demand forecasting in retail

To achieve optimal fulfilment performance, retailers must now forecast how consumers will shop and understand the influences behind purchasing decisions.

 Forecasting demand for products and positioning inventory has never been a simple task for retailers, but it has become increasingly difficult in the digital age when shoppers’ purchasing behaviour has become varied and unpredictable.

Demand forecasting gives businesses the ability to use historical data on markets to help plan for future trends. As more data on consumers and products becomes available, the need to use this data to anticipate demand is critical for establishing a long-term model for growth.

In a sense, demand forecasting is attempting to replicate human knowledge of consumers once found in a local store. Long ago, retailers could rely on the instinct and intuition of shopkeepers. They knew their customers by name, but, more importantly, they also knew buying preferences, seasonal trends, product affinities and likely future purchases.
Demand forecasting attempts to replicate that sophistication through analytics-based evaluation of available data. By examining buying behavior and other bits of data left behind by the consumer, a retailer can mimic that knowledge on a broader scale.


Anticipating demand for the modern buyer
Today’s consumer often journeys from digital space to physical space and back again, moving among devices, apps and displays. The buying process might start with researching a product online, continue with comparing prices from a mobile device, and finish with an in-store purchase. Or consumers may see merchandise in a store, then search on their phones to score a last-minute deal.
In a world where you can have practically any item shipped to your door, it’s important for retailers to make a connection with the buyer. A variety of buying options is a delight to consumers – and a rich source of intelligence for retailers, if you know how to capitalize on it.
This omnichannel retail environment intensifies the need for better answers to the perennial questions of retail planning. What merchandise should be stocked, in what sizes/colors, at what quantities, in which locations? How, where and when should products be displayed, priced, promoted, ordered or shipped? How can we maximize profit without eroding the quality of the shopping experience and customer satisfaction?
Demand forecasting gives you the ability to answer these questions. But the sheer number of variables involved in the omnichannel world makes demand forecasting and merchandise planning on a global scale highly complex.

Specifically, the winners were the ones who engaged in seven productive habits:

1. They put a high value on analytics. Fully 77 percent of winning retailers rated analytics as “very important” to their retail success, compared to 48 percent of average or lagging retailers. They know they can’t get by without integrating more predictive capabilities into their decision-making processes, and they understand why this is so.
   
   
2. They value more rigorous forecasting. Nearly two-thirds of winning retailers (74 percent) rated demand forecasting technologies as “very important” to their success, compared to 58 percent for the others.
   
   
3. They bring analytics into the process earlier. They know that basing future plans on prior year or season sales will create self-fulfilling prophecies. The lost opportunities of the past will be repeated in the future. Instead, they bring insights from customer analytics into the demand forecasting process upfront, not just as a sanity check at the end of the planning process.
   
   
4. They are optimistic about information. Higher-performing retailers are twice as likely to expect big things from demand forecasting. They are also nearly twice as likely to “highly value” attribute-based merchandise planning systems.
   
   
5. They are ahead of the game in implementation. Winners were twice as likely to have technologies already in place for attribute-based merchandise planning and demand forecasting to help with price, promotional or assortment planning. And they are twice as likely to have integrated their planning, allocation and replenishment systems.
   
   
6. They make smarter allocation decisions. Winners know they need to put products where they are most wanted or needed, and they trust demand forecasting to help them make the best localized store assortments and fulfillment decisions for direct-to-consumer orders.
   
   
7. They blend art and science. While information is seen as a critical asset – along with tangible assets such as stores, distribution centers and inventories – 62 percent of the winners also credit their success to “a healthy blend of art and science.” They value decision makers’ years of experience and understand the importance of well-tuned internal operational processes.

The Retail Systems Research report closes with a checklist of do’s and don’ts related to demand forecasting, customer analytics and localized assortments for retailers who want to be (or remain) winners.
“If retailers can follow these simple steps, they’ll go a long way towards optimizing their merchandising life cycle and creating a more compelling buying experience for customers,” the report states. “If they don’t, they risk being consigned to the dustbin of history.”

Drive Profitable Planning and Operations

• Translate detailed and summarized data into forward-looking actionable insights
• Automate data management and cleansing, and make manual updates by exception
• Apply the appropriate forecasting method to observed or predicted selling patterns
• Adjust forecasts to correct for out-of-stocks, seasonality, recent trends, and other causal factors
• Improve forecast accuracy for promoted, non-promoted, short- and long-lifecycle items
• Deploy via Oracle Cloud for Industries to simplify integration, improve agility, and ease the burden of implementation for your IT resources (optional deployment method)

Demand forecasting or demand planning?

