Reducing Food Retail Waste Losses: Connecting Expiration Dates, Markdowns, and Ordering Through DX

Target Readers: Executives, site managers, store operations managers, and purchasing/logistics/administrative staff at Japanese companies operating food retail, supermarkets, convenience stores, or food processing divisions attached to retail outlets in Thailand and ASEAN. This article is written for operations that handle short-shelf-life products such as fresh produce, prepared foods, daily essentials, and frozen foods, and that are struggling with waste losses and markdown losses.

Profitability in food retail is determined less by dramatic sales growth and more by how effectively you minimize the small, daily losses. Particularly in short-shelf-life categories such as fresh produce, prepared foods, and daily essentials, unsold products are first marked down and eventually discarded, creating a triple cost burden: cost of goods, labor, and disposal fees. For Japanese companies operating stores in Thailand, managing this “hard-to-see but steadily margin-eroding loss” is one of the highest-priority management challenges of 2026.

The World Bank has expressed a cautious outlook on Thailand’s growth in 2026, pointing to risks in the external environment as well as logistics and energy costs. In an environment where significant sales growth is difficult to achieve, “protecting margins by eliminating waste” has a greater management impact than “growing by selling more.” In other words, reducing waste and markdown losses is not merely a shop-floor improvement initiative — it is a management strategy that directly boosts gross margins.

This article breaks down the structure of why food retail waste losses occur, analyzing it through three key nodes — expiration management, markdowns, and ordering — and explains how to implement DX that connects these nodes with data. It also covers how to distinguish investments to pause from those to pursue, how to evaluate implementation on a three-year payback basis, the unique challenges of the Thai operation environment (Japan-Thailand communication, over-reliance on key personnel, labor shortages, and BOI utilization), and specific failure patterns along with phased implementation strategies. Our goal is to provide decision-making guidance for those who seek not DX as a buzzword, but DX that changes real numbers on the shop floor.

Why Waste Losses in Food Retail Never Seem to Decrease

The primary reason waste fails to decrease at most stores is that “waste only becomes visible after it has already occurred.” Products go unsold, their expiration dates pass, they are disposed of in the back room, and only then is the quantity recorded. By that point, however, no one can accurately trace the cause — whether it was because the items simply did not sell, because too much was ordered, because markdowns were applied too late, or because there were issues with product placement or restocking.

Waste losses are the result of a chain of three interconnected decisions. First, ordering (what to stock and how much); second, expiration and freshness management (in what sequence to sell through received inventory); and third, markdowns (when and by how much to reduce prices on items at risk of going unsold). These three decisions are frequently made by different people at different times, and they are typically fragmented across paper order forms, store managers’ gut instincts, and handwritten notes near the register. Because the data is not connected, it is impossible to verify after the fact how a 10% reduction in ordering volume affected waste levels.

In Thai stores, this fragmentation tends to be even more acute. Ordering decisions are made by Japanese managers or headquarters, restocking and freshness checks are handled by Thai staff, and the final call on markdowns is concentrated with the store manager. Because language and roles are divided, shop-floor insights — such as “this product always has leftovers on Fridays” — never feed back into order adjustments, and the same waste is repeated week after week. The combination of over-reliance on key personnel and communication barriers structurally preserves waste.

Another frequently overlooked cause is a sense of resignation at the shop floor level — the attitude that “some waste is inevitable.” When the premise is that a certain level of waste is unavoidable for fresh and prepared food categories, waste never rises to the level of a management agenda item. However, once you present data on where your waste rate stands relative to the industry, which categories are outliers, and whether there are patterns by day of week or time of day, you can separate “reducible waste” from “unavoidable waste.” At most stores, waste that could be reduced is simply left unaddressed, hidden among the “unavoidable” category. Visualizing waste and placing it on the table for discussion is the starting point for improvement — and a mindset shift that must precede any tool implementation.

Managing Waste Losses and Markdown Losses as Distinct Metrics

The first step toward improvement is recognizing “waste loss” and “markdown loss” as separate concepts. The two are often lumped together as “loss,” but their management implications are entirely different.

Waste loss means the full cost of goods sold for unsold products becomes a direct loss, further compounded by disposal fees. It has the most severe impact on gross margin and is something you want to drive as close to zero as possible. Markdown loss, on the other hand, represents the difference between the original price and the discounted price — but the cost of goods is still recovered and disposal is avoided. In other words, markdowns serve a purpose as a “necessary cost to prevent waste” and are not purely negative.

