Handy Terminal Implementation 2026: Cost, 2D Barcodes and ROI

Friday evening, a customer calls: the wrong part number arrived. Month-end stocktaking costs eight people their Saturday, and the book still does not match the floor. In the office, figures from paper tags are re-keyed into the core system. Handy terminal implementation is usually where the conversation goes next, and that is where it stalls: what does this actually cost, and who do we even ask? This article works through the same order you will follow in practice: what really drives the price, the one hardware decision you cannot postpone in 2026, and how the payback period moves in a model calculation built around a Japanese-owned plant in Thailand.

Why Handy Terminal Implementation Is Back on the Agenda

Handheld terminals are not new technology. Yet they keep returning to the agenda at Japanese-owned plants across Thailand and ASEAN, and there are two reasons: one on the quality side, one on the labour cost side. The starting point is being able to state your own current situation as a number.

Mis-shipment Rates Are Measured in PPM

“We ship the wrong thing too often” is a feeling. To put it into a capital request, you need a unit. Logistics operations use PPM (parts per million). The formula is:

Mis-shipment rate (PPM) = number of mis-shipments / total transactions × 1,000,000

If you ship 100,000 orders a year and 80 of them go out wrong, that is 80 / 100,000 × 1,000,000 = 800 PPM. Looking at the raw count, you cannot tell whether “80 a year” is good or bad. Converted to PPM, it becomes comparable against industry levels.

The levels discussed in the Japanese logistics industry are broadly summarised as follows.

On top of that, the cost of handling a single mis-shipment (re-delivery freight, complaint handling, goodwill gifts and so on) averages 3,000-5,000 JPY. For in-house warehouses, one source notes that an investment in the range of 100,000-300,000 JPY up front is often enough to improve PPM substantially.

Be precise about the scope here. These are benchmarks drawn mainly from Japanese e-commerce and mail-order logistics. You cannot assert that “the mis-shipment rate of manufacturers in Thailand is 800 PPM”, and the 800 PPM used later in this article is only an assumption for the model calculation. The unit and the formula themselves, however, transfer to any industry. Start by counting your shipments and mis-shipments over the last 12 months and calculating your PPM. Whether or not you have that single number completely changes the quality of every subsequent conversation with a vendor.

One more thing: in B2B parts shipping, the weight of a single mis-shipment is different from e-commerce. If a wrong part number reaches the customer’s line, you pay not only for return freight but for an emergency delivery, sorting work at the customer’s site and, in the worst case, the response to a line stoppage. So rather than applying the Japanese e-commerce benchmark directly, pull up a handful of your own recent mis-shipment cases and build up the actual cost and man-hours they consumed. That build-up becomes your company-specific equivalent of the “1,800 THB per case” used in the model calculation below.

Labour Cost in Thailand Is Rising in Total Employer Burden, Not Daily Wage

The other reason is labour cost. This is where the misunderstanding usually sits.

The daily minimum wage in Thailand ranges from 337 to 400 THB, with a national average of roughly 374 THB per day. The 400 THB daily rate applies in Bangkok, Phuket, Chachoengsao, Chonburi, Rayong, and Ko Samui district of Surat Thani. The fact that Chonburi and Rayong, where manufacturing clusters, sit in the 400 THB band matters directly to plants in the Eastern Seaboard.

What moves the number more than the daily rate itself is the total employer burden. From January 2026, the ceiling on the earnings base used to calculate Social Security Office (SSO) contributions was raised from 15,000 THB to 17,500 THB per month. That raises the monthly employer contribution cap. On top of that, the Employee Welfare Fund (EWF) is scheduled to start in October 2026, adding one more layer of mandatory contributions.

In other words, reading the daily wage table alone does not capture the rise in labour cost. In processes run on “headcount × hours” – verification, stocktaking, re-keying – the effective hourly cost keeps climbing quietly. When you put labour cost into a capital request, use an hourly rate inclusive of employer contributions, not the daily wage. That is exactly why the model calculation in this article puts shop-floor operators at 65 THB per hour: 400 THB per day / 8 hours = 50 THB per hour, plus an assumed allowance for employer contributions.

