Solving the Retail Inventory Discrepancy Problem: A DX Roadmap for Stocktaking and Ordering at Stores in Thailand

Target Readers: Executives, branch managers, store operations managers, and administrative staff at Japanese companies with stores or retail locations in Thailand. This article also addresses retailers, wholesalers, and service businesses operating multiple stores, as well as manufacturers running direct-sales shops or outlet stores attached to their factories. It is written for those facing challenges such as: “Book inventory and physical inventory do not match,” “Discrepancies appear every stocktake but the cause is unclear,” and “Ordering relies on intuition and experience, leading to stockouts and excess inventory simultaneously.”

The deepest and most chronically neglected problem on the retail floor is inventory discrepancy. Sales are recorded in the POS register, goods pile up in the warehouse and back room, and purchases and cost of goods sold appear side by side in the accounting system. Each component is functioning—yet when stocktaking time comes, the books and physical counts do not align. Small gaps are processed as “variance,” larger ones are booked as “loss.” Does this cycle repeat every month, every quarter?

This inventory discrepancy is not simply a counting error. In most cases, there are structural causes hidden somewhere in the daily flow of receiving, sales, returns, transfers, and disposal—points where records and physical reality diverge. And when inventory does not balance, the foundation for ordering, gross profit calculation, and stockout prevention all rest on sand. In Thai store operations, the added barriers of language differences between Japan and Thailand, business custom gaps, high workforce turnover, and over-reliance on individual know-how make the problem even harder to see.

This article first breaks down “why inventory discrepancies occur” from a structural perspective, then clarifies what to invest in and what to stop given the 2026 business environment. It then provides a practical, step-by-step retail DX roadmap centered on stocktaking and ordering—including investment decision criteria (such as a 3-year payback), common failure patterns to avoid, and how to sidestep them—grounded in the reality of operating in Thailand. The goal of this article is that by the time you finish reading, you will be able to see clearly which single store and which single form to tackle first, starting tomorrow.

One thing worth stating upfront: this problem is not the kind that “inserting the latest expensive system will solve in one shot.” In fact, the reverse is true—layering sophisticated tools on top of a shaky inventory accuracy foundation will only produce erroneous numbers at high speed. What this article consistently argues is the unglamorous but reliable sequence: solidify the foundation of inventory accuracy, then build ordering, integration, and advanced capabilities on top of it, step by step. Rather than being driven by buzzwords, change the numbers in the field one by one. That steady accumulation is what makes your Thailand operation one step stronger than the competition.

Why Does Retail Inventory Become Inaccurate in the First Place?

Attributing inventory discrepancies solely to “employee error” or “theft” leads to the wrong countermeasures. In reality, there are many more pathways by which inventory falls out of balance—and they are embedded in everyday operations. The following are typical patterns seen in stores in Thailand.

• Receiving inspection discrepancies: The quantity on the delivery note differs from the actual quantity received, but is recorded as “per delivery note” without verification. When delivery slips mix Thai, Japanese, and English, unit mix-ups (cases vs. individual units) are also common.
• Unrecorded sales transactions: Bundle deals, complimentary items, sampling, and in-store discount adjustments are not correctly reflected in the POS. “Special handling” that bypasses the register accumulates over time.
• Unprocessed returns and exchanges: Physical goods from customer returns are taken back, but the system inventory is not restored—or it is restored twice.
• Unrecorded inter-store and inter-warehouse transfers: Items moved “just for now” go unrecorded, appearing as excess inventory at Store A and a stockout at Store B.
• Unposted disposal and loss: Items past their expiration date, damaged goods, or items converted to samples are physically discarded without being removed from inventory.
• Master data inconsistencies: The same product is assigned multiple codes, or JANs are not linked to internal codes. This alone makes it impossible for inventory to ever balance.

The key point is that most of these are “unintentional mistakes.” This is precisely why countermeasures relying on individual attention (“let’s all be more careful”) result in recurrence. Solving inventory discrepancy requires changing the recording system itself.

The Real Cost of “Inventory Discrepancy” to the Business

While inventory discrepancies are treated lightly as “adjustment items at stocktaking,” the business is quietly incurring losses. The real damage lies not in the discrepancy amount itself, but in the losses occurring behind it.

• Lost sales opportunities due to stockouts: Because the system shows inventory as available, no order is placed—yet the item is actually out of stock, and sales are lost. Customers move to competitors.
• Excess inventory and cash tie-up: Conversely, because inventory appears insufficient, over-ordering leads to unsold stock, markdowns, and disposal. Cash is locked up in inventory.
• Reduced reliability of gross profit calculations: If inventory does not balance, cost of goods sold cannot be accurately calculated, making it impossible to identify which products or stores are truly profitable. Pricing strategy and product mix decisions go wrong.
• Bloated stocktaking labor: The larger the discrepancies, the more time is spent investigating causes and recounting, consuming valuable store staff time.
• Cost of explaining to headquarters: Branch managers are worn down by repeatedly having to explain to Japanese headquarters “why losses occurred.”

