From Deplorability to ROI Evaluability: Expert Perceptions of AI-Enabled Capabilities in Digital Marketing, Armenia

Abstract

Artificial intelligence (AI) is widely used in digital marketing and e-commerce, but its strategic value is often harder to prove than its operational usefulness. This tension is especially visible in small and emerging markets, where tool adoption may outpace measurement and governance capacity (Gama & Magistretti, 2023; Islam et al., 2024). The study adopts a capability-based and governance-oriented perspective on AI value in digital marketing management (Abrardi et al., 2021; Mariani et al., 2021). It examines how Armenian marketing experts evaluate four AI domains—predictive analytics, sentiment analysis, personalization, and programmatic advertising—and tests how these evaluations relate to perceived ROI-oriented impact, adoption outlook, and governance perceptions (Mariani et al., 2021; Haleem et al., 2022; Dumitriu & Popescu, 2020). A cross-sectional expert survey (N = 29) was used because objective performance data are limited and fragmented in this small-market context. The survey collected 5-point Likert ratings and one open-ended response on adoption barriers. Differences in perceived tool effectiveness were assessed using the Friedman test with Durbin–Conover post hoc comparisons. Hypothesized relationships were examined using Spearman correlations, supplemented by exploratory regression models. Personalization and programmatic advertising were rated highest in perceived day-to-day effectiveness, but only predictive analytics showed a clear positive association with perceived ROI-oriented impact. Perceived ROI-oriented impact was positively associated with adoption outlook. Governance perceptions were asymmetric: privacy concerns were negatively associated with adoption outlook, whereas bias/fairness concerns were positively associated with oversight/accountability expectations. Open-ended responses identified weak measurement discipline, capability gaps, and poor strategic integration as the main barriers to realizing AI value. These findings imply that AI value realization in digital marketing depends not only on tool deplorability but also on measurement discipline, integration routines, and governance capacity. The study is limited by its small, non-probability, expert sample, perception-based measures, and cross-sectional design, so the findings should be interpreted as exploratory and context-bound. The study contributes to theory by distinguishing deplorability from ROI evaluability, and to practice by showing that AI value realization in small and emerging markets depends on measurement discipline, integration, and governance (Avram et al., 2020; Pop, 2020; Akter et al., 2023; Schmauder et al., 2023).