Why did Apple miss the mark? Learning from the 'MacBook Neo' production surge: The limits of demand forecasting in 2026
To succeed in overseas expansion, it is essential to grasp the realistic needs on the ground. Recently, I met with the overseas sales manager of a beauty equipment manufacturer in Gangnam, Seoul.
The company, preparing for a full-scale entry into the Japanese market, reportedly prepared three times the usual amount of inventory for test marketing.
'Japanese buyers told us they were very interested, so we rushed to produce two containers' worth. However, concrete discussions have not progressed at all since then. Only the warehouse storage fees are piling up every month.'
We receive consultations every day from export companies that are at the mercy of 'forecasts' for overseas markets and are struggling with mountains of inventory and costs.
In this modern age where data analysis technology has evolved so much, why do we continue to miss the mark on demand forecasting?
From the 'forecasting trap' that even Apple, with its world-class supply chain, fell into, we will unravel the limits of demand forecasting in the latest global business of 2026 and its solutions.
The limits of demand forecasting: The '30% gap' that shook the world's best supply chain
Failure in demand forecasting is not just a problem for small and medium-sized enterprises with limited resources.
Apple, a global IT company, is also facing the wall of demand forecasting.
According to data from market research firm IDC, Apple's Mac shipments in the first quarter of 2023 recorded a dramatic decline of 40.5% year-on-year.
The cause is said to be a misreading of the speed at which the 'special demand' during the COVID-19 pandemic would end, leading to delays in supply chain production adjustments.
'The growth curve calculated from past trends becomes nothing more than a mathematical formula in the face of discontinuous market changes.'
These are the words of a former procurement manager at a major electronics company, reflecting on the market turmoil at the time.
Apple possesses the world's highest-level AI algorithms and thousands of data scientists.
Even so, they could not perfectly predict the 'complex variables' of global inflation, rising interest rates, and changes in consumer behavior.
The more sophisticated the forecasting model becomes, the more it becomes trapped by the curse of past data.
This is the essence of the 'limits of demand forecasting' that every global company is facing as of 2026.
The 'forecasting trap' that companies blindly trusting data fall into when expanding into the Japanese market
Why are demand forecasts so often wrong?
The '2023 White Paper on Manufacturing' published by the Ministry of Economy, Trade and Industry presents interesting data regarding the status of digital technology utilization in the Japanese manufacturing industry.
Only 34.1% of companies overall report that they are effectively utilizing data in production management and demand forecasting.
The remaining approximately 70% of companies face challenges such as a lack of data itself or a failure to improve forecast accuracy even when data is utilized.
As we observed B2B sales in South Korea and Japan, three structural factors that lead to inaccurate forecasts became apparent.
1. The 'Bullwhip Effect' that paralyzes the supply chain
The 'Bullwhip Effect' refers to a phenomenon where minor fluctuations in customer demand are amplified into significant volatility as they travel upstream through the distribution channel (to manufacturers and raw material suppliers).
A buyer's casual remark of 'I'm a little interested' becomes inflated into 'signs of a major hit' as it passes through trading companies and export agents.
As a result, the manufacturer's production floor receives instructions for a 'production surge,' leading to an overflow of excess inventory in the market.
2. Survey data that ignores the 'reasons not to buy'
Many companies conduct local surveys or FGI (focus group interviews) before entering the Japanese market.
However, the responses to questions like 'Would you buy this if it were available?' do not translate into actual purchases.
This is because people cannot truly realize their own needs until the moment they pay.
'A buyer who says it's wonderful in a survey goes silent the moment you send a quote. This is the most common scene in international sales.'
3. The 'time axis' gap between South Korea and Japan
This difference in time axes significantly distorts forecasts, especially when South Korean startups approach the Japanese market.
If the decision-making speed of a Korean company is '1,' the decision-making process of a traditional large Japanese B2B company requires '3 to 6' times as much time.
Forecasts stating that 'this much revenue will be generated within this quarter' almost certainly get pushed back because they fail to account for Japan's internal approval systems and consensus-building processes (nemawashi).
I noticed something while organizing the data.
When we analyzed the internal platform data at RINDA, we found an interesting trend.
The companies with the highest probability of receiving inquiries from overseas buyers were not those that had prepared perfect market forecast reports in advance.
Rather, they were the companies that took only their product specifications and a minimum number of samples and approached 100 buyers directly in the first month.
We call this the 'empirical approach' within our company.
No matter how detailed a market report you read, you cannot know 'how much, at what price, and how many' a buyer in front of you will purchase right now.
Even in JETRO's '2023 Survey on Overseas Business Operations of Japanese Companies,' 'communication with local partners' and 'understanding customer needs' are consistently ranked as the top challenges in overseas business.
Rather than piling up predictions on a desk, it is about repeating validations through actual sales activities.