In short, inventory discrepancy simultaneously affects stockouts, excess inventory, gross profit, labor hours, and governance—it is the vital pressure point of retail management. Correcting it generates far more reliable returns than flashy new investments.

Conditions Unique to Stores in Thailand: Three Walls That Make Inventory Discrepancy Hard to See

Even with the same “inventory discrepancy” problem, stores in Thailand have unique conditions that obscure the issue and delay resolution. The reason Japanese approaches do not simply transfer over lies here.

Wall 1: Communication and Language Barriers Between Japan and Thailand

A typical structure has Thai staff handling floor operations while Japanese expatriates manage administration and headquarters reporting. In this setup, the “facts” and “causes” of inventory discrepancies often fail to be accurately conveyed. Staff may know that a discrepancy occurred, but if there is no system or habit for articulating and reporting the cause, the root issue stays buried. When slips and forms mix Thai, Japanese, and English, unit and product name mix-ups occur as well. Preparing a predefined set of discrepancy cause codes so that anyone can record findings using the same terminology is a practical solution for crossing the language barrier.

Wall 2: Workforce Turnover and Reliance on Individuals

In Thailand, employee turnover is faster than in Japan. If the ordering and stocktaking methods exist only in the head of a particular capable individual, operations collapse the moment that person leaves. Inventory discrepancies also tend to occur in the gaps during handovers. This is precisely why moving stocktaking procedures and ordering criteria from “people” to “systems” is even more important in Thai stores than in Japan. Standardization is both an efficiency measure and insurance against turnover risk.

Wall 3: “Conflicting Numbers” Between Headquarters and the Field

When inventory does not balance, the numbers the field sees and the numbers headquarters sees diverge. Headquarters feels “field management is lax,” while the field feels “headquarters does not understand the reality.” This cycle of distrust drains the energy available for improvement. Having POS, inventory, and accounting integrated so that everyone references one single set of numbers eliminates the discrepancy itself. What appears to be a technology problem is, in fact, also an organizational trust problem.

The 2026 Thailand Business Environment: Investments to Stop and Investments to Pursue

The 2026 Thailand economy presents a challenging backdrop, with the World Bank expressing a cautious growth outlook and the OECD noting risks related to external conditions, logistics, and energy costs—making it difficult to say that strong tailwinds are blowing. Personnel and logistics costs continue to rise, and securing and retaining talent remains a persistent challenge. In such an environment, a single-track approach of “growing revenue to solve problems” carries high risk; defensive DX focused on reducing current losses and protecting gross margins delivers reliable results.

At the same time, stopping all investment is not the answer. BOI (Thailand Board of Investment) continues to support investments including automation, AI, data analytics, and enterprise management IT. The axis for investment decisions is not “is it trendy?” but “will it change the numbers on the floor?” Use the framework below to sort which investments to stop and which to pursue.

Stocktaking DX: Rebuilding Inventory Accuracy First

The foundation of all retail DX is “correct inventory.” Demand forecasting, automated ordering, and AI analytics—if the underlying inventory data is wrong, they will only produce precise errors at scale. For this reason, the starting point of the roadmap must always be improving stocktaking accuracy.

Break Away from Paper and Excel-Based Stocktaking

In stores in Thailand, it is still common to count by hand on paper and then transcribe into Excel afterward. This approach generates transcription errors, double counts, missed items, and a lack of record of “who counted what and when.” Switching to barcode/QR reading via handheld devices or smartphones dramatically improves counting accuracy and speed.

Switch to Cycle Counting

The traditional “full stocktake”—closing the store once or twice a year to count everything—is burdensome and makes it impossible to trace the cause when discrepancies appear. Switching instead to “cycle counting,” where product groups are counted in small batches daily or weekly, enables early detection of discrepancies and easier identification of root causes. The recommended approach is to prioritize high-value and fast-moving items first.

Record “Cause Codes” for Discrepancies

When a discrepancy occurs, rather than simply adjusting the number, attach a cause code—”damage,” “expiry,” “receiving error,” “unrecorded transfer”—and log it. This makes it visible which process in the workflow is causing inventory leakage, and enables concrete countermeasures for prevention of recurrence.

Ordering DX: Turning Intuition and Experience into Visible Standards

Once inventory accuracy has been rebuilt, the next step is ordering. When ordering depends on individual intuition, the quality of orders collapses the moment that person leaves. Given workforce turnover in Thailand, standardizing the ordering process is also insurance against the risk of over-reliance on individuals.

