With the help of ChatGPT and Bard, the hype around generative AI has soared over the past few months, with experts shouting daily from the mountaintops about its potential to revolutionize the retail industry.
From generating product descriptions and dynamic pricing to improving data quality and automating redundant tasks, AI has been monikered as the magical fix-it-all technology that promises to slash time-to-market, quadruple employee productivity and deliver larger-than-life customer experiences.
And while these claims may technically be possible, there’s an implicit caveat that AI bandwagoners can often gloss over: both your product information and technology stack need to be optimized and up-to-date before you can start to take full advantage of generative AI.
AI is only as effective as the data it receives. When it comes to training language models and leveraging AI algorithms, scattered and incomplete information poses significant challenges. One of the most common examples of AI in the retail industry today is in the use of chatbots to provide customer support. In order to answer questions about the available colors included in the latest product release, the bot needs access to product design files stored in the product lifecycle management (PLM) system and product descriptions or material lists from the product information management (PIM) system.
Advertisement
Similarly, when customers inquire about the status of their orders, the chatbot relies on the order management system (OMS) for inventory and shipping information, as well as a customer management system or CRM for address verification.
Or what if a customer comes to your chatbot looking for personalized product recommendations? Your bot needs to know previous searching, browsing and purchasing history of both that customer and similar customers, along with availability information, shipping estimates and pricing information of the products it’s recommending, all of which is stored in a number of different software platforms, solutions and tools.
And all that’s just for one use case of AI.
In order to leverage AI effectively, organizations need seamless integration of these various technologies involved in managing product data. And that leads us to the caveat of the caveat: proper product information requires a flexible, modular system where different software services can be selected, assembled and adjusted as needed — AKA a composable architecture.
Composable commerce offers two key advantages: scalability and flexibility. By breaking down the silos of individual software platforms and fostering synchronicity, a composable architecture enables seamless communication across the organization.
AI algorithms thrive on data. As AI applications often require access to multiple systems and data sources, a composable commerce framework enables retailers to create a centralized repository for all product data by seamlessly connecting their PIM, ERP, OMS, DAM and other systems, making product data easily accessible to these AI applications.
This unified data environment enables retailers to create a foundation of rich product information, including descriptions, attributes, images and pricing, to enable AI algorithms to utilize accurate, complete and up-to-date information to generate dynamic pricing, personalized recommendations, targeted marketing campaigns and other intelligent functionalities.
With a composable framework, organizations can efficiently manage and activate their entire product record across all channels without being hindered by the complexity of their tech stack, the quantity of custom integrations they’ve created or the number of stakeholders involved.
The Future of AI Technology
AI technology evolves rapidly, and retailers need the flexibility to adopt new AI capabilities as they become available. The modular nature of a composable architecture allows retailers to plug in new AI solutions without disrupting their entire system. This flexibility ensures that retailers can leverage the latest AI advancements to enhance customer experiences, optimize operations and drive business growth.
AI technology often requires experimentation and iterative improvements. A composable commerce framework empowers retailers to experiment with different AI algorithms and models without extensive system reconfigurations. By decoupling the components of their commerce infrastructure, retailers can test and deploy AI solutions rapidly, measure their impact and make necessary adjustments. This agility enables retailers to fine-tune AI algorithms, improve accuracy and uncover new insights to refine their customer experiences continuously.
AI technology has the potential to deliver highly personalized and engaging customer experiences across various touch points, but an outdated monolithic architecture will hinder the extent to which you can utilize the technology. A composable commerce framework provides the scalability and flexibility needed by retailers to leverage AI to offer dynamic pricing, personalized recommendations, chatbots for customer support, visual search capabilities, voice assistants and more.
So before you too jump on the AI bandwagon, take a step back to ensure that your technology infrastructure and product information management is in place, so that when you do begin integrating AI into your customer journey, you’re able to provide convenient, tailored experiences that drive customer satisfaction, loyalty and ultimately, revenue growth.
Kristin Naragon is Chief Strategy and Marketing Officer at Akeneo, the Product Experience company. Before joining Akeneo, Naragon was the global go-to-market strategy leader for Adobe’s marketing automation offering. She brings many years of experience spearheading alliances, sales, strategy, product marketing and go-to-market capacities for B2B tech companies, from high-growth startups to category-defining major corporations. Naragon earned her MBA from Harvard Business School and an undergraduate degree from Pennsylvania State University. Outside of work, she enjoys fitness and contributing back to the community with her husband and two kids.