This is the only logic to minimize the probability of failure in overseas expansion in 2026, where uncertainty is increasing.
Case studies of successful overseas expansion by giving up on 'prediction' and using 'empirical measurement'
Here, I would like to introduce the case of Company H, a Korean precision parts manufacturer that we supported.
Company H was planning to export new industrial sensors to the Japanese market.
Initially, they hired a major consulting firm and spent several million yen to create a 'sensor demand forecast report for the Japanese market.'
The report's outlook was bright, stating that a demand of 5,000 units per year could be expected.
However, although they believed that forecast and started developing Japanese distributors, only 12 units were sold in six months.
The 'market size' written in the report was real, but it lacked 'living variables' such as existing competitor penetration, compatibility with standards, and the difficulty of customization required by Japanese worksites.
'The time spent reading the forecast report was like a drug to gain a sense of security.'
The representative of Company H looks back on the situation at that time like that.
So, Company H switched its strategy from 'prediction-dependent' to 'high-speed empirical measurement.'
Utilizing RINDA's AI agents, they began a direct approach to 200 niche Japanese manufacturers that required specific specs.
The results were astonishing.
In three weeks, there were inquiries about specific specifications from 18 companies, and from four of them, concrete feedback was obtained: 'If you can change the size of this part, we would like to introduce it starting next month.'
This real-time buyer voice is the living demand forecast.
Company H immediately modified the specifications and acquired its first full-scale large order in just three months.
The 'Three Empirical Approaches' for 2026 Overseas Expansion and B2B Sales
In an era where perfect demand forecasting does not exist, how should we approach the global market?
I propose three concrete actions derived from primary information on the ground.
First Principle: Lower the Planning Ratio, Raise the Feedback Ratio
Reduce the time spent creating expansion plans by half and invest those resources into 'test approaches to buyers'.
The market's answer is not found in consultant reports, but only in the reply emails from buyers.
Second Principle: Keep Supply Chain 'Buffers' Dynamic
Keep initial production volumes low and build a system that responds to demand fluctuations by 'shortening lead times' (time from production to delivery).
In most cases, the cost of emergency air freight is cheaper than the cost of holding inventory.
Third Principle: Deploy 'Multi-Angle Sales' Utilizing AI
Instead of narrowing down to one target segment, approach multiple industries simultaneously on a small scale.
For example, a company manufacturing cosmetic containers could simultaneously approach the 'pharmaceutical packaging' and 'food container' industries.
By utilizing AI agents, you can execute these simultaneous approaches to multiple segments without incurring additional human costs.
This is an approach of casting a wide net and immediately strengthening the areas where you get a bite, rather than trying to predict where demand will spike.
How to walk through an unpredictable future
No matter how much technology advances, it is impossible to predict tomorrow's weather or a buyer's budget changes next month with 100% accuracy.
It is an extremely high-risk challenge for us to pour our limited resources into demand forecasting that even Apple gets wrong.
The companies that win in global business are not those that predict the future most accurately.
**They are the companies that realize their predictions were wrong the fastest and pivot on the spot.**
Is your company spending too much time and budget creating 'forecast reports' for overseas markets that you have yet to see?
First, try starting your live demand measurement by sending a 'single letter' to the buyers right in front of you.
Regarding the challenge of entering overseas markets, please feel free to share your own experiences from the field or any struggles you have with demand forecasting in the comments section.
Frequently Asked Questions (FAQ)
Q1. Why are traditional demand forecasting methods failing in overseas expansion?
A1. The modern global market has too many 'discontinuous variables,' such as inflation, geopolitical risks, and rapid shifts in consumer behavior. It has become difficult to accurately capture market volatility and the live need of buyers using only demand forecasting models based on historical data.
Q2. What is the most effective B2B sales approach for successfully entering the Japanese market?
A2. Instead of a 'predictive approach' that creates perfect market research reports at a desk, use a 'measured approach' where you take minimal samples to approach actual buyers directly and verify market reactions. This allows for a realistic approach that accounts for the decision-making processes and time-gap characteristics unique to Japanese business.
Q3. Please tell me how to accurately identify a buyer's true need.
A3. Do not take simple surveys or verbal promises of 'interest' at face value; instead, measure 'actual feedback' through real business negotiations, price quotes, or test implementations. True need only becomes clear when a customer pays money or makes specific customization requests.
[Free] Will your product sell overseas? Information on our 'Market Demand Measurement Diagnosis'
At RINDA, we provide a service that utilizes advanced AI technology to rapidly measure the real demand for your company's products in the global market.
Why not test the potential for market entry based on actual buyer data, rather than theoretical arguments?
If you are interested, please feel free to contact us via the link below.
Click here for inquiries and casual meetings
Rinda | B2B Global Sales AI Agent for Overseas Expansion
Please feel free to contact us via LINE at any time for consultations or inquiries.