• Establish reorder points: For each product, set a threshold below which an order is triggered, along with an order quantity, so anyone can make the same decision.
• Factor in lead times: Understand the delivery lead time for each supplier, and set safety stock levels to prevent stockouts. In Thailand, account for public holidays and logistics delays.
• Start with “operations” rather than “accuracy” for demand forecasting: Rather than aiming for sophisticated AI forecasting from the outset, begin with order proposals that reflect past sales history and seasonality, in a format where field staff can override judgments.
• Review promotional ROI daily: Ensure that products subject to discounts or promotions are genuinely selling with positive gross margin, verifiable by the next day.

Standardizing ordering simultaneously reduces both lost sales from stockouts and markdowns/disposal from over-ordering. These mechanisms can only function correctly when inventory is “accurate” to begin with.

Connecting POS, Inventory, and Accounting: Eliminating Data Silos

In many stores, POS, inventory management, and accounting operate on separate systems (or spreadsheets) and are connected manually through transcription. This fragmentation is the breeding ground for inventory discrepancies and gross profit opacity. The core of retail DX is creating a seamless flow in which POS sales automatically reduce inventory, receiving automatically increases inventory, and the results are correctly reflected as cost of goods sold in accounting.

When integration is achieved, it becomes possible to see in near-real-time “which store, which product, how many units, at what gross margin are selling right now.” Headquarters reporting shifts from month-end manually compiled Excel files to a dashboard where the same numbers are always accessible. Reducing the unproductive back-and-forth from “conflicting numbers” between Japan and Thailand lowers the explanation burden on branch managers—a significant value to the field.

There is no need to build perfect integration all at once. Even starting with just one point—”POS sales automatically reduce inventory”—eliminates manual transcription and raises inventory accuracy by one level. From there, receiving, transfers, and disposal can be connected incrementally. What matters is eliminating one by one the places where data is entered manually two or three times. The fewer manual entries, the fewer errors, and the more inventory balances. Viewing integration not as a large-scale system overhaul but as a daily accumulation of eliminating transcription lowers the barrier to getting started considerably.

Daily Store Reports and Converting Findings into Tasks: Connecting Data to Action

A common DX failure is becoming satisfied once a dashboard is built. Visibility alone does not change field operations. What matters is converting identified issues into tasks—”who does what by when”—and tracking execution.

For example, digitize the daily store report so that stockouts, complaints, equipment failures, and observations can be recorded on the spot. Convert recorded issues into improvement directives as tasks, and make response status visible. This moves the operation from a state where paper daily reports are filed away and never read, to a state where the loop of “record → directive → action → verification” keeps turning. Tools for digitizing field forms, such as i-Reporter, serve as the entry point for this recording and task conversion loop.

A Phased Implementation Roadmap: Start with One Store, One Form

Retail DX that launches simultaneously across all stores with all features at once almost always fails. The field cannot master it, results cannot be measured, and the push-back of “the old way was easier” reverses progress. What TOMAS TECH consistently recommends is starting with a small unit—one store, one warehouse, one form, one meeting—measuring the effect, ensuring adoption, and only then rolling out horizontally.

IoT, Automation, AI, and Accounting DX: Where Do They Deliver Results?

When people hear “automation” and “AI,” they tend to imagine large-scale investments right away—but in the context of retail inventory and ordering, there are more grounded applications. Applied in the right sequence, these technologies become reliable tools.

• IoT and sensors: For stores handling refrigerated or frozen goods, IoT-based temperature monitoring reduces quality loss and disposal. Applications also exist for tracking inventory depletion physically through weight sensors in back rooms or sensors on shelves. The established approach is to start with “one high-loss item.”
• Automation: Handheld reading for stocktaking, automatic generation of purchase orders, and automatic consolidation of headquarters reports—replacing manual transcription tasks with machines delivers the highest cost-effectiveness. Verify at the planning stage whether these qualify for BOI automation incentives.
• AI and demand forecasting: Order proposals based on past sales history, seasonality, weather, and events reduce both stockouts and excess inventory. However, the prerequisite is accurate inventory and sales data. Layering AI on top of dirty data produces only precise errors. This is why AI is positioned as the final step.
• Accounting DX: Connecting POS and inventory to accounting to automatically and accurately calculate cost of goods sold and gross profit frees operations from month-end manual aggregation and improves both the accuracy and speed of headquarters reporting. A connected system is also more stable for meeting Thai tax and accounting requirements.

The common principle is one: “Do not start with flashy technology—solidify the foundation of inventory accuracy first.” Only when the foundation is in place do automation and AI return results commensurate with the investment.

Investment Decision Criteria: Think in Terms of a 3-Year Payback

When presenting to Japanese headquarters, qualitative appeals of “this will be more convenient” or “this is the latest technology” do not carry weight. Investment decisions must be expressed in numbers. The basic approach is to total the annual savings and improvements that can be realized, and determine whether initial and operating costs can be recovered within three years.

