Production Cost Optimization and Market Acceptance Analysis: A Comparative Study of Coconut Milk vs. Palm Kernel Milk.

1. Executive Summary        

Business Problem: The need to reduce the Cost of Goods Sold (COGS) without compromising sensory acceptance and consumer purchase intent in key markets (Indonesia & Malaysia).

Key Findings (Based on Industrial Hypothetical Data):

1. COGS per mL: Coconut milk is estimated at IDR 23.7/mL; palm kernel milk is estimated at IDR 16.3/mL. This represents a potential raw material cost saving of 31% with palm kernel milk.
2. Physicochemical Performance: Palm kernel milk demonstrates higher emulsion stability, whereas coconut milk excels in traditional aromatic profile.
3. Preference & WTP (Willingness-to-Pay): Statistically significant differences exist based on country and demographic segments (urban/rural, income level, and product exposure).

Initial Recommended Decision: It is recommended to implement a dual-strategy:

1. Premium Products (Taste-First): Maintain coconut milk for premium culinary products that prioritize authentic taste.
2. Industrial Products (Cost-First): Adopt palm kernel milk for industrial applications, Ready-to-Drink (RTD), and creamers in cost-responsive market segments.

Decision Rule (Rule of Thumb): The transition to palm kernel milk is considered a "GO" if the delta (\Delta) in the liking score is \le 0.3 points (on a 9-point hedonic scale) AND the WTP for palm kernel milk is \ge WTP for coconut milk minus 5%.

2. Background & Study Objectives        

Operational Context: Palm kernel milk offers functional advantages, including better emulsion stability, consistent supply chain assurance, and a lower cost structure compared to traditional coconut milk.

Market Context: Consumer perception, especially for signature dishes, still strongly favors coconut milk due to its inherent "natural" and "traditional taste" attributes.

Case Study Objectives:

1. To validate the differences in physicochemical performance and cost structure (COGS per mL) between the two materials.
2. To measure and analyze consumer preferences and Willingness-to-Pay (WTP) across key demographics in Indonesia (ID) and Malaysia (MY).
3. To identify priority market segments and formulate commercial implications for product launch.

3. Research Design & Methodology        

This study utilizes a combined experimental design consisting of two main analysis packages:

3.1 Laboratory Analysis Package (Physicochemical & Shelf-Life)

1. Fixed Factors: Type (Coconut vs. Palm Kernel), Country (ID vs. MY).
2. Random Effect: Batch/Supplier (to control for raw material variability).
3. Response (Dependent) Variables: Emulsion stability (% separation at 24–72 hours), Viscosity (mPa·s), pH, Fat Content (%), Color Profile (CIELAB: L^*a^*b^*).
4. Covariates (ANCOVA): Extraction temperature, water:material ratio, homogenization pressure.
5. Statistical Model (Linear Mixed Model):$$$$$$

3.2 Consumer Analysis Package (Preference & WTP)

1. Design: Mixed ANOVA/ANCOVA.
2. Within-Subject Factor: Type (each respondent tests both samples in a blind test with randomized order).
3. Between-Subject Factors: Country (ID vs. MY), Demographics (Urban/Rural), Age (18–34, 35–49, \ge50), Income (Quartile Q1–Q4), Education, Cooking frequency, Exposure to palm kernel milk.
4. Response (Dependent) Variables: 9-point Hedonic Score (overall, taste, aroma, thickness), WTP (IDR/mL) via Gabor-Granger/Van Westendorp, Purchase Intent.
5. Covariates: Income, education, cooking frequency, exposure score.
6. Conceptual Model (Simplified): Score ~ Type * Country * (Urban/Rural) + Age + Income + Education + CookFreq + Exposure + (1|Respondent) + (1|Site)
7. Sample Size:

• LAB: 2 (Type) \times 2 (Country) \times 4 (Batch) = 16 batches, with triplicate measurements.
• Consumer: n=150-200 respondents per country (Total N=300-400) to detect a medium effect size with 0.8 power.