Vietnam Raised Wages by an Average of About 7.2% in January 2026

For companies running production across ASEAN, the Vietnamese side is worth checking too. Decree 293/2025/ND-CP (promulgated 18 November 2025, effective 1 January 2026) replaced the previous Decree 74/2024/ND-CP.

Hourly rates were revised at the same time.

The increases range from 250,000 to 350,000 VND per month, or about 7.2% on average. Watch how you round this. It is not “7.2% in every region”; it is 7.05% to 7.25% depending on the region, averaging about 7.2%. In documents that line up multiple countries side by side, rounding differences like this come back later as inconsistent numbers.

If you have sites in both Thailand and Vietnam, evaluate handy terminal implementation with an additional question in mind: can the setup you build in Thailand be rolled out to Vietnam? If you can transplant the same screens and the same operating rules, the cost at the second site drops sharply because the development work does not repeat. Build something bespoke at each site and you double both the maintenance and the training.

But “Labour Is Expensive, So Automate” Will Not Get Approved

Having just spent several paragraphs on labour cost, here is the opposite point, and it is the central message of this article.

The cost of handy terminal implementation is decided by integration development, not by the price of the devices. Payback comes faster when you phase the rollout and limit both the number of units and the number of functions. And if you are choosing hardware in 2026, a 1D-laser-only device rather than a 2D imager risks a replacement purchase after GS1 Sunrise 2027.

Rising labour cost is a legitimate trigger for the review, but it is not the main driver of payback. As the model calculation later shows, stacking up time savings alone stretches the payback period to nearly seven years in the scenario where you build every function. What actually moves the needle is the reduction in mis-shipments, and the decision to keep the investment small in the first place. Hold on to that distinction from the start, and you will read vendor quotations differently.

The Cost of Handy Terminal Implementation Splits into Five Layers

The main reason quotations cannot be compared side by side is that every vendor draws the boundary of “what is included” in a different place. Break the cost into five layers first, then re-sort each vendor’s quotation into those five layers. That alone surfaces mismatches such as a quotation that looked cheap because it never included layer 3.

Layer 1, Devices: Look at Service Life Before Price

Device prices in Japan run roughly 50,000-100,000 JPY for entry models, 100,000-200,000 JPY for standard models, and from 200,000 to over 300,000 JPY for high-spec models. Another source puts the benchmark at 100,000-200,000 JPY per unit with a service life of 5-7 years.

What matters here is service life rather than the price itself. Once you accept that you are choosing, today, a machine you will use for 5-7 years, the 1D versus 2D decision discussed below stops being a matter of preference.

Note also that these are Japanese domestic benchmarks. Procurement prices in Thailand shift with import costs and the local service structure, so do not convert them directly; treat them as an order-of-magnitude reference. That is precisely why the model calculation in this article is built in THB.

Layer 2, Peripherals: Beware Quotations Missing Cradles and Access Points

Charging cradles, spare batteries, protective cases and straps, and wireless access points. These are smaller amounts than the devices, but a quotation that omits them will not run on the floor. The most commonly overlooked item is the wireless access point. Office Wi-Fi may reach fine, but coverage drops at the back of a warehouse lined with metal racking, and at the truck bays. Decide as a specification what happens when someone scans in a dead spot: does the device buffer offline and send later, or does it error out on the spot?

Layer 3, Application and Integration: This Is the Biggest Line Item

Now the main point. Of the five layers, the one most likely to be the largest is layer 3, integration development, not the devices.

In the model calculation later in this article (Scenario A, where four functions are built from scratch), development accounted for about 55.8% of the initial investment, and the devices themselves for about 26.3%. In other words, arguing about the unit price of a device has far less effect on the total than deciding how tightly to scope the integration work.

The “anywhere from tens of thousands to several hundred thousand JPY” figure in the source refers to lighter integration: configuring the standard handheld functions a package already ships with, or exchanging data via CSV. Developing a dedicated real-time interface to your core system for four functions – shipping inspection, goods-receipt verification, stocktaking and item tag printing – is a different order of magnitude. When you read a quotation, always unpack what the single word “integration” is actually referring to.

