned out to be, which gives you the deviations from the forecast – the Bias.
6. Apply that Bias to your most recent predictions.
7. Devise an algorithm between the past and the future to create your first sensing model.
8. Apply the same and compare with real output. Keep refining till you reach to satisfactory outcome with the defined model.

Once Demand Sensing model will be ready for use, follow below method in day to day routine work.

1. Import short-term demand data on an hourly/daily rather than a weekly or monthly basis
2. Immediately sense demand signal changes as compared to a detailed statistical demand pattern
3. Evaluate the statistical significance of the change
4. Analyse partial period actual demand and execute short-term forecast adjustments using automated routines
5. Identify and rapidly react to replenishment issues or sudden changes in customer demand via advanced statistical analytics

REQUIREMENT OF SUPPLY CHAIN CAPABILITIES TO START DEMAND SENSING

Every Supply Chain Solutions should have the following capabilities to start Demand Sensing process:

1. The supply chain shall be able to model demand at the most atomic level and for the shortest time period – Items/Ship-to locations/Daily – Ship-to locations, sell-in channels, geographical territory, etc.
2. The ability to model demand variability which can understand and segregate the relevant data from noise (Irrelevant data)
3. The supply chain shall be able to capture and use downstream data like ship-to data, VMI feeds, POS data, collaborative planning, etc.
4. The supply chain shall be able to model the impact of external variables like the weather forecast, economic conditions, oil price, or similar causal factors into demand forecasting to predict short-term demand. It makes quick repositioning of inventory possible.
5. The system should have big data processing ability to address data from thousands of item-location combinations per hour or per day based on the type of industry and model applied.
6. To get the integrated networking benefits, the platform must support a seamless environment between planning and execution, as well as the ability to replenish the high-frequency demand signals with optimized execution.
7. Increased process automation is required to ensure that the resulting demand signal used to drive the execution environment does not require significant amounts of manual effort.

To start the Demand Sensing process, you need Ship-to information on distribution, replenishment, or sales order are key data feeds along with respective line-order details. Most manufacturing and distribution companies don’t have retail POS data, but sensing demand using sell-in data has been found to improve forecast accuracy by 30%.

DEMAND SENSING – ARTIFICIAL INTELLIGENCE AND 21^ST CENTURY SUPPLY CHAIN

Demand sensing combines big data and artificial intelligence to push supply chain planning into new levels of capability. In this ever-changing and highly connected world, the only way to adapt and thrive is to provide the right product or service, at the right time, in the easiest, fastest, and most personalized way possible for your customer without overstocking.

The most elusive and important change in the supply chain will be the ability to predict the future accurately – predictive demand. Traditional planning and forecasting techniques are slow, unresponsive, relatively inaccurate, and based on legacy technology whose days are gone.

“Lean culture” is now the history and time of the ‘Now Culture’ has arrived in Supply chain strategy. It shall support demand unpredictability and strategy should cover the aspects from the bottom line to employee scheduling, so it makes sense to focus on your ability to sense demand.  

As the field of business forecasting develops, technological improvements allow companies to experiment with more advanced concepts. These include Machine Learning, Artificial Intelligence, IoT, Voice Enablement, and now, Demand Sensing. This approach has been around for more than a decade but has been steadily gaining interest as the significance of accuracy and insight increase. Because this is the competitive battleground, even leaders in forecasting accuracy must look at new opportunities to maintain their dominant position as #1 in market share, efficiency, and cost-savings.

While the term is not new, its application in the era of AI and big data is undoubtedly pushing new boundaries

Value creation through demand sensing

It is a phenomenal opportunity for the supply chain to further its case as a revenue generation activity rather than a cost center in the business.

Demand sensing gets you on the path to predictive demand by taking historical data, combined with future predictions of external factors. This really is predicting the future – using years of data derives an algorithm with an output that tells you what demand will be in the future, given a set of circumstances.

For supply chain planners, this allows their role to shift from batches, manual activities, and Excel spreadsheets, to strategic and data-driven scenario planning for the future. For the consumer, this goes even further, because it allows us to be in stock at the right time at the right location and to avoid the damaging implications of out of stock, which leads to consumers buying another brand, or leaving your portfolio of brands. Anyone in the supply chain, sales, or marketing world, would agree that if you know an accurate precision of demand, you will be at a competitive advantage forever.

HOW DEMAND SENSING HELPS IN SUPPLY CHAIN

Demand sensing uses technology that takes huge amounts of data and recognizes patterns so supply chains have actionable signals and can make accurate decisions. It helps companies understand consumer behaviour and variables by synthesizing big data in real-time. Data sets may include shifting weather patterns, point-of-sale data streams, competing prices, social media, and economic indicators.

In the age of digital transformation, demand sensing has been adopted by some supply chains that recognize the need for new solutions for advanced inventory planning. Demand sensing software is a set of forecasting models that can complement traditional forecasting-related techniques to create a broader framework of forecasting.

Building the right skills and teams

Demand sensing requires significant upskilling. It’s a new science for a supply chain that’s being taught in some of the best universities of Supply chain, but finding the skills is a challenge. It requires data scientists, people with deep supply chain and demand planning backgrounds and strong business sense to understand the markets where these products play, so you can ensure you capture all the external factors that influence your brand.

It needs cross-functional teams that draw skills from across the business because to make this work as a revenue-generating machine. It’s another example in the modern era of significant technical skills and supply chain skills coming together as organizations have to transform their IT to support the supply chain.

DEMAND SENSING CAPTURES OF A FULL RANGE OF DATA

Demand sensing is mainly depending on collecting and processing data. For this purpose, it collects structured as well as unstructured data, the same way also collects internal as well as external data.