4. Production Operation Process (Summary)        

1. Raw Materials: Mature coconut (de-husked) vs. Palm kernels (moisture content \le 7\%).
2. Extraction:
3. Coconut: Wet process (grating \rightarrow pressing).
4. Palm Kernel: Mechanical disintegration + hot water addition (60–70°C) \rightarrow screw press \rightarrow 3-phase decanter for separation.
5. Stabilization: Pasteurization (85°C for 10–15 minutes); Homogenization (pressure \approx 15 MPa).
6. Asepsis/Shelf-Life: UHT option (135–140°C for 2–4 seconds) followed by aseptic packaging.
7. Quality Targets: Fat content 25–35%, pH 6.5–7.0; separation <5\% per 24 hours (room temp); neutral aroma profile (palm kernel) vs. savory (coconut).

5. Cost & Price per mL Analysis        

Cost estimation based on 2024–2025 market condition assumptions:

Coconut Milk:

• Raw Material: IDR 10,000/kg (yield \approx 450 mL)
• Process Cost: \approx IDR 1.5/mL
• Estimated Total COGS: \approx IDR 23.7/mL

Palm Kernel Milk:

• Raw Material: IDR 5,000/kg (yield \approx 350 mL)
• Process Cost: \approx IDR 2.0/mL (higher due to decanter process)
• Estimated Total COGS: \approx IDR 16.3/mL.

Difference: Palm kernel milk is \approx IDR 7.4/mL cheaper (\approx IDR 7,400/L).

Sensitivity Analysis (Quick Scenarios):

1. If kernel price increases by 10% \rightarrow Palm kernel COGS \approx IDR 17.0/mL (still maintains a 28% advantage).
2. If palm kernel yield drops by 10% \rightarrow Palm kernel COGS \approx IDR 18.1/mL (maintains a 24% advantage).

6. Key Results (Hypothetical Data)        

6.1 Laboratory Results (ANOVA/ANCOVA)

Interpretation: Palm kernel milk is statistically more stable (emulsion). Coconut milk is significantly thicker and has a higher fat content. The effect of country (ID vs. MY) was not significant, and controlling for process covariates did not alter the main significances.

6.2 Consumer Results (Mixed ANCOVA - Overall Liking)

Effect of Type: Small to moderate effect. Coconut milk excelled in aroma and taste attributes (delta 0.2–0.4 points).

Effect of Country & Interaction:

• Indonesia: Stronger preference for coconut (delta +0.3 points), especially for traditional dish applications.
• Malaysia: The preference gap was smaller (delta +0.1–0.2 points), particularly in urban segments with high exposure to creamers and RTD products.

WTP (IDR/mL, Marginal Means):

• ID-Urban (Q3–Q4): Coconut (25–27); Palm Kernel (24–26) \rightarrow Equivalent/Slightly below.
• MY-Urban (Q2–Q3): Coconut (24–25); Palm Kernel (24–25) \rightarrow Non-inferior.
• Rural & Q1: Highly price-sensitive. Palm kernel milk excels due to its lower effective selling price, with neutral sensory preference.

Significant Covariates: Cooking frequency (positively correlated with coconut preference); palm kernel product exposure (positively correlated with palm kernel acceptance); income (positively correlated with WTP).

Preference Conclusion: Non-inferiority (equivalent acceptance) was achieved in urban segments (ID & MY) and for middle-to-upper income brackets. Palm kernel milk can be adopted in these segments without a significant sensory penalty, assuming flavor calibration is performed.

7. Commercial Strategy & Product Positioning        

Coconut Milk (Taste-First Premium):

• Application: Rendang, curries, premium desserts, premium Horeca (Hotel/Restaurant/Cafe).
• Positioning: "Authentic, savory, traditional."

Palm Kernel Milk (Cost-First & Industrial):

• Application: Frozen sauces, instant soups, non-dairy creamers, RTD beverages.
• Strengths: Emulsion stability, flavor neutrality as a carrier.
• Positioning: "Stable, creamy, plant-based, industrial-process friendly."

Hybrid SKU:

• Formulation: 30–50% blend of palm kernel milk with coconut milk.
• Objective: To reduce costs by 10–18% while maintaining a flavor profile close to pure coconut.

