Staff at Nigerian Breweries who report using AI forecasting and inventory tools also report better sustainability performance, with those measures explaining 56.7% of variation — but the evidence is correlational from a purposive single-firm survey and does not prove causation.

Main Finding

AI-driven demand forecasting and AI-based inventory optimization are positively and significantly associated with improved sustainability outcomes in the manufacturing supply chain at Nigerian Breweries (Lagos State). A multiple regression model including these AI applications explains 56.7% of the variance in sustainability measures; estimated coefficients reported were 4.287 for AI-driven demand forecasting and 3.549 for AI-based inventory optimization.

Key Points

• Context: Study examines how AI applications in the supply chain (demand forecasting and inventory optimization) affect sustainability in a Nigerian manufacturing setting.
• Main quantitative results:
  • Coefficient (AI-driven demand forecasting) = 4.287 (positive, significant)
  • Coefficient (AI-based inventory optimization) = 3.549 (positive, significant)
  • Model R² = 0.567 (56.7% of variance in sustainability outcomes explained)
• Practical interpretation (authors’ conclusions): AI forecasting reduces overproduction and spoilage; AI inventory optimization reduces excess stock and improves resource use — both supporting economic and environmental sustainability.
• Recommendations (from the paper): Nigerian Breweries should adopt AI-powered demand forecasting for production planning and implement AI-based inventory optimization systems to reduce waste and excess inventory.
• Theoretical framing: Resource-Based View (RBV) — AI capabilities treated as strategic, VRIN-like resources that can improve firm efficiency and sustainability.

Data & Methods

• Population and sample:
  • Population: 950 supply-chain employees at Nigerian Breweries, Lagos State.
  • Sample target (Yamane formula): 281; responses received: 278 (98.9% response rate).
  • Sampling: purposive (judgmental) sampling targeting employees involved in forecasting and inventory.
• Data collection:
  • Instrument: Structured questionnaire with 5-point Likert scale (1 = Strongly Agree … 5 = Strongly Disagree).
  • Constructs: AI-driven demand forecasting, AI-based inventory optimization, sustainability outcomes.
  • Reliability & validity: Face and content validity checked; Cronbach’s alpha > 0.70.
• Analysis:
  • Software: SPSS v27.
  • Descriptive statistics (frequencies, means, SDs) and inferential analysis (multiple regression).
• Sample demographics (n = 278):
  • Gender: 71.2% male, 28.8% female.
  • Age: largest group 30–34 years (45.3%).
  • Education: 77.7% HND/B.Sc, 22.3% OND/NCE.
  • Experience: 44.6% had 6–10 years’ experience.
• Methodological limitations to note (inherent in design):
  • Cross-sectional, self-reported survey from a single firm limits causal inference and generalizability.
  • Purposive sampling may introduce selection bias.
  • Sustainability outcomes were survey-measured rather than independently audited (potential measurement bias).

Implications for AI Economics

• Microeconomic/firm-level implications:
  • Efficiency gains: AI forecasting and inventory optimization can lower inventory carrying costs, reduce waste and spoilage, and improve matching of production to demand — increasing firm productivity and lowering unit costs.
  • Resource allocation: Better forecasts reduce overproduction and capital tied up in stock, freeing resources for other investments or lowering working-capital needs.
  • Competitive advantage: Under RBV logic, AI capabilities can be strategic assets that are hard to replicate, potentially yielding sustained performance differences.
  • Labor and skills: Adoption increases demand for data/AI skills; potential for reallocation of labor toward analytics and process oversight, with implications for wages and training costs.
• Macroeconomic and environmental externalities:
  • Aggregate adoption across firms could reduce resource waste at the sectoral level, lowering environmental footprints (material use, emissions from unnecessary production/transport).
  • Scale effects: widespread AI adoption could shift industry cost structures, influence prices, and affect market entry/exit dynamics in manufacturing.
• Policy and investment implications:
  • Adoption barriers noted in literature (and relevant for Nigeria): infrastructure deficits, implementation costs, regulatory uncertainty, and skills shortages — suggesting roles for public policy (subsidies, training programs, digital infrastructure investment, standards/regulation).
  • Policymakers should consider incentives (grants, tax relief) and capacity-building to lower adoption frictions and magnify sustainability gains.
• Research and measurement directions for AI economics:
  • Quantify monetary effects: estimate cost savings, ROI, and payback periods from AI forecasting/inventory systems.
  • Environmental accounting: measure direct reductions in material waste and emissions attributable to AI adoption (life-cycle or input-output approaches).
  • Causal identification: use panel data, quasi-experimental designs, or randomized rollouts to estimate causal impacts on productivity, employment, and emissions.
  • Distributional effects: study how AI adoption affects wages, employment composition, and skill premiums in manufacturing.
  • Market structure: evaluate how differential adoption alters competition, concentration, and market efficiency in manufacturing sectors.
• Operational recommendation for researchers and practitioners:
  • For practitioners: prioritize pilots with measurable KPIs (inventory days, spoilage rates, waste volumes, production-match metrics), track financial and sustainability outcomes, and invest in workforce training.
  • For researchers/policymakers: support firm-level experiments and multi-firm panels to generate generalizable causal evidence on economic and environmental impacts.

If you’d like, I can (a) produce a one-page executive summary formatted for managers highlighting costs/benefits and implementation steps, or (b) draft suggested empirical designs to identify causal effects of AI adoption on sustainability and productivity.