If your scope extends to collecting production results, the design has to cover the upstream system alongside the handhelds, not the handhelds alone. How far you go on production-result integration changes the cost, and that is covered in our article on the cost of process management systems.

Layer 4, Annual Maintenance and MDM: Comparing Initial Cost Alone Does Not Work

Hardware maintenance, software maintenance and MDM (mobile device management) licences recur every year. MDM is the mechanism for managing multiple devices together: pushing application updates, distributing settings in bulk, remote lock and remote data wipe. Even at around five devices the operation itself is workable by hand, but every application update costs you the man-hours of collecting all the units and swapping the software one by one, every single time. Somewhere past ten devices, that approach stops being realistic at all. Note that the model calculation later in this article includes MDM licences in Scenario B as well, even at five devices.

Annual costs are less visible than the initial investment, but they accumulate across the 5-7 year service life. Always compare vendors on initial cost + annual cost × expected years of use.

Layer 5, Operations and Training: The Layer Most Often Missing

Master data cleanup, operator training, and the labour cost of running on paper in parallel. This layer tends to vanish from the quotation and reappear as internal man-hours. Master data cleanup in particular is the number one failure pattern discussed below. You cannot attach a barcode to an item that is not registered in the part master, and if you do, it will not match on scan.

What You Must Decide in 2026: 1D Laser or 2D Imager

This is the decision most often deferred during hardware selection, and the one that is hardest to undo later. The conclusion first: if you are buying in 2026, buy a 2D imager.

What GS1 Sunrise 2027 Actually Is

GS1 Sunrise 2027 is an international initiative led by GS1 to migrate from 1D barcodes (EAN/UPC) to 2D barcodes at the point of sale. By the end of December 2027, retail POS systems are expected to be able to accept 2D barcodes. Preparation is under way across 48 countries and territories, representing roughly 88% of global GDP.

Let us clear away the misunderstanding people get hyped about first. 1D barcodes are not disappearing in 2027. During the transition, 1D and 2D coexist, and many products will carry both. GS1 has asked manufacturers to add 2D codes to packaging in addition to the existing 1D codes, at least through 2027.

And this is a retail distribution story that starts at the POS. If someone tells you that “all the barcodes in your factory stop working in 2027”, that is not accurate. There is no deadline after which the 1D barcode on the item tags you use inside your own site becomes unreadable.

The Reason to Choose a 2D Imager Anyway Is Service Life

So why does this affect hardware selection? The reason takes us back to cost.

A device has a service life of 5-7 years. A unit purchased in 2026 will be in use until somewhere around 2031 to 2033. Over that period, there is a real possibility that more of the goods and delivery labels arriving from your trading partners will carry 2D codes. A 1D-laser-only device reads linear barcodes by design, so it cannot read 2D codes such as GS1 QR Code or GS1 DataMatrix. A 2D imager reads an image with a camera, so it handles both 1D and 2D.

This is therefore not a story about “something happening in 2027”. It is a procurement decision about whether it is acceptable to lock equipment you will use for 5-7 years into 1D only, today. If unreadable codes multiply a few years from now and you have to swap out the fleet, that replacement cost is one you would never have incurred by choosing a 2D imager at the outset.

2D Codes Can Carry Data, and That Is What Helps Traceability

There is a second, more direct benefit for manufacturers. 2D codes (GS1 QR Code and GS1 DataMatrix) can carry the following:

• Product identification (GTIN)
• Lot and batch number
• Expiry date
• Serial number
• Consumer-facing links

In addition, GS1 Digital Link turns the identifier into a web link, connecting to product information, instructions for use, sustainability information and promotional content. GS1 DataMatrix can “hold structured product information while keeping the footprint on the packaging small”, and is widely used in healthcare and manufacturing.

This points in the same direction as the traceability requirements your customers are already imposing. If a single code can carry lot and expiry date as well as the part number, you can detect lot reversal at the point of shipping inspection, and follow-up investigations can be traced from the code on the physical item.

