AI-Powered Sales: Bridging Opportunity and Value Gap
By Japjit Ghai, Abhinav Duggal, and Archith Mohan
Over the past year, AI has been central to most go-to-market discussions. The increase in available first- and third-party data has created exciting possibilities for AI in sales. Additionally, recent macro shifts, such as slowing demand, rising customer acquisition costs, and decreasing sales productivity, have made it essential to use AI to improve efficiency, create new revenue streams, and boost shareholder value.
The economic potential of AI-powered sales is enormous. It promises better results and sets a new standard for customer experience. Predictive and hyper-personalized insights powered by AI can significantly boost customer lifetime value (LTV). Based on our experience, there are five predominant use cases where AI can help deliver transformative P&L impact (Exhibit A).
Exhibit A: AI delivers transformative P&L impact along 5 vectors
The applications of AI in sales have rapidly evolved over the past decade. In recent years, we’ve seen a shift from relying on intuition and tribal knowledge to intelligent selling, leveraging third-party signals and machine learning for predictive insights and hyper-personalized strategies. More recently, the emergence of Generative AI (GenAI) has introduced “assisted” selling capabilities, providing real-time, personalized steer and coaching to enhance seller performance. We are also witnessing the rise of “autonomous” digital selling through chat, voice, and video avatars — not as a replacement but as an extension to the human sellers. This evolution reduces sales drag, allowing sellers to focus on building relationships and closing deals — embodying the expectation of “letting sellers sell” (see Exhibit B).
Exhibit B: Value realization from AI is lagging the pace of tech evolution
Despite a decade of rapid innovation, value delivery has been lagging. Our industry-leading experience shows that most companies haven’t realized the promised commercial benefits of intelligent selling, even with significant investments in data platforms, analytics, and enablement. Many experiments start with great design and pilots but struggle to graduate to AI-driven motions at scale. Additionally, the proliferation of a large variety of GenAI-based sales tools with information and attention scattered across multiple tabs is causing tool fatigue among sales teams. Four key reasons for AI-powered sales initiatives not delivering value:
1. Unreliable data foundation: Companies often rush into designing propensity algorithms without first establishing a reliable single source of truth on data. Two major issues frequently arise. First, CRM data often suffers from undefined account hierarchies, incomplete data fields, vertical tagging, and duplicates, making it ineffective for large-scale analysis. Second, fuzzy matching algorithms with partial accuracy are used to integrate costly third-party data (such as firmographics, intent, purchase signals, and contact information) with first-party CRM and ERP datasets. This results in “garbage in, garbage out” scenarios, where the data might appear insightful but fails to deliver real business value.
2. Lack of Experimentation and Feedback Loops: The design of AI models and initiatives are often siloed within data intelligence teams, who build proof-of-concept models in a sandbox. Companies frequently overlook the importance of establishing cross-functional feedback and experimentation loops with the field teams. This results in “black box” models that, while capable of generating algorithmic recommendations, lack business explainability & trust. Moreover, without a robust experimentation engine to test and learn, these recommendations quickly lose relevance in ever-evolving market dynamics.
3. Inadequate Tooling: At-scale AI-driven sales motions require synchronous, omnichannel, and seamless activation to drive seller adoption. This demands skillful front-end customization and comprehensive back-end integration to create a single unified interface and experience where sellers can easily access all customer data, insights, and recommendations to act on them. However, companies often rely on off-the-shelf tools with loosely integrated datasets and insufficient customization to align with their sales processes. This combination results in tools consuming investment dollars and contributing to seller fatigue, instead of streamlining sales workflows.
4. Failure to influence culture change: Every artificial intelligence journey is different: each one is shaped around an organization’s starting point, circumstances, and goals. But the common recipe to success in our experience is what we call the 10/20/70 rule: dedicating 10% of the effort to algorithms, 20% to data and technological infrastructure, and 70% to business and people transformation. While algorithms and infrastructure can help determine who to target, we often see a lack of focus on designing the right operating model, investing in top-performing capacity to lead pilots, ensuring specialization, building centers of excellence, equipping the salesforce with the right incentives, and driving enablement and behavior change.
The potential for AI to revolutionize sales is immense, and the opportunity to realize this potential is more tangible than ever. By implementing a finely tuned AI stack, investing in data quality and governance, and redesigning operating models to manage change effectively, businesses can unlock substantial value. Whether you’re just starting your AI-powered sales journey, are already deep into it, or have innovative ideas to share, reach out to us. You can read more BCG perspectives on the transformative impact of the latest developments in AI in our new series, AI Transformation for Future-Ready Functions.
The authors would like to thank Alfonso Abella, Ignacio Hafner, Tarandeep Ghai, Patrick Mueller, Vaibhav Malhotra, Eshaan Puri, Aayushi Jangid for their contributions to this article.