One further overlooked issue is that the “timing” and “rate” of markdowns are left entirely to individual shop-floor judgment. Even for the same product, there are days when a 20% reduction three hours before closing sells out everything, and days when a 50% reduction one hour before closing still leaves items unsold. This variation is driven by factors such as foot traffic that day, weather, and nearby competitor activity — but if the only decision-making tool available is the store manager’s experience, results are not reproducible. By accumulating markdown history and its outcomes (sold out or leftover), you can identify data-driven patterns such as “for this category, a 30% markdown two hours before closing minimizes waste most effectively,” moving decision-making from an individual’s judgment to a standardized process.

The core problem is that no one is monitoring the balance between these two metrics with actual numbers. If you fear markdowns and apply them too late, waste increases. If you apply markdowns too early, you end up discounting items that would have sold at full price, eroding gross margins. The optimal point is not “zero waste, minimum markdowns” in the abstract, but rather “minimize waste while keeping markdowns to what is necessary and sufficient” — and that optimum can only be found through data, not intuition. Only by connecting POS sales data, inventory, expiration dates, and markdown history can you identify the optimal approach by category, day of week, and time of day.

The Big Picture: DX That Connects Expiration, Markdowns, and Ordering

The core focus of this article is connecting the three fragmented decision nodes through data. The ideal form is the following loop.

First, record expiration (use-by) dates at the time of receiving goods. When registering received goods via barcode or handheld device, attach the expiration information to the item so that inventory is managed as “date-tracked inventory.” Next, cross-reference POS sales data against date-tracked inventory to make visible “how many units need to be sold and within how many days.” For products where the sell-through pace is insufficient, the system automatically flags them as markdown candidates. Once a markdown is applied, record the markdown history and the final outcome (sold out or discarded), then feed that back into the next ordering cycle. Once this loop is running, waste shifts from “something we record after it happens” to “something we act on before it happens.”

One critical point is not to aim for a perfect system from the outset. Many failures begin with large-scale investments such as “simultaneous POS integration and AI demand forecasting across all stores and all categories,” only to stall because the shop floor cannot effectively use the system. As discussed later, the standard approach is to start with one store and one category (for example, prepared foods only), measure results, and then expand from there.

Implementing only one element of this loop in isolation produces limited results. For example, if you begin recording expiration dates but do not connect that data to POS sales pace, you still cannot determine “when to apply markdowns.” Conversely, if you establish markdown rules without understanding your date-tracked inventory, you have no visibility into which products the rules apply to. This is precisely why “connecting” is the essence of the matter — simply aligning individual tools side by side is not DX. Only when they are connected does the improvement cycle of ordering → freshness → markdowns → waste outcomes → next order begin to turn, allowing you to break free from repeating the same waste week after week.

Investments to Pause vs. Investments to Pursue

In a year like 2026 where growth is difficult to predict, the goal is not to stop all investment but to be selective. Categorizing DX investments in food retail as “pause/proceed with caution” versus “pursue” yields the following breakdown.

The standard is simple: “Can it measurably reduce waste, markdowns, or labor hours in a quantifiable way?” If an investment cannot answer that question — regardless of what trending keywords it carries — it is perfectly acceptable to pause it for now. Conversely, measures that deliver small but verifiable results, when compounded, protect gross margins.

Where to Apply IoT, Automation, and AI

Automation and AI in food retail are powerful tools when applied at the right entry point. The following outlines priorities in order of importance.

1. Freshness and Temperature Monitoring (IoT)

Temperature deviations in refrigerated and frozen cases directly affect both waste and food safety risk. Simply recording temperatures around the clock with sensors and generating alerts when deviations occur can prevent large-scale waste events where “by the time anyone noticed, everything had spoiled.” This delivers clear, easily understood results and involves relatively low investment, making it the ideal first automation initiative.

2. Automated Alerts for Expiration and Inventory

This involves automatically flagging products where sell-through pace will not meet expiration deadlines by cross-referencing date-tracked inventory against POS sales pace. Having a person visually check every SKU every day is unrealistic, but a system can compile a “list of items to mark down today” each morning. To ensure adoption, markdown rates should also be standardized in rules that allow Thai staff to act decisively without hesitation.

3. AI Demand Forecasting Can Come Last

AI demand forecasting is appealing, but without a solid foundation of clean data on sales, inventory, expiration dates, weather, and events, accuracy will suffer. The sound approach is to first improve data quality through the “visualization” and “alerting” measures described above, allow order accuracy to stabilize through manual processes, and then pilot AI forecasting in a limited set of categories. Getting the sequence wrong risks making predictions that miss the mark — and when shop-floor staff lose confidence in the system, recovery is difficult.

