Why forecasting new products is like driving without a map — and how AI can help

The New Product Paradox

Picture this:  you’re driving in an unfamiliar city and your phone dies, leaving you without GPS and navigation. You’re on your own, without any history of the streets or landmarks to lead you efficiently to your destination.  Many businesses, especially those in the retail and manufacturing sectors, face similar challenges when launching new products without any historical data to guide them. The difference is a wrong turn can cost their companies millions of dollars.

Every new product launch carries a significant amount of uncertainty and risk. Without historical sales data, business planners lack insights into performance trends and patterns to adequately predict future demand for a new product. In fact, The Institute of Business Forecasting & Planning (IBF) reports an average forecast error rate of 64% for new products.  Consider a clothing company that wants to branch out into a casual shoe line. Having never offered footwear before, they can’t look to past performance to predict demand, nor can they easily correlate with other products in their portfolio.

The Institute of Business Forecasting & Planning (IBF) reports an average forecast error rate of 64% for new products.¹

Market uncertainty adds another layer of complexity. Customer preferences shift rapidly, competitors make unexpected moves, and external factors like economic conditions or regulatory changes can dramatically impact success.  For our clothing company above, social influencers can make or break a new product, with a single viral post potentially driving unexpected demand spikes or sudden drop-offs. These forecasting challenges apply across a wide range of industries as companies look to bring innovative products and solutions to market.  As another example, an automotive manufacturer launching a new electric vehicle may rely on market research suggesting strong demand for sustainable transportation, however factors like charging infrastructure, government incentives, and supply chain constraints can reshape the landscape overnight.

Adding to the complexity of forecasting for new products is the potential for cannibalizing existing products. Building on our previous example, traditional car manufacturers entering the EV market not only have to predict demand for their new electric vehicle offerings, but also how they will impact sales of their conventional vehicle counterparts. This internal competition makes revenue forecasting complex, as higher-than-expected sales of the new EV vehicles might not translate into higher overall revenue for the company if they're simply replacing sales of higher-priced conventional vehicles.

Applying AI to an Old Problem

The advent of advanced AI modeling techniques is significantly shaping the way businesses forecast and plan for new products, driving greater accuracy, explainability, and adaptability to dynamic market conditions. At Ikigai, we lead this transformation with cutting-edge AI-driven solutions designed specifically to address the challenges of new product launches.

Built on the patented Large Graphical Model (LGM), Ikigai’s AI technology transforms enterprise time series data into highly accurate and actionable forecasts and plans. While valuable across a multitude of industry use cases, it is unparalleled in its effectiveness in predicting and planning for new products across sectors such as retail, consumer goods, and manufacturing, where data is scarce and uncertainty is the norm.

Addressing the Challenges of New Product Launches

Lack of Historical Data

Launching a new product without historical data relies heavily on guesswork when utilizing traditional rules-based forecasting approaches. Conversely, Time2Vec leverages sophisticated AI-based statistical modeling to reveal hidden relationships and patterns across similar products to infer likely demand patterns, driving forecast accuracy even with limited data.

Complex Business Hierarchies

Planning for a global product launch demands consistent, reconciled forecasts at every level of the organization to ensure sufficient inventory and efficient logistics. Ikigai’s hierarchical reconciliation aligns forecasts from the lowest level SKU up to high-level aggregations, such as regions or categories, ensuring consistency across all levels. This supports collaboration across the organization and facilitates rapid response to changing market dynamics which is critical when venturing into a new product or market.

Real-time Influences

In many, if not most, industries, forecasts are heavily influenced by external data, such as weather, customer sentiment, and supply chain disruptions. The ability to incorporate real-time data empowers businesses to react quickly to external influences, dynamically adjusting forecasts to gain competitive advantage and reduce risk.

Shifting Business Priorities

What-if scenario planning ensures that businesses can react quickly to unexpected outcomes or conditions. Using aiCast^TM, planners can model a variety of scenarios to understand the impact of different business drivers. For example, a company in the CPG sector may want to model how allocating a greater percentage of marketing spend to launch a new energy drink could impact sales of their existing soda products.

Domain Expertise

In new product launches where limited to no historical data exists, domain experts can significantly impact the accuracy of forecasts with real-time insights and intuition. Expert-in-the-loop is a hybrid approach employed in the Ikigai platform where human expertise is integrated into the machine learning process.

Conclusion

Launching a new product can feel like navigating an unfamiliar city without a map—uncertainty reigns, and wrong turns can prove costly. Advanced AI technologies, like those pioneered by Ikigai, are transforming how businesses approach forecasting and planning with sophisticated techniques that uncover hidden patterns, integrate human expertise, and generate scenario plans. No GPS needed.

Explore our interactive demo and learn more at ikigailabs.io.

‍

‍

¹ https://demand-planning.com/2017/12/18/new-product-forecasting-planning-benchmark-report-lifecycles-shorten-forecasting-becomes-harder/

‍

‍

‍