There is an operational implication on the shop floor, though. EAN/UPC codes are normally pre-printed on packaging, but GS1 2D codes often contain production data (date, line, lot number and so on) and therefore need to be printed at the point of packing or immediately before it. In other words, the assumption shifts from applying pre-printed labels to issuing them on the spot. Decide where the label printing system sits – at the end of the line, or in the packing area – at the same time as you select the devices.

Dedicated Handhelds, or Smartphones plus a Business App?

This is another branch that comes up constantly. “Wouldn’t smartphones be cheaper?” is a proposal that often comes from the executive level. Here is what you need to decide it.

The axis for the decision is not the device price but the annual cost per unit. The source notes that because dedicated handhelds are rugged and long-lived, their annual cost per unit is often close to that of a smartphone, and not uncommonly lower. A cheaper device that has to be replaced more often narrows the gap on an annual basis.

A practical way to settle it: processes where continuous scanning is the core of the work, such as full-day shipping inspection or stocktaking, call for dedicated handhelds. Supporting uses – checking an item a few times a day, or a supervisor querying status – work fine on a smartphone. If you mix both, keep the application identical across them; training and maintenance get much easier.

Model Calculation: Scenario A (Build Everything) versus Scenario B (Shipping Inspection First)

This is the core of the article. Everything below is a model calculation based on assumptions, not the actual results of any specific company. Substitute your own numbers when you use it.

Shared Assumptions

A few notes. The 65 THB/hour for shop-floor operators is an assumption: 400 THB per day / 8 hours = 50 THB per hour, plus employer contributions such as social security. The 1,800 THB per mis-shipment is an assumed total of emergency replacement freight, response man-hours and customer handling; it is set above the Japanese benchmark of 3,000-5,000 JPY (approx. 667-1,111 THB) because B2B parts shipping adds the cost of responding to line-stoppage risk. The 800 PPM mis-shipment rate is an assumption placed near the middle of the “500-2,000 PPM without barcode verification” range. That gives 100,000 × 800 / 1,000,000 = 80 cases per year.

Scenario A: Build Every Function (Full Scratch Integration, 10 Devices)

Scope covers four functions: shipping inspection, goods-receipt verification, stocktaking and item tag printing. A dedicated interface to the existing production management system is newly developed.

Initial investment (THB)

At the assumed rate of 1 THB = 4.5 JPY, the 33,000 THB device price converts to a level that sits inside the Japanese standard-model band (around 100,000-200,000 JPY). Note that these 10 units do not include a spare. If you want to hold a spare, request quotations with one additional unit of the device and its peripherals added on (33,000 + 3,000 + 1,500 + 600 = 38,100 THB). The same applies to the 5 units in Scenario B.

What this table is meant to show is not the total but the ratio inside it. Development accounts for about 55.8% of the initial investment; the devices themselves for about 26.3%. It is immediately clear that a decision to narrow the development scope moves the number far more than negotiating the device price.

Annual cost (THB)

Annual benefit (THB)

The breakdown: benefit 2 is 4.5 hours/day × 300 days = 1,350 hours/year, and 1,350 × 65 = 87,750. Benefit 3 is goods-receipt verification at 1.5 hours/day × 300 days = 450 hours, plus stocktaking saving 28 hours per monthly count (8 people × 8 hours = 64 hours, reduced to 8 people × 4.5 hours = 36 hours) × 12 counts = 336 hours; 450 + 336 = 786 hours, and 786 × 65 = 51,090. Benefit 4 is 2 hours of re-keying per day × 300 days = 600 hours, and 600 × 90 = 54,000.

Note that benefit 2 is verification work on the floor while benefit 4 is re-keying into the core system in the office. They are separate processes, so this is not double counting. Someone will raise this in almost every internal review, so write the distinction into your materials in advance.

Annual net benefit A = 318,840 – 135,000 = 183,840 THB

Simple payback period A = 1,255,000 / 183,840 = approx. 6.8 years (6.83 years)

Scenario B: Shipping Inspection First, Using Standard Package Functions (5 Devices)

Scope is shipping inspection only. It uses the standard handheld functionality already included in the existing inventory or production management package, with no scratch development.

