October 17,2023 | Navigating the AI Revolution in CPQ and B2B Commerce: Separating Hype from Practicality

In recent years, the impact of Artificial Intelligence (AI) on our daily lives and commercial applications has been nothing short of revolutionary. This technological transformation has been fueled by the emergence of Large Language Models (LLMs), which have significantly lowered the cost of AI deployment for applications that may not boast millions of users. This paradigm shift has led to a race among enterprise app vendors to integrate AI into their offerings, including Configure, Price, Quote (CPQ), and B2B Commerce solutions.

In this blog post, we'll explore the potential and pitfalls of AI in enterprise applications, with a focus on CPQ and B2B Commerce, including what Logik.io’s perspective is on the initial and future impact that AI can and will have on these technologies: 

The Pitfall of Overhyping AI in Transactional Enterprise Apps

While it's tempting to jump on the AI bandwagon and add chatbots to every application as the primary means of introducing AI, it's essential to be discerning. For instance, many CPQ vendors highlight the idea of recommending additional products, similar to how Amazon does. However, it's crucial to question why such innovations have not been available in the CPQ space for the past 15 years, especially when B2C companies like Amazon and Netflix have successfully implemented similar recommendations.

The primary reason behind this discrepancy is data sparsity in most B2B enterprise use cases. Simply amassing data in data lakes won't substantially improve recommendations, even with generative AI. To bring AI to your admin or end users successfully, it's essential to objectively assess your data's state and collaborate to identify viable use cases.

A Strategic Approach to AI Implementation

So, how should businesses approach the integration of AI into transactional applications like CPQ and B2B Commerce? 

Here are some key steps that we at Logik.io are taking to heart with our AI roadmap:

• Prioritize Impactful Dimensions: Focus on areas where AI can have the most significant impact while minimizing the risk of inaccurate responses. For instance, consider automating the connection of actionable information to reduce cognitive overload for users.

• Evaluate Use Cases Carefully: Determine whether a use case calls for generative AI, supervised learning, reinforcement learning, or a combination thereof. Conduct proof of concepts to manage expectations and mitigate potential failures.

• Gradual Feature Launch: Introduce AI features incrementally to allow for iterative learning and to avoid disrupting customer business processes, which are often mission-critical.
• Co-innovate with Customers: Collaborate with customers to build trust and determine what data can be shared with AI models. Start with the least sensitive data for training to comply with evolving regulations and protect intellectual property.

• Consider Cost Implications: Discuss the additional costs or price impact of AI adoption in your CPQ or Commerce applications. Be mindful of the economics of LLM models and avoid unexpected expenses.

Beyond LLMs: Embracing Different AI Approaches

While LLMs like ChatGPT have garnered significant attention, it's crucial to understand that AI's value extends beyond generative AI. Supervised learning, as explained by AI pioneer Andrew Ng, plays a substantial role in AI applications. LLMs can complement supervised learning by helping users express their intent more naturally in a conversation, while supervised models execute targeted tasks effectively.

Attribute-Based Configuration for Enhanced AI

To harness AI's intelligence in transaction workflows, an attribute-based configuration approach is essential. This approach provides a clear signal of user intent, which is critical for AI inference engines to recommend the right actions. Unlike part-based configurators, attribute-based configurators allow for more contextual recommendations, making AI more effective in providing valuable insights.

Conclusion

The rapid advancement of AI, especially in transactional enterprise applications like CPQ and B2B Commerce, presents both opportunities and challenges. To make the most of AI's transformative potential, businesses should adopt a thoughtful and strategic approach, focusing on real impact, data collaboration, cost considerations, and a diverse range of AI techniques. By doing so, they can navigate the AI revolution successfully and deliver valuable experiences to their users.