AI: Fluent Commerce

Nicola Kinsella
Senior Vice President, Global Marketing
Fluent Commerce

Hardware Retailing (HR): What are some of the practical applications AI has for the independent channel?
Nicola Kinsella (NK): There are a lot of great tools emerging for product information enhancement using GenAI, including product descriptions, merchandising imagery and even automated generation of product attributes. Marketers are of course using GenAI to help generate copy for all sorts of customer communications.

And in the area of predictive machine learning (ML), retailers have seen great results in improving demand forecast accuracy, pricing recommendations and commerce search.

HR: How are these AI tools helping retailers be more efficient, save money and boost overall operations?
NK: When it comes to GenAI, it’s all about making humans more effective and productive. If you can reduce the time it takes to update product information or create campaigns, there are real labor savings. And if you can optimize your inventory and pricing, and maximize online conversions, it can have both significant financial gains, including reducing inventory carrying costs and improving margins, as well as improve the customer experience by reducing canceled orders.

HR: Data is king when it comes to AI. Can you speak to the importance of good data? What are some of the ways retailers can gather and maintain good data?
NK: Absolutely! Data is fundamental and powers informed decision-making. The more good, clean data you have, the better the long-term outcomes. When it comes to demand forecasting and optimizing search results, the more product attributes you track, the better. You also want to have a granular view of your inventory levels over time. And if you have an event-driven platform for managing inventory availability that can track the ratio of stock availability checks to orders, which can vary a lot by item, it can also boost your digital demand forecast accuracy.

Another area where an event-driven platform is key to gathering good data is order management. If you have a system that tracks how long an order takes in each status, and tracks the status of fulfillment locations, including inventory availability, fulfillment capacity, etc., it will give you a lot more data that can be used to optimize a sourcing and allocation ML model in the future.

HR: Along with data, what are some of the other aspects retailers need to consider when using AI?
NK: Large Language Models (LLMs) are going to fundamentally change the way business users interact with enterprise software. Today we’re only scratching the surface. The best is yet to come, which means three things. First, it’s important to understand the different types of AI and their use cases: machine learning is best for predictive use cases like demand forecasting or sourcing optimization and GenAI is for creating new content.

Secondly, as you evaluate new solutions, look for vendors taking an AI-first approach to their product roadmaps. The evolution of AI is moving at a rapid pace. New models and tools are being released daily, so it’s important that when selecting new tech it’s flexible and can track all the data needed to power AI use cases, even if you’re not ready to take advantage of them yet. And be sure that the vendor has a strong, innovative and AI-driven vision for the future.

For example, in the area of distributed order management, we’re developing a Large Language Model (LLM) based UI that will allow business users to create new dashboards and reports by asking a question. It can update their order workflows and business logic, and it will configure notifications and alerts—all using natural language queries. It’s an exciting time.

Finally, it’s important to establish good data governance policies. The results you achieve from AI will only be as good as the data that sits behind it, so make sure you invest in the right team to manage your data and address any ethical concerns up front. That way you’ll reap all the rewards of AI and avoid risking your brand.