Firms adopt generative-AI CRM when they trust its functionality and task benefits, while information shortfalls and a preference for human contact slow uptake; neural-network analysis ranks perceived usefulness as the top predictor of trust and information barriers as the top predictor of inertia.

Main Finding

Organizations’ adoption of Generative AI–enabled CRM systems (GAICS) is jointly shaped by perceived consumption values (especially functional and instrumental values) that build initial trust, and by innovation-resistance barriers (need for human interaction, information gaps, and normative/ethical concerns) that create inertia. Initial trust and inertia both significantly affect adoption. In predictive models (ANN), functional value is the strongest predictor of trust, information barriers drive inertia, and the need for human-interaction barrier is the single most important predictor of GAICS adoption. Management support and flexible management practices help moderate resistance and enable adoption during digital transformation.

Key Points

• The study integrates Theory of Consumption Values (TCV) and Innovation Resistance Theory (IRT) to capture both positive drivers (values) and barriers to GAICS adoption.
• From qualitative interviews (37 participants), six first-order factors emerged:
  • Values: functional, convenience, instrumental.
  • Barriers: need for human interaction, information barriers (insufficient/ambiguous information), norm barriers (ethical/privacy/social acceptance).
• Hypothesized and empirically tested relationships:
  • Functional and instrumental values → positive effect on initial trust.
  • Need for human interaction, information, and norm barriers → associated with inertia/resistance.
  • Initial trust and inertia → significant antecedents of GAICS adoption.
• Analytical approach combined structural equation modeling (SEM) for hypothesis testing and artificial neural networks (ANN) for nonlinear predictive importance.
  • ANN importance ranking: functional values (top predictor of trust), information barriers (top predictor of inertia), need for human interaction barriers (top predictor of adoption).
• Emphasis on flexible management: adaptive managerial support, resource alignment, and executive advocacy moderate resistance and help align GAICS with strategic goals.

Data & Methods

• Mixed-methods sequential design:
  • Qualitative phase: grounded theory, 37 semi-structured in-depth interviews (employees, marketers, senior leadership; predominantly India). Interviews 35–45 minutes. Iterative coding produced zero-, first-, and second-order categories; coding reliability κ = 0.75.
  • Quantitative phase: conceptual framework derived from qualitative insights tested using SEM (to assess hypothesized causal paths) and ANN (to rank predictor importance and capture nonlinearities). Management support was tested as a moderator.
• The study focuses on SME and emerging-market contexts and situates findings in the flexible management / digital transformation literature.

Implications for AI Economics

• Modeling adoption: Economic diffusion models for generative-AI CRM should incorporate both value-driven demand (functional/instrumental benefits) and frictional resistance (inertia from information deficits, preference for human interaction, normative concerns). Trust and inertia are distinct state variables that jointly determine adoption probability.
• Investment priorities and ROI: Functional value (efficiency, real-time insights, automation) yields the largest trust payoff; investments that demonstrably improve functional performance will have the highest marginal impact on adoption. Conversely, investments in information transparency (explainability, documentation, training) reduce inertia and are high-leverage for uptake.
• Labor and task reallocation: The prominence of the “need for human interaction” barrier highlights complementarities and substitution effects. Where stakeholder interactions require emotional, cultural, or complex judgment, GAICS adoption will be constrained unless firms reallocate tasks (augment human roles rather than fully substitute). Economic models should account for heterogeneous task susceptibility to automation.
• SME and emerging-market dynamics: Resource constraints and sensitivity to normative/privacy concerns imply slower adoption without managerial advocacy and tailored support. Policies or incentives (subsidies, training, certification, standards for privacy/explainability) can correct market frictions and accelerate beneficial diffusion.
• Policy and externalities: Information barriers and norm/ethical barriers point to market failures from asymmetric information and social externalities (privacy reputation, societal trust). Regulators and industry bodies can reduce adoption frictions via transparency standards, data governance frameworks, and certification of GAICS outputs.
• Forecasting and valuation: Valuation and demand forecasts for GAICS vendors should incorporate nonlinearity: functional improvements produce outsized trust gains early, but growth can be capped by persistent human-interaction and information frictions unless explicitly addressed.
• Research priorities for AI economists: quantify macro and firm-level productivity gains from GAICS, estimate adoption elasticities with respect to functional performance improvements and information provision, model heterogeneous adoption across tasks and sectors, and measure labor reallocation and wage effects tied to human-interaction-intensive activities.

If you want, I can convert these implications into a simple formal adoption model (specifying trust and inertia as state variables), or produce short policy recommendations tailored to SMEs or to GAICS vendors.