Initial investment (THB)

Annual cost (THB)

Annual benefit (THB)

Because Scenario B only covers shipping inspection, the assumption is that the mis-shipment rate does not fall to 100 PPM but stops at 150 PPM, since mix-ups at goods receipt remain. Being optimistic here breaks the whole calculation, so hold the discipline: if you narrow the scope, discount the effect accordingly.

Annual net benefit B = 231,750 – 80,500 = 151,250 THB

Simple payback period B = 442,500 / 151,250 = approx. 2.9 years (2.93 years)

A versus B: 2.84 Times the Investment, 1.22 Times the Net Benefit

Side by side:

• On initial investment, A costs about 2.84 times what B costs (1,255,000 / 442,500 = 2.836).
• On annual net benefit, A delivers only about 1.22 times what B delivers (183,840 / 151,250 = 1.2155).

You multiply the investment by 2.84, and the annual net benefit multiplies by only 1.22. That is why a phased rollout wins.

The reason is straightforward. The two pillars of the benefit – reducing mis-shipments and cutting shipping inspection time – are almost fully captured by that single function, shipping inspection. Add goods-receipt verification, stocktaking and item tag printing, and the cost rises honestly as development spend while the benefit grows only modestly.

What the 10-Year ROI Looks Like

Taking the 5-7 year service life into account and assuming hardware is reinvested at the same amount in year 6, here is the 10-year view.

Over a 10-year span the gap widens further. Scenario A only just reaches positive territory after a full decade, and a single assumption going wrong along the way tips it back into the red. Scenario B builds a cumulative total well above the amount invested over the same period.

The practical conclusion is a sequence: stand up shipping inspection alone on standard package functionality first, measure the actual effect, then extend to goods-receipt verification, stocktaking and item tag printing. If phase one has a measured payback record, phase two is far easier to get approved.

Whether standard package functionality is sufficient depends on what upstream system you are running. A handheld only works if there is an upstream system holding the inventory and instruction data to check against, so if that upstream layer does not exist, deploying handhelds alone will not produce results. The options for the upstream system are covered in our comparison of inventory management systems.

The depa 200% Deduction: Some Companies Qualify, Many Do Not

Thailand offers a tax incentive for digital investment. Under Royal Decree No. 802, a 200% deduction is allowed on expenditure for computer software, hardware, smart devices and digital services, whether purchased, commissioned or subscribed (an additional 100% deduction on top of the amount spent). It applies only to products and services registered in the Thailand Digital Catalog (registered with depa, the Digital Economy Promotion Agency).

The eligibility conditions and caps are as follows.

For a company that does qualify, taking the 300,000 THB cap, the additional 100% deduction and an assumed corporate tax rate of 20%, the tax saving is 300,000 × 100% × 20% = 60,000 THB. The effective initial investment for Scenario B becomes 442,500 – 60,000 = 382,500 THB, and the payback period = 382,500 / 151,250 = approx. 2.5 years (2.53 years).

Here, however, is the point that deserves the most caution. The conditions of 5 million THB or less in paid-up capital and 30 million THB or less in revenue are conditions that many Japanese-owned manufacturers in Thailand do not meet. The majority exceed both thresholds.

So the sequence is this. First, confirm with your accounting and finance team whether your company qualifies. If it does not, do not build the capital request on the assumption of this incentive. Getting approval on a payback period that bakes in a 60,000 THB tax saving, only to discover later that you are not eligible, damages confidence in the entire plan. Companies that do not qualify should leave the 60,000 THB out of the calculation and use the approx. 2.9 year figure instead.

Even if you do qualify, there is one more check. The incentive covers only products and services registered in depa’s Thailand Digital Catalog, so you need to ask the vendor whether the product is already listed in the catalog before you request a quotation. On top of that, the tax authorities require you to retain, for each item of expenditure, a formal tax invoice, proof of payment and documentation confirming depa registration. Fix those documentation requirements internally from the outset.

Five Patterns That Make Implementations Fail

With the money settled, the next topic is the shapes failure takes. Sites where handy terminal implementation does not work out share a common set of causes.