8. Recommended Decisions & KPIs        

Go/No-Go Gate (per Target Segment):

1. \Delta Overall Liking (Palm Kernel vs. Coconut) \le 0.3 points (on a 9-point scale).
2. WTP (Palm Kernel) \ge 95\% WTP (Coconut) OR Margin increase \ge 6\%.
3. Emulsion stability (LAB) > 90\% (at 24 hours, 25–30°C).
4. Consumer complaint rate (phase separation) < 0.5\% of total sales in the first 3 months.

Launch KPIs (90-Day Target):

1. Modern trade penetration (urban): \ge 70\% of target listing.
2. Consumer repeat rate: \ge 30\%.
3. COGS reduction (industrial line): \ge 12\%.

9. Implementation Plan (8 Weeks)        

• Weeks 1–2: Finalize pilot batch SOPs (ID & MY); calibrate QC equipment (homogenizer, centrifuge); validate halal process.
• Weeks 3–4: Execute LAB tests (D0–D7); conduct consumer panel in Indonesia.
• Weeks 5–6: Conduct consumer panel in Malaysia; consolidate and clean all data.
• Weeks 7: Statistical analysis (ANOVA/ANCOVA, EMMs, post-hoc); create results dashboard.
• Weeks 8: SKU decision session; prepare for limited commercialization (A/B test in 2 cities).

10. Key Risks & Mitigation        

• Risk: Raw material variability.
• Mitigation: Contract with 2–3 suppliers; include 'Batch' as a random effect in the statistical model.
• Risk: Label bias (perception of "palm" vs. "coconut").
• Mitigation: Sensory testing must be blind; product information sessions held separately after testing.
• Risk: Perception of being "less natural."
• Mitigation: Educate on functional benefits (stability, plant-based); flavor tuning to mask any off-notes.
• Risk: Long-term stability.
• Mitigation: Optimize homogenization pressure; use natural emulsifiers (lecithin/carrageenan) if required and compliant with regulations.

11. Practical Appendices        

11.1 Data Schema (Column Design)

1. LAB Data: country | type | batch | temp_C | ratio | pressure_MPa | day | viscosity | stability_pct | pH | fat_pct | L | a | b
2. Consumer Data: respondent_id | country | site | urban_rural | age_group | income_quartile | education | cook_freq | exposure_score | condition(blind/info) | type | overall_liking | aroma | taste | thickness | WTP_per_ml | buy_intent

11.2 Analysis Template (R Syntax)

11.3 Draft Questionnaire (Key Sections)

1. Hedonic Test: (9-point scale: 1=Dislike extremely, 9=Like extremely)

• Overall Acceptance
• Taste
• Aroma
• Thickness

2. CATA (Check-All-That-Apply):

• [ ] Creamy
• [ ] Savory (Gurih)
• [ ] Neutral
• [ ] Leaves an aftertaste
• [ ] Looks stable/unbroken

3. WTP (Van Westendorp):

• At what price (per 100mL) would you consider this product too cheap?
• At what price would you consider this product cheap?
• At what price would you consider this product expensive?
• At what price would you consider this product too expensive?

4. Demographics & Habits: (Including cooking frequency and palm product exposure score).

12. Final Recommendations        

Based on the hypothetical data analysis, the strategic recommendations are as follows:

1. Industrial/RTD/Creamer Line: Full adoption of palm kernel milk is recommended to maximize cost advantages and functional stability.
2. Traditional Culinary Line (Premium): Retain coconut milk to protect brand equity. Consider developing a value SKU (hybrid/blend) for more price-sensitive segments.
3. Launch Priority: Focus on urban segments in Malaysia and urban Q3–Q4 segments in Indonesia, where data indicates strong sensory non-inferiority and moderate price sensitivity.
4. Financial Target: Target a minimum production cost saving of IDR 6–8 million per 1,000 L of production on converted lines, with no statistically significant sensory degradation.

Conclusion: Palm kernel milk provides financial and operational advantages (on the spreadsheet), while coconut milk retains its traditional sensory superiority (in the kitchen). The recommended strategy is to optimize margins from segments that are not chasing "nostalgic flavor" (industrial/RTD),

while aggressively protecting brand equity in the segments that value it (premium culinary). This strategy represents a balance between profitability optimization and brand equity protection.