Connecting to Accounting and Management DX to Visualize “Gross Margin”

The impact of waste loss reduction only becomes meaningful in management terms when it appears as “improved gross margin rate” in accounting and management reports. Yet in most stores, waste quantities are tracked in a backroom notebook, markdowns are in the POS system, and cost of goods is in a headquarters spreadsheet — all existing in separate silos. Connecting these to produce monthly — ideally weekly — figures for “waste loss amount,” “markdown loss amount,” and “effective gross margin rate” by category is what sustains the momentum for continuous improvement.

In Thailand, accounting and tax requirements (VAT, withholding tax, BOI-related reporting, etc.) also apply, and the disconnect between operational data and management accounting creates delays in headquarters reporting. DX that connects ordering, inventory, waste, and sales to accounting delivers benefits beyond waste reduction — including faster month-end closes and reduced burden for headquarters reporting. Aim for a state where you can speak in numbers: not “things became more convenient,” but “gross margin improved by X points and administrative hours decreased by Y hours.”

The BOI Perspective: Incorporating Automation, Data, and Enterprise Management IT Into Your Investment Story

The Thailand Board of Investment (BOI) supports investments that include automation, AI, data analytics, and enterprise management IT. In the context of food retail and distribution, inventory management, automation, and data infrastructure development are worth considering as potential areas of support. The key is not to decide on investments first and then think about BOI applications as an afterthought, but to build your investment story with BOI utilization in mind from the planning stage.

Since the specific program details, eligible scope, and incentive conditions change over time, always verify the latest BOI information and consult with specialists (this article does not specify particular incentive rates or amounts). What matters is being able to present “inventory and automation investment for waste reduction” to headquarters not as a standalone expense, but framed within the context of BOI support and a three-year payback.

BOI support also involves a certain administrative workload for applications and reporting. Here too, having your ordering, inventory, waste, and sales data organized makes it easier to prepare documentation explaining investment outcomes and equipment utilization rates. Conversely, if shop-floor data is scattered across paper or fragmented spreadsheets, both applications and reporting become significantly more burdensome. In other words, building the data foundation for waste reduction also becomes the groundwork for advancing BOI utilization. Rather than running investment, support, and operational improvement as separate projects, adopting the perspective of bundling them into a single investment story makes internal consensus-building easier.

Implementation Decision-Making: Think in Three-Year Payback Terms

When presenting to Japanese headquarters, “the shop floor will become more convenient” is not enough to secure a budget. What needs to be communicated is a story backed by numbers: three-year payback, risk reduction, quality improvement, and reduction in administrative hours. Waste loss reduction is a topic that lends itself well to this type of explanation.

For example, if you can identify the monthly waste loss and markdown loss amounts for a store’s prepared food category, you can project “if we reduce waste by a certain percentage through expiration management and standardized markdown rules, what would be the annual gross margin improvement?” Laying that gross margin improvement figure alongside the costs of the system, devices, and operations, and assessing whether payback is achievable within three years, is the basic approach. A critical point here is to always measure a baseline (current waste and markdown amounts) before implementation. Without a baseline, you cannot prove improvement.

The Reality of Thai Operations: Key-Person Dependency, Communication, and Labor Shortages

Stores where waste does not decrease even after system implementation share a common characteristic: improvement depends on specific individuals. Markdown decisions are concentrated with a single store manager, and when that manager is absent, markdowns are delayed and waste increases. Order expertise exists only in the heads of certain veteran staff, and when they transfer, accuracy drops. This is the key-person dependency problem.

In Thai stores, the additional challenge of Japan-Thailand communication compounds the issue. Information does not flow smoothly between the Japanese staff making ordering decisions and the Thai staff observing freshness conditions on the shop floor, making it difficult for shop-floor insights to be reflected in improvements. The root cause is not only a language barrier but also the absence of a mechanism for “knowing where to report what you’ve noticed.” This is precisely why making freshness checks, markdown decisions, and daily store reports operable by anyone following the same procedure (standardized, digitized, converted to task workflows) is more important than implementing any tool. In an environment of continued labor shortages, operations that do not depend on specific individuals are what protect shop-floor capability.

Failure Patterns and How to Avoid Them

The following are common failure patterns in food retail waste reduction DX, along with avoidance strategies.