1. The Master Data Is Not Clean

This is the most common cause. Items on the floor that are not registered in the part master, duplicate registrations of the same part number written differently, records left behind for discontinued parts. With those in place, a scan will not match, and the floor arrives at the conclusion that “it throws errors, so we do not use it”.

The countermeasure is simple but not light: start the master data cleanup during the review stage, running it in parallel with requirements definition and the quotation process. In terms of the 90-day approach below, treat it as work that runs alongside the measurement in Day 0-15. Who fixes the master data, at what point, and from which screen? Hand out handhelds before those rules exist, and the floor stops every time a master data gap appears after go-live.

2. Operating Design on the Floor Was Left Until Later

“We will work out the operating rules once the system is built” fails. Things you need to decide up front include:

• When a scan does not match, who handles it and by what procedure?
• Can a quantity discrepancy be passed on the operator’s judgement, or does it always require supervisor approval?
• If a label is damaged and unreadable, is manual entry permitted? If so, is it recorded?
• If the shipping cut-off is at risk, may the system be bypassed? (Allow this exception and the whole thing unravels.)

Defer the design of exception handling and an exception will occur on day one of go-live, someone will decide on the spot to “use paper today”, and it will stay that way.

3. Nobody Looked at the Network

The back of the warehouse, the aisles between metal racking, the truck bays, the outdoor staging area. Office Wi-Fi may reach, but in these places it drops. Before implementation, take actual signal measurements in every area where work happens. If additional access points are needed, that is a line item belonging in the initial investment. That is exactly why the model calculation above includes additional access points in the initial investment.

4. Spare Devices and MDM Are Not in the Budget

Devices get dropped. When they are dropped, work stops. On a floor running five devices, one failure means covering the same workload with four for the several days it is out for repair. Get quotations for a unit count that includes spares.

The same goes for MDM. Collecting every device and swapping software by hand each time the application is updated does not scale with the number of units. The annual licence is a cost, but it is cheaper than paying the man-hours for every update round.

5. Paper Stayed On as Permanent Parallel Running

The last one. Start parallel running without deciding how long it lasts, and paper stays forever. And as long as paper is still there, benefit 4, the elimination of office re-keying, never materialises. Recall that in Scenario A of the model calculation, eliminating re-keying accounted for 54,000 THB per year. Set a calendar date for stopping paper before you go live.

Parallel running itself is necessary. The point is to define the exit condition in advance – “X days after go-live”, or “once the error rate falls below X%” – rather than leaving it at something vague like “until it stabilises”.

A 90-Day Approach

Here is a way to structure the path from review to investment decision in 90 days. Follow this order and, by the time you are requesting quotations, your own numbers will already be in hand.

Day 0-15: Measure

Spend the first two weeks buying nothing and counting. Count four things: mis-shipments and total shipments over the last 12 months (which gives PPM), the hours per day currently spent on shipping inspection, the total hours for one stocktake (headcount × hours), and the hours spent re-keying in the office.

Those four become the input values for the benefit calculation directly. The figures used in the model calculation (8.0h/day, 28 hours saved per monthly count, 2 hours of re-keying per day) are assumptions, so replace them with your own measurements.

Day 16-30: Target PPM and a One-Page Requirements Definition

Once you have your current PPM, set a target. The target is not “zero”; put a number on it. Then write the requirements definition on a single A4 page. Requirements that do not fit on one page are too many at this stage.

What to write: the functions in scope (shipping inspection only, or also goods receipt and stocktaking), the sites in scope, the expected number of devices, the touchpoints with existing systems, and the policy for exception handling.

Day 31-50: Judge Whether Standard Package Functions Suffice

This is the single biggest branch determining cost. If your existing inventory or production management package includes standard handheld functionality, start by confirming what it can do. Scenario B of the model calculation is only viable if this assessment comes back yes.

Do the assessment not by skimming a feature list, but by taking the one-page requirements from Day 16-30 and sorting each line into “standard”, “achievable through configuration” or “requires development”. If only a handful of items require development, examine whether those few can be handled through operating rules instead.