Does your company do ‘demand forecasting’ or ‘demand planning’? I have seen numerous consultants address this semantic difference and argue for one over the other – most often for ‘planning’ over ‘forecasting’ as it gives a more active echo.
However both are needed in supply chain management and are indeed different things. Planning involves a collection of actions you are going to perform to obtain the results that you are seeking. In sales that might include, for example, pricing, campaigning and other marketing activity. Forecasting is always a numeric estimate of future outcome, and is based on historical performance, a selected plan, and the potential changes in other factors in the environment. A plan can be controlled, but the forecast always has an element of uncertainty; and the more you understand that uncertainty the better.

What’s the difference?

The difference between a plan and a forecast is best seen in stock management in a retail DC. For example for a future sales promotion you need to calculate a demand forecast for each product for each store – just to be able to order the right amount to fulfill the demand. Then you come to the delivery plan side – you can plan to ship e.g. 70% of the forecast volume 2 days before the promotion to give the stores enough time and goods to build up promotional displays. The pre-delivery is the biggest demand fluctuation of the promotion from the DC point of view.

Even though retail demand forecasting and replenishment are often seen as separate processes, they cannot be totally separated.

However it doesn’t even need to be forecast – it is basically a delivery plan, and you can get a precise picture of the volumes to be sent to each store and thus the required goods and resources to make those deliveries beforehand. This example also shows that even though retail demand forecasting and replenishment are often seen as separate processes, they cannot be totally separated.


Most future predictions hardly turn out to be true due to some unexpected circumstances or changes in the external environment. The same can be said for demand forecasting in the retail industry as well. But, retailers still perform demand forecasting as it is important for inventory management, production planning, and measuring future capacity requirements. Precise demand forecasting provides businesses with valuable information about their potential in the current market to make smart decisions on market potential, pricing, and business growth strategies. In this blog, Quantzig has listed the top demand forecasting trends in the retail industry.
According to the demand forecasting experts at Quantzig, “The retail industry simply can’t survive without demand forecasting as they risk making poor decisions about their inventory and products, which might result in lost opportunities.”
Speak to an expert to know more about the scope of our research.
Top demand forecasting trends in the retail industry

• Rising popularity of bottom-up forecasting: Today, the retail industry functions over many channels, which demands inventory positioning in several locations. As a result, retailers have to emphasize on bottom-up forecasting to meet the demand through various channels. Using such an approach helps them fulfill orders from both e-commerce and traditional retail channels for a wide array of varieties. It allows the retailers to meet customer demands more rapidly and distribute goods through the customers’ choice of channel. When the need arises, such methods can also help retailers balance inventory between distribution centers and stores through high-frequency inter-depot transfers.
• Focus on forecast quality: When carrying out demand forecasting, retailers usually look at demand signals. But, retailers with less sophisticated planning abilities often seek constancy in demand signals, which is frequently fragmented. As a result, they look for a combined model that lets all stakeholders cooperate via “what-if” simulations. Consequently, retailers have started looking for ways to quantify forecast quality by looking at external associations, including end users and suppliers to get better forecasts, which can then be shared with the sales team and suppliers.
• Fresh view towards long-tail items: A majority of the slow-moving or long-tailed items sell because they are in the inventory. Ensuring service levels is the key to mastering demand forecasting for slow-moving items. Such items cannot be formed consistently, so the retailers turn towards supply chain planning software to robotically model stock-to-service level, which precisely lists how much stock they need.
• Visit our page, to view a comprehensive list of top demand forecasting trends in the retail industry

Quantzig is a pure-play analytics advisory firm concentrated on leveraging analytics for prudent decision making and offering solutions to clients across several industrial sectors.
Request a free demo to see how Quantzig’s solutions can help you.
View the complete list of the top demand forecasting trends in the retail industry:
https://www.quantzig.com/blog/trends-demand-forecasting-retail-industry
About Quantzig
Quantzig is a global analytics and advisory firm with offices in the US, UK, Canada, China, and India. For more than 15 years, we have assisted our clients across the globe with end-to-end data modeling capabilities to leverage analytics for prudent decision making. Today, our firm consists of 120+ clients, including 45 Fortune 500 companies. For more information on all of Quantzig’s services and the solutions they have provided to Fortune 500 clients across all industries, please contact us.