Failure 1: Large-scale simultaneous rollout that stalls. Implementing across all stores and all categories at once causes shop-floor confusion and the system goes unused. The avoidance strategy is to start with one store and one category, measure results, and then expand.

Failure 2: Building a dashboard and considering the job done. Clean charts are produced, but no one changes their next action. The avoidance strategy is not to stop at “seeing” the data, but to drive it all the way to specific actions such as “products to mark down today” and “products where ordering should be reviewed.”

Failure 3: Starting without measuring a baseline. After implementation, you find yourself unable to determine “whether it actually had any effect,” losing the basis for justifying continued investment. The avoidance strategy is to always record current waste and markdown amounts before implementation.

Failure 4: Leaving decisions to individuals without standardizing rules. Markdown decisions are left entirely to shop-floor discretion, preserving individual dependency and waste. The avoidance strategy is to standardize markdown timing and rates into documented rules that anyone can follow.

Phased Implementation Roadmap

The following outlines a realistic approach across three phases.

Phase 1 (Building the Foundation): Begin with one category at one store (e.g., prepared foods) — recording expiration dates at receiving, monitoring temperatures, and measuring current waste and markdown amounts. This phase establishes the baseline.

Phase 2 (Connecting the Data): Cross-reference POS, inventory, and expiration data; introduce automated markdown alerts and standardized markdown rules. Begin feeding results back into ordering decisions and review waste and markdown trends on a weekly basis.

Phase 3 (Expanding): Roll out the proven approach to other categories and stores, and integrate with accounting and management reporting. For categories where order accuracy has stabilized, begin piloting AI demand forecasting.

At each phase, confirm “have the numbers improved?” and only move forward after improvement is visible. This discipline is what keeps DX from ending as a trend and transforms it into a system that protects gross margin.

One important note when expanding: do not simply copy the exact approach from the successful pilot store to all others. Location, customer demographics, product mix, and staff proficiency vary by store, and optimal markdown timing and order volumes will differ. What should be transferred is not “specific markdown rates” but the framework and mindset: “record expiration dates, cross-reference with sell-through pace, apply markdowns according to rules, and feed results back into ordering.” Creating a state where each store can fine-tune its rules based on its own data is what sustains continuous improvement. If you also standardize the baseline measurement method established in Phase 1, you can compare across stores and benchmark performance, enabling early identification of underperforming locations.

The TOMAS TECH Perspective

At TOMAS TECH, we are here to help not through hard selling, but through the lens of how technology can address our readers’ challenges. In the context of reducing food retail waste losses, the following combination is effective.

The inventory management system PEGASUS serves as the foundation for centralized management of receiving, inventory, and shipping, supporting date-tracked inventory and category-level loss visibility. Making visible “which products had leftovers, when, in what quantities, and at what waste/markdown cost” is the starting point for improvement. The paperless application i-Reporter digitizes paper order forms, freshness check sheets, and inventory records, enabling anyone to input and report following the same procedure. This transforms Thai staff’s shop-floor observations directly into data, lowering the barriers of individual dependency and communication. Furthermore, our operations monitoring system and smartwatch system make work status and notifications in store back rooms and processing departments visible, supporting operations that never miss the timing for markdowns or restocking.

What matters is not implementing all of these at once, but starting small with one store and one category, measuring results, and building adoption gradually. For implementation inquiries, please feel free to reach out at https://tomastc.com/contact.

Summary

Waste losses in food retail recur because the three decision nodes of ordering, expiration management, and markdowns are fragmented and disconnected. Connecting these through data, and shifting operations from “recording waste after it occurs” to “acting before it occurs,” is the core of DX that directly protects gross margin.

In a year like 2026 where growth is difficult to forecast, the right move is not to stop all investment, but to selectively pursue measures that can demonstrably reduce waste, markdowns, and labor hours in measurable terms. Avoid large-scale simultaneous rollouts and the self-satisfaction of dashboard creation. Start with one store and one category, measure baselines, standardize markdowns into rules, confirm results, and then expand. This disciplined, phased approach is what protects shop-floor capability even amid individual-dependency challenges and labor shortages, and what produces outcomes that can be presented to headquarters as a three-year payback. Begin your DX not as a buzzword, but as a practice that changes real numbers on the shop floor — and start with a single small step.

References

• World Bank Thailand
• Thailand BOI (Thailand Board of Investment)
• JETRO Thailand
• Ministry of Economy, Trade and Industry — Manufacturing White Paper 2025
• S&P Global PMI