Day 51-70: Request Quotations on Identical Conditions and Compare Across the Five Layers

Request quotations from several vendors. Nothing is comparable unless the conditions match, so standardise the unit count, the functions in scope, the integration method, the maintenance coverage and whether training is included, all in writing.

Re-sort the quotations you receive into the five layers described earlier (1. devices, 2. peripherals, 3. application and integration, 4. annual maintenance and MDM, 5. operations and training). Compare on initial cost + annual cost × expected years of use, not on initial cost.

Day 71-90: On-Site Testing and Investment Decision

Borrow real devices and test them in the actual working environment. Check whether the signal reaches every area, whether your existing labels can be read, whether the device can be operated while wearing gloves, whether the battery lasts a full day’s work, and whether the operators say “yes, we can work with this”.

Calculate the payback period from the test results and move to the investment decision. Over these 90 days you will have assembled four inputs: PPM, hours, quotations and hands-on experience with the hardware.

Frequently Asked Questions

What is a handy terminal?

It is a ruggedised mobile computer for business use that reads barcodes and 2D codes and lets you verify and register data on the spot. In factories and warehouses it is used for shipping inspection, goods-receipt verification, stocktaking and item tag printing. Its main role is to send scan results to the upstream system so that book inventory and physical inventory agree. These devices are rugged, with a service life benchmark of 5-7 years.

How much does handy terminal implementation cost?

The cost splits into five layers: 1. devices, 2. peripherals, 3. application and integration, 4. annual maintenance and MDM, 5. operations and training. On Japanese domestic benchmarks, devices run around 50,000-100,000 JPY for entry models, 100,000-200,000 JPY for standard models, and from 200,000 to over 300,000 JPY for high-spec models. Peripherals include charging cradles at 5,000-20,000 JPY and spare batteries at 3,000-10,000 JPY.

What drives the total most, however, is layer 3, integration development. In the model calculation in this article (based on assumptions), Scenario A, which builds four functions from scratch, has an initial investment of 1,255,000 THB, of which development is about 55.8% and the devices about 26.3%. Scenario B, which stands up shipping inspection alone on standard package functionality, comes to 442,500 THB. Because the cost varies so heavily with the number of devices and functions in scope, we recommend narrowing to a single process before you request quotations.

Can smartphones do the job instead?

It depends on the use case. For processes centred on continuous scanning, such as full-day shipping inspection or stocktaking, a dedicated handheld is the better fit. For supporting uses, such as checking an item a few times a day, a smartphone works.

Base the decision on the annual cost per unit rather than the device price. Because dedicated handhelds are rugged and long-lived, their annual cost per unit is noted as being close to that of a smartphone, and not uncommonly lower. Either way, you still need device management through MDM.

How long does implementation take?

This article sets out a 90-day approach from measurement to investment decision. The actual build period varies considerably depending on whether you use standard package functionality or develop new integration with your core system. If the work is mainly configuring standard functions and cleaning up master data, it is shorter; developing a dedicated interface extends it accordingly.

If you want to compress the timeline, the single most effective move is to bring master data cleanup forward. Master data work can begin during the review stage, and if it slips, go-live slips with it regardless of which approach you choose.

Can it integrate with our existing production management system?

In most cases there are several integration options: using the standard handheld functionality of the package, exchanging data through files such as CSV, or developing a dedicated interface. Because cost and duration change substantially with that choice, ask the vendor “which integration method will you use?” rather than “can you integrate?”.

The prerequisite to confirm is whether the upstream system actually holds the inventory and shipping instruction data. A handheld only works when there is something to check against. It is not unusual to conclude that the upstream system has to be sorted out first.

Should we choose RFID instead?

It depends on the objects and the volumes. RFID’s major advantage is bulk reading of tags, but tag cost scales with the number of objects. Barcodes, by contrast, have a low label issuance cost and suit workflows that verify one item at a time.

Useful criteria include the unit value of the objects, how many items you handle at once, and whether a tagging step can be inserted into your existing work. The cost structure is covered in detail in our article on RFID implementation costs, which is worth reading alongside this one. The two are not mutually exclusive; some sites use barcodes for shipping inspection and RFID for asset management.