Certainly, not all the data in this grid applies in the same way to every sector, region, and stage in the product life cycle or consumer type. Its influence might change over time or apply differently in different contexts.

This underlines the fact that without understanding context, you cannot properly apply the data. But once you understand the context, that understanding serves as a key foundation for other activities as well, such as the design and implementation of new data-driven business models.

Demand sensing should be an integral part of a real-time, connected supply chain. It is not a replacement for demand planning, which used to be applied for supply chains — that is, creating forecasts using internal and structured data such as sales history. But demand sensing does enhance demand planning and makes supply chains far more responsive to demand.

BENEFITS OF DEMAND SENSING

• Drives forecast accuracy from the traditional 60% to over 80% at the SKU/shelf/location level to help supply chains to increase efficiencies and customer satisfaction when implemented correctly as part of a comprehensive approach.
• Increase inventory accuracy by up to 20%, and Reduces inventory requirements and transportation costs by optimally deploying the same whether it’s in transit, shipped from the Distribution Centre, cross-docked, flowed through or shipped directly from the manufacturer.
• This jump in forecast accuracy helps companies manage the effects of market volatility and gain the related benefits of a demand-driven value network, including more efficient operations and increased service levels
• A range of financial benefits including higher revenues, improved profit margins, decreased inventory levels, better order performance, and a shorter cash-to-cash cycle time.
• Scales to process the high volumes of data associated with hundreds of thousands of item and location combinations
• The underlying platform supports a seamless environment between planning and execution enables the ability to replenish the high-frequency demand signals with optimized execution
• providing a more responsive framework for supply chains to fulfil demand near-term with precise execution
• Specifically, demand sensing enables automation of short-term planning, freeing up supply chain experts for more mid-and long-term strategic tasks and enabling them to focus on alerts and exceptions (such as unexpectedly running out of stock). Demand sensing delivers the greatest value if its results are used in processes such as sourcing, smart replenishment, dynamic warehousing, and real-time production scheduling.
• In general, companies face both external “pull” movement toward digital transformation (such as more demanding consumer expectations for things such as instant order fulfilment) and technological “push” movement (universal connectivity and availability of real-time data). Demand sensing is well-placed to kick-start digital transformation because it serves both kinds of movement.
• By preparing the ground for a wider digital transformation and enabling a truly connected end-to-end supply chain, demand sensing not only delivers substantial benefits in sourcing, manufacturing, warehousing and distribution but also lays the foundation for new data-driven business models with increased consumer interaction.

LIMITATIONS OF DEMAND SENSING

As such Demand Sensing is a complementary technology to Demand Planning and having no limitations if adopted and programmed properly. However, blindly relying on the same without applying logic may lead to a wrong decision making and one will incur a major mistake.  Demand sensing can send false signals with over-sensitive models that subsequently trigger overreactions.  Some element of human management is required, in addition to a comprehensive view of the overall forecasting framework.

False signals can come from:

• Weak data sets: Corrupted data, independent events, small sample sizes, and historical drivers that don’t relate to future demand
• Non-linear drivers: Weather changes, market variables, and promotions
• Overconfidence: Using forecast models alone

Businesses need a combination of both human logic and demand sensing to ensure they don’t rely on, and make decisions solely based on, forecasting models.  Additionally, businesses should not rely on on-demand sensing alone. 

It is certain that with the right use of technology and processes in place, Demand Sensing will improve short-term forecast accuracy. But forecast accuracy plays a big role in service levels only for fast-moving items. For slow-moving, highly volatile items, safety stock modeling becomes more important than forecast accuracy in meeting service level targets. It cannot be satisfactorily addressed by Demand Sensing.

OBSTACLES TO DEMAND SENSING

Nevertheless, some obstacles prevent many companies from moving to demand sensing technology.

1. Company leaders are simply unaware of the value that Demand Sensing can provide. For such businesses, the demonstration with a small prototype model is important to make them aware of benefits.
2. Some companies are wary of trusting automated systems as they feel their data might be at risk or having whim that technology never replaces human intervention. However, even trained professionals are limited in the extent to which they can process vast amounts of data and deal with the factors that must be considered when planning demand for the supply chain. Such complexity is something at which machines excel without fatigue as like humans. A side-by-side test is a good way to develop trust. Compare the results of your analytics model with your human forecasters’ results in an area where historical data is available. In any case, your system can always be overruled by your human planners. You can compare the system’s predictions again later to see which were more accurate. Crucially, the automated system will learn from its experience and human input, tweaking the algorithm and becoming more accurate over time.

CONCLUSION

Companies that use the Demand Sensing model as a part of their Demand Forecasting strategy may certainly realize that they are more prepared with a fully enabled real-time connected Supply Chain and can get benefitted from artificial intelligence, machine learning, and new data-driven models with optimized consumer interaction.

With ever-increasing dynamism in demand, one needs to be clear about what is on the radar to improve; forecast accuracy or service levels (order fulfillment performance). For the latter, look for an integrated solution that can both model/sense demand and also models safety stock on the same platform based on service level-oriented policies (stock to service). Just like Demand Sensing, modeling of safety stock has to be a dynamic ongoing process rather than a static quarterly process, because your long-tail products are as important or becoming more important than your fast-moving items.

Over the next few years, any company that wishes to maintain or expand its current market position will need to embrace a connected, real-time supply chain enabled through demand sensing. The pace of change is too fast and the number of influencing factors too great for traditional models, which rely on the statistical analysis of historical data, to remain serviceable.