Demand Forecasting

With customers expecting immediate responses irrespective of location or time, you need to meet these needs rapidly, and cost-effectively. Our experts can build predictive algorithms to better forecast demand based on internal/external factors and ensure that your Supply Chain is well-oiled to cater to these dynamic needs. We can create custom solutions? - reflecting your unique sales history while utilizing correct statistical methods and the most effective process to refine statistical results and capture collective knowledge.

Point of Sale: POS-based forecasting at retailer, distribution center, and store level Deviation Analysis: Forecasted and actual sales are compared at both the store and SKU level Dashboards and Customized Solutions: Key performance enterprise dashboards with trend analysis and predictive analytics for a multi-dimensional drilldown to increase visibility and insight on a global level. Customized solutions adjust to the needs of each individual organization, taking instead of retrofitting a pre-defined product or application Most retailers are facing a shrinking operating “margin for error”. So it’s not surprising that many are looking for more accurate demand forecasting and intelligent stock replenishment. In a report entitled Market Guide for Retail Forecasting and Replenishment Solutions, Gartner analyst Mike Griswold spotlights seven recent trends in this area. 1/ Multichannel retailing is requiring inventory positioning in more locations than ever before and causing retailers to focus on “bottom up” forecasting, says Griswold. Until now, he says, many retailers have planned less than half of their assortment at the item/location level, but now they’re looking for platforms that can scan disaggregated demand streams down to the channel and stock-keeping unit (SKU) level. These retailers want their supply chains to be able to fulfill both e-commerce and brick-and-mortar purchases for a wide item assortment of items in a way that is “inventory agnostic”. They look to match the dynamic evolution of demand and give customers multiple order collection and delivery options. If needed, they look to balance inventory between stores and DCs via high-frequency inter-depot transfers. 2/ Less mature retailers are also focused on the demand signal. Griswold reports that retailers with less mature planning capabilities are seeking more consistent ownership of the demand signal, which is often fragmented and often owned by merchandising (especially in apparel). They want a single, unified model that allows stakeholder collaboration via “what-if” simulations of trade-offs. More retailers are now measuring forecast quality, forecast value-add (FVA) and bias, Griswold adds. They’re also looking for more external collaboration, to get better forecasts and share them with sales channels and suppliers. 3/ More sophisticated retailers understand that servicing “lumpy,” long-tail demand is driven by inventory and not forecast accuracy. For the long tail—slow-moving items with unpredictable demand—the key to meeting demand is to ensure service levels. For these items that can’t be reliably planned, retailers want supply chain planning (SCP) software that can accurately and automatically model stock-to-service levels to offer a clear picture of demand variability: how much stock they need, the mix, and where and when they need it. 4/ Retailers of all maturities are looking to automate forecasting and replenishment to improve planner productivity. With increasing pressure on margins, retailers don’t want to add more planners, and automation is no longer a dirty word, he says. They want the automated application to do the “heavy lifting”, allowing their planners to add value and business acumen. 5/ Product returns are increasingly costly. More e-commerce means more returns and some retailers see that more diligence on the forecasting and replenishment side—as well as analytics—can help them better predict and minimize returns at the outset, and then better manage and reposition the returned goods across their inventory. 6/ Cloud-based applications are becoming more acceptable as deployment platforms. Smaller retailers have been quicker to adapt to the cloud than larger ones, Griswold says, but retailers in general are becoming more comfortable with cloud platform viability and scalability. They see that cloud offers the flexibility to scale to demand, help companies get up and running quickly without extensive IT resources, and hedge against IT infrastructure obsolescence. 7/ Retailers are beginning to understand how machine learning differs from statistical techniques. Griswold says that retailers are beginning to comprehend the benefits of machine learning in the retail supply chain. Machine learning can analyze demand variables and their complex interactions and patterns in automated fashion, and “self-learn” demand profiles. Machine learning’s ability to collate and analyze clusters of big data can also help predict demand beforehand for launches, promotions, and markdowns involving products that share similar characteristics.

Forecasts that Improve Over Time: An explicit process creates a benchmark against which future forecasts can be compared and improved Forecast Integrated with Marketing and Promotions: Procurement and production managers can respond to demand planning projections from sales and marketing activities