Is There a Journey Beyond the Gen AI Prototype? Why Do Retail Companies Fail at Adopting Gen AI


Generative AI is at the top of retailers’ to-do lists in 2024. Companies are diving into GenAI solutions, with BCG forecasting that by 2025, GenAI technology will capture 30% of the AI market.

Retail businesses are embracing GenAI with open arms. A striking 87% of shoppers who’ve tried a GenAI tool are thrilled with the positive impact on their shopping experience, while 88% of companies are actively exploring GenAI to enhance customer interactions. Moreover, 73% of consumers are now open to AI-powered chatbots for customer service. We’re in the experimentation phase—companies are testing use cases, building prototypes, and navigating the complex technological, legal, and compliance challenges landscape. Success stories from industry giants like H&M and Wayfair showcase the transformative potential of GenAI at the enterprise level.

However, the 2024 outlook for GenAI adoption is mixed. Many GenAI projects stall in the experimental phase and fail to make it to production. In this article, we delve into the reasons behind these setbacks and propose strategies to overcome them, paving the way for successful GenAI implementations.

GenAI Adoption: Reality Beyond the Hype

While 2023 saw a surge in retail adoption of generative AI, the outlook for 2024 appears less optimistic. According to a report by Everest at the beginning of 2024, several key factors are preventing GenAI prototypes from scaling to the production enterprise stage:

  • Lack of clarity on success metrics
  • Budget and cost concerns
  • The fast-evolving technology landscape
  • Security and privacy concerns
  • Talent shortage

At DataArt, we’ve observed increased interest in GenAI technology. However, few successfully delivered prototypes have been scaled to enterprise-ready solutions.

Retail businesses we’ve consulted with typically face challenges that align with findings from independent industry reports:

  1. Lack of AI Company Strategy. Implementing GenAI within a company demands a cohesive strategy and vision. Success hinges upon integrating GenAI use cases into a broader AI platform approach aligned with the company’s overarching AI strategy. This strategy should address key business challenges, adopt compatible technological stacks, and adhere to established AI and data governance frameworks. Attempting to build GenAI solutions without this strategic compass is akin to navigating a ship without a compass, risking eventual wreckage or aimless drifting.
  2. Navigating an Uncertain Regulatory Environment. Navigating the regulatory landscape adds another layer of complexity. Beyond data privacy and compliance, issues such as copyright and intellectual property rights present challenges. With laws in flux, adoption becomes inherently risky. For instance, in e-commerce, generating product text descriptions may clash with SEO compliance, while image generation could raise intellectual property concerns.
  3. Understanding Infrastructure & Technology. Misconceptions about infrastructure and technology abound. While some believe GenAI can work off-the-shelf, it thrives on custom data. Training on the organization’s data significantly improves accuracy and ensures results meet production standards, reducing the risk of failure. Company data is crucial for achieving tailored accuracy; with it, LLMs can perform well.
  4. Lack of Change Management Processes and Culture. Like any AI project, this requires organizational buy-in and a phased approach to adoption to prevent missteps. For instance, the infamous Meal bot incident underscores the importance of meticulous planning and verification before deployment.
  5. Uncertain ROI on AI Solutions and Hard-to-Assess Measurable Impact. Additionally, there’s a need to recalibrate ROI expectations. While GenAI minimizes human effort, it isn’t a wholesale replacement. Accurately estimating operational and maintenance costs is essential for a realistic ROI assessment.
  6. Assessing Company Talent. Lastly, talent assessment is crucial. While outsourcing can aid implementation, internal Data, and AI leadership are indispensable for aligning efforts with the company’s strategic vision. Collaborating with the right partner can expedite progress, but internal expertise ensures alignment and sustainability.

GenAI holds immense promise, but realizing its potential requires a strategic, holistic, and adaptive approach tailored to each company’s unique circumstances.

Path to Scaled GenAI: Ways to Address the Challenges

Despite the numerous challenges, the industry boasts notable success stories of GenAI adoption and its scaling to an enterprise level. For instance, this trend exemplifies H&M’s virtual styling studio, Wayfair’s home styling visualizer, and Under Armour’s team talk.

Whether pursued solely in-house or through collaborative efforts with partners, retail businesses must prioritize cultivating data-driven cultures and investing in digital and data maturity. Success necessitates a holistic approach integrating technological advancements, organizational readiness, and ethical considerations.

To embark on this transformative journey, several foundational steps warrant consideration:

  • Formulating the North Star Vision: Develop a clear vision for data maturity and AI assessment to guide the organization’s strategic direction.
  • Appointing an In-House Leader of Data and AI: Designate an internal leader of Data and AI to spearhead strategy and coordinate efforts across internal teams and external partners.
  • Exploring Business Challenges and Prioritizing Use Cases: Delve into specific business challenges and prioritize use cases for future solutions based on their potential impact and alignment with organizational goals.
  • Developing the Right Technical and Organizational Strategy: Create a technical and organizational strategy tailored to effectively address identified business challenges.
  • Formulating the Project’s Vision and Prioritization: Establish a clear vision for the project and prioritize initiatives with a well-defined roadmap.
  • Establishing Success Metrics and Monitoring Procedures: Define success metrics and implement monitoring procedures to track progress and ensure alignment with objectives.

These steps align with the nine building blocks of AI strategy, encompassing AI strategy formulation, establishment of AI committees, legal and compliance approaches, risk management frameworks, data management, information security, platform construction, monitoring and observability approaches, and skills and expertise development. These are common elements observed in successful GenAI adoption in retail.

In conclusion, GenAI’s effectiveness in retail depends on meticulous planning and digital proficiency. Retailers who plan diligently and grasp GenAI’s intricacies stand to gain the most. The aim is to implement GenAI thoughtfully, guaranteeing substantial benefits that drive business expansion. Let’s prioritize these savvy decisions to capitalize on GenAI’s full potential.