Artificial Intelligence Based Predictive Analytics for SKU Performance and Revenue Optimization in Competitive Markets

Abstract

This study addresses the problem that many cloud-enabled enterprises invest in AI predictive analytics but still experience inconsistent SKU portfolio performance and avoidable revenue leakage because analytics capability, data integration, governance, and user adoption are uneven across functions. The purpose was to quantify how strongly AI Predictive Analytics Capability (AIPAC) influences SKU Performance (SKUPerf) and Revenue Optimization (RevOpt) in enterprise settings. Using a quantitative, cross-sectional, case-based design, data were collected via a structured 5-point Likert questionnaire from N = 210 professionals drawn from cloud and enterprise operational cases (forecasting, pricing and promotion, inventory and replenishment, and analytics roles). Key variables were AIPAC (overall construct and five capability dimensions: forecasting support, pricing and promotion decision support, inventory and replenishment decision support, data integration quality, and governance plus user adoption), SKUPerf, and RevOpt. The analysis plan included internal consistency reliability (Cronbach’s alpha), descriptive statistics, Pearson correlation, and OLS regression models predicting (1) SKUPerf from AIPAC, and (2) RevOpt from AIPAC and SKUPerf, plus a dimension-level regression to identify the most influential capability components. Reliability met accepted thresholds with AIPAC α = .91, SKUPerf α = .88, and RevOpt α = .90. Descriptively, perceived capability was high (AIPAC M = 4.02, SD = 0.61) while outcomes were moderate to high (SKUPerf M = 3.92, SD = 0.62; RevOpt M = 3.87, SD = 0.65). Correlation results showed strong positive relationships among the constructs, including AIPAC and SKUPerf (r = .62, p < .001), AIPAC and RevOpt (r = .58, p < .001), and SKUPerf and RevOpt (r = .66, p < .001). Regression findings confirmed that AIPAC significantly predicted SKU performance (β = .59, t = 10.21, p < .001; R² = .38; F(1,208) = 127.60, p < .001). In the dimension model, forecasting support (β = .24, p = .002), inventory and replenishment support (β = .19, p = .011), data integration quality (β = .16, p = .018), and governance plus user adoption (β = .27, p < .001) were significant, increasing explained variance to R² = .46. Revenue optimization was jointly explained by AIPAC and SKUPerf (R² = .52; F(2,207) = 112.40, p < .001), with SKUPerf the strongest predictor (β = .49, t = 8.02, p < .001) while AIPAC retained a direct effect (β = .29, t = 4.71, p < .001). These results imply that enterprises can improve SKU outcomes and revenue by strengthening predictive analytics capability end to end, prioritizing governance and adoption, disciplined forecasting, integrated data pipelines, and replenishment decision support so AI insights translate into measurable commercial gains in cloud analytics environments.