Summary

The key points of this article:

1. Measure mis-shipments in PPM. Number of mis-shipments / total transactions × 1,000,000. In-house warehouses without barcode verification can run at 500-2,000 PPM, and 100 PPM or below is treated as the first target in e-commerce logistics. Start by calculating your own PPM for the last 12 months.
2. Cost splits into five layers, and the largest line item is integration development, not devices. In Scenario A of the model calculation (based on assumptions), development was about 55.8% of the initial investment and the devices about 26.3%.
3. 2.84 times the investment yields only 1.22 times the net benefit. This too is the result of a model calculation based on assumptions, but Scenario A at 1,255,000 THB pays back in approx. 6.8 years while Scenario B at 442,500 THB pays back in approx. 2.9 years. The 10-year ROI figures are approx. +10.7% and approx. +134.5%. A phased rollout that starts with shipping inspection comes out ahead.
4. If you buy in 2026, buy a 2D imager. GS1 Sunrise 2027 is an international initiative aiming for retail POS systems to be able to accept 2D barcodes by the end of December 2027, with preparation under way across 48 countries and territories representing roughly 88% of global GDP. 1D codes are not disappearing, but with a device service life of 5-7 years, a unit bought in 2026 will be in use until around 2031 to 2033.
5. The depa 200% deduction is limited to companies that qualify. Many Japanese-owned manufacturers in Thailand do not meet the conditions of 5 million THB or less in paid-up capital and 30 million THB or less in revenue. Check first whether your company is eligible, and if it is not, do not build the capital request on this incentive.
6. Structure it over 90 days. Measure in Day 0-15, set the target PPM and a one-page requirements definition in Day 16-30, assess standard package functionality in Day 31-50, compare quotations on identical conditions in Day 51-70, and run on-site testing and the investment decision in Day 71-90. Follow that order and your own numbers will be ready by the time you request quotations.

Note once more that all model calculations in this article rest on assumptions and are not the actual results of any specific company. Replace the figures with your own measured values.

Talk to Us

TOMAS TECH is a Bangkok-based integrator supporting Japanese-owned manufacturers with the implementation of production management systems and shop-floor systems. We are happy to talk at the review stage, before anything has been decided: “we want to calculate our PPM first”, “we want to assess whether the standard functions of our existing package are enough”, “we want quotations from several vendors lined up on the same conditions”. Tell us how your current process flows, and we will work through it with you from the angle of which process to tackle first for the fastest payback. Please get in touch via our contact form.

References

• Benchmarks for mis-shipment rate (PPM): https://stockcrew.co.jp/insights/article_kpi_miss_ship_rate_2026-05-04
• How to calculate and improve the mis-shipment rate (APT): https://n-apt.com/info/wrong-shipment-rate/
• Handheld terminal price ranges (Busicom): https://www.busicom.co.jp/misekatsu/all/device/handy_kakaku_b
• Criteria for selecting handheld terminals (LogiShift): https://logishift.net/glossary/%E3%83%8F%E3%83%B3%E3%83%87%E3%82%A3%E3%82%BF%E3%83%BC%E3%83%9F%E3%83%8A%E3%83%AB/
• GS1 Sunrise 2027 FAQ (Toshiba): https://business.toshiba.com/blog/sunrise-2027-faq-what-businesses-need-to-know
• What is GS1 Sunrise 2027? (GS1 US): https://www.gs1us.org/industries-and-insights/by-topic/sunrise-2027
• Minimum wage in Thailand (ThaiLawOnline): https://www.thailawonline.com/minimum-wage-in-thailand/
• Vietnam Decree 293/2025/ND-CP (Crowe Vietnam): https://brochure.crowevietnam.vn/en/news/decree-no-293-2025-nd-cp-increase-in-regional-minimum-wages/
• Thailand 200% tax deduction for SMEs (Mahanakorn Partners): https://mahanakornpartners.com/thailand-approves-new-tax-incentive-to-accelerate-sme-digital-transformation/