Yahoo Lifestyle AI’s Role in Overcoming Post-pandemic Supply Chain Challenges
Roger Mayerson, senior vice president and industry principal of apparel and soft goods at AI supply chain planning software company Logility, sees this post-pandemic period as challenging for retailers and brands from an inventory management and forecasting perspective. Mayerson said a shift in buying behavior, new channels of distribution and changes in online shopping preferences make the old forecasting model obsolete.
Here, Mayerson discusses these challenges and how AI-powered technology can make forecasting easier and more accurate while transforming the supply chain.
More from WWD
Product Design Firm Whipsaw Lends Its Award-winning Prowess to the Wellness Space
Alo Brings Its Studio to Street Style to an Exclusive Partnership Launch with Beats
WWD: From an inventory management and forecasting perspective, what are the challenges facing apparel retailers and brands?
Roger Mayerson: Over the past few years, we’ve seen a few primary challenges facing apparel retailers and brands.
The biggest challenge by far has been the unreliability of historical sales to create a good representation of future sales. The traditional way of forecasting, by relying on the past two to five years’ worth of historical sales data to build a statistical forecast, doesn’t work in a market that is facing constant disruptions. Brands are finding that the sales data from 2020 to 2022 is producing low-quality forecasts and they need to find a better way to build their demand plans.
There has also been an increase in the number of products being managed in the portfolio and the number of channels being used. This proliferation of complexity is stressing traditional inventory planning processes and tools. It’s not just the continuing shift to e-commerce, but also within e-commerce, there has been a change in buying behaviors. For example, for many consumers delivery date has become more important than price.
We’ve also seen changes in the way products are promoted and the associated marketing spend. It’s not just about discounts and ads anymore. Tailoring avatars or customer images with similar build, height or hair color can influence demand as much as a commercial or print ad once did.
WWD: How are new technologies such as AI, predictive AI and generative AI helping to transform traditional forecasting and inventory management?
R.M.: With AI first forecasting, the forecaster or demand planner needs to shift how they work. Understanding business trends, and customer shifts and explaining outlier data to the AI tools is more important now than being a statistical genius or data scientist.
Baseline results are more accurate and planner input or overrides can easily be measured to determine when to aid AI and when to step away.
Artificial intelligence is having a profound effect on the supply chain planning space right now. Leveraging the full spectrum of AI in an integrated platform is allowing brand owners to focus on forecast value added activities, take advantage of demand sensing and demand shaping opportunities, while simultaneously reducing their inventory levels without sacrificing customer satisfaction. It’s truly an exciting time to be in the supply chain space.
AI is enabling a new way to collaborate on the forecast. The old way of negotiating the “one-number” is out. Our clients have found they achieve much better results and more accurate forecasts with driver-based forecasting and collaboration around the components of demand. For example, collaboration now happens around the forecast inputs such as planned promotions and new product introductions. The AI is used to surface trends and identify outlier events that the planner can then use to train the AI. It is a fundamentally different approach, with a significant improvement in business outcomes.
One of the clearest use cases for generative AI is in-demand planning. There is so much data flowing in from POS systems to consumer sentiment that it’s hard to manage it and keep track of all the dashboards and reports. GenAI has democratized access to this critical supply chain data in a revolutionary fashion. It has opened the forecast and demand drivers to groups outside of demand planners — such as sales, marketing and finance. It has proven to be a great way to bridge departmental silos and remove decision latency.
WWD: How does your company’s platform work?
R.M.: Logility has the broadest end-to-end supply chain planning platform in the market. It covers many diverse supply chain planning needs like vendor management and compliance, product life-cycle management and traceability, manufacturing and supply optimization, demand and inventory optimization, and even network design optimization and business planning such as integrated business planning (IBP) or sales and operations planning (S&OP).
In addition to a fundamentally different approach to demand forecasting and industry-leading multiechelon inventory optimization capabilities, the breadth of our solutions supports clients facing the increasing complexity of global supply chains and the growing importance of data analytics and automation — especially the convergence of sourcing and planning. An integrated platform eliminates organizational silos which enhances operational performance. Many of our clients are using the Logility Digital Supply Chain Platform to share data and collaborate more closely between sourcing and planning teams, ensuring sourcing and procurement decisions align with production and inventory needs.
WWD: How would you describe the value proposition? And what are some of the ROIs retailers and brands can experience?
R.M.: Some practical accomplishments made possible with convergence by our clients include a major global apparel brand having 3,500 tier one and tier two suppliers cataloged for necessary certificates of compliance. They mapped 350 at-risk vendors and their suppliers through tiers of production in China for compliance. Having an alternate source for every material with a reasonable amount of effort and onboarding in a short time frame is one of the greatest examples of the power of convergence.
As consumer expectations of sustainable products grow, the convergence of sourcing and planning includes the ability to release information about the documentation of the physical products’ journey from origin to consumer. Using recent vendor performance to auto-adjust lead times and inventory policies is another great example of convergence paying off in operational performance.
The key to accomplishing these great feats lies in Logility’s business analytics layer, sourcing management capabilities, vendor management and supply planning optimizer.
WWD: How can technologies such as your platform help with improving sustainable practices?
R.M.: No one will argue that supply chains shouldn’t be more sustainable. But the challenge is how do you actually get there, how do you manage it? There is a misconceived perception that sustainability is cost-prohibitive for a lot of organizations and that it is too expensive to operate a sustainable supply chain. We think that couldn’t be further from the truth and the way we help brands achieve their sustainability goals without sacrificing profit is with the right applications of artificial intelligence and machine learning.
Logility has leveraged these technologies to help companies maximize their expansive data sets, getting the best information to guide their current sustainability strategies. What’s more, AI enables brands to employ a digital supply chain twin, a virtual representation of their supply chain designed to reflect a physical reality that assists in analyzing different scenarios.
With the right tools and technology, modeling and continuous improvement are possible. Additionally, performing “what if” scenario modeling for alternative options and balance between objectives can foster the best sustainability outcomes. Long-term sustainability projects with higher upfront costs and longer payback periods demand rigorous data analysis to evaluate potential pain points, trade-offs and other potential pitfalls.
For example, constructing a new distribution center or reworking a network can produce higher short-term costs for long-term savings. Shuttering an old plant to build a new one requires a large investment but pays off with increased productivity, efficiency and lower costs in the long run. Similarly, while nearshoring requires upfront work and can carry higher production costs, it offers significant savings on fuel and transportation.
Modeling and remodeling these scenarios empowers decision-makers to leverage artificial models to inform real-world decisions.
Best of WWD