The benefits that can be accumulated through retail DX include, for example, the following. Please substitute your own company’s actual figures when calculating (specific amounts and ratios are not provided here—it is important to calculate using your own company’s data).

• Reduction in inventory discrepancies and losses (disposal, theft, unrecorded shortfalls)
• Gross profit increase from recovering lost sales due to stockouts
• Reduction in inventory holding costs and disposal from compressing excess inventory
• Labor cost (man-hour) savings on stocktaking, ordering, and headquarters reporting
• Gross profit improvement through optimized discounting and promotions

Estimate these on an annual basis, and determine whether the cumulative three-year benefit exceeds the investment amount. In addition, confirm before committing to investment whether BOI incentives applicable to automation, AI, and enterprise management IT can be utilized—this can change the payback period assumption.

Common Failure Patterns and How to Avoid Them

Retail DX failures follow highly reproducible patterns. Simply knowing them in advance allows you to avoid many detours.

• Installing systems before cleaning up master data: Deploying a system while leaving duplicate product codes or unregistered JANs in place means inventory will never balance.
  Avoidance: Make master data cleanup the first task before deployment.
• Rolling out to all stores simultaneously: Expanding without a pilot causes problems to erupt at every store at the same time.
  Avoidance: Build the model at one store, solidify the results and operating procedures, then expand.
• Failing to involve the field: Tools decided by headquarters are rejected by store staff as “something imposed on us.”
  Avoidance: Incorporate field voices into the design; make operation as simple as possible. Do not underestimate the importance of Thai-language support.
• Becoming satisfied with a dashboard: Visualization becomes the goal, and behavior does not change.
  Avoidance: Include as a set a mechanism to convert issues into tasks and track responses.
• Not measuring results: The system was implemented, but it is impossible to explain what improved and by how much.
  Avoidance: Capture baseline KPI values (inventory discrepancy rate, stockout rate, stocktaking labor hours, etc.) before deployment and compare before and after.

The TOMAS TECH Perspective

We at TOMAS TECH are based in Bangkok and have been supporting the development of systems used on the floors of Japanese-affiliated manufacturing, retail, and logistics operations in Thailand and across ASEAN. For the problem of inventory discrepancy, our inventory management system PEGASUS plays the role of connecting receiving, sales, transfers, and stocktaking in a single dataset—eliminating the recording fragmentation that creates the “gap between books and physical stock.” Only when inventory is correctly organized does the next step—standardizing ordering and achieving real-time gross margin visibility—begin to carry real meaning.

Daily store reports, receiving inspection records, and stocktaking checklists—these field forms can be digitized using the paperless form application i-Reporter, enabling the loop from recording to improvement directives and task conversion to operate continuously. This moves operations away from transcription errors on paper and “daily reports that were written but never read,” and connects field insights to management action. Additionally, an operations monitoring system for tracking equipment and store operations status, and a smartwatch notification system for reliably delivering communications to staff, can be combined incrementally to match the realities of each operation.

What we value is not selling a multitude of features, but solving the reader’s challenges—”inventory does not balance,” “ordering is dependent on individuals,” “explaining to headquarters is exhausting”—by starting small and solving reliably. Starting with one store and one form, demonstrating results in numbers, and rolling out only after the approach has taken root: this is the approach we believe leads to reproducible outcomes at Thailand locations. Please feel free to reach out for a consultation at https://tomastc.com/contact.

Summary

The chronic retail problem of “inventory discrepancy” does not stem solely from employee error or theft—it arises from structural recording gaps spanning receiving, sales, returns, transfers, disposal, and master data. And this problem quietly but reliably damages the business through stockouts, excess inventory, gross profit opacity, stocktaking labor, and the cost of explaining to headquarters.

Thailand in 2026 is an environment in which it is difficult to rely on strong growth. For this reason, defensive DX that reduces current losses and protects gross margins is effective. The order of what to do is clear: first rebuild inventory accuracy through stocktaking, then standardize ordering, integrate POS, inventory, and accounting, and finally advance with demand forecasting and AI. Rather than launching across all stores at once, start with one store and one form; use a 3-year payback as the decision standard; measure results; and roll out horizontally. This grounded approach reliably strengthens store operations in Thailand. Start with your first step by measuring your own inventory discrepancy rate.

References

• World Bank Thailand (Thailand Economic Outlook)
• Thailand BOI (Investment Promotion for Automation, AI, Enterprise Management IT, etc.)
• JETRO Thailand (Investment and Business Information for Thailand)
• S&P Global PMI (Manufacturing and Business Sentiment Trends)
• Ministry of Economy, Trade and Industry – Manufacturing White Paper 2025