← 部落格← Blog

標準入門Standards

一物一碼：GTIN、批號、序號到底差在哪GTIN, lot, and serial: what is the difference

「一物一碼」常被當成一句口號，但要真正做對，得先分清楚你貼在產品上的那組碼，到底識別的是「一種產品」、「一批產品」，還是「一個產品」。這三件事是完全不同的層級，用錯層級，後面的追溯、召回與防偽都會跟著走樣。"One item, one code" sounds like a slogan, but getting it right starts with a simple question: does the code on your product identify a kind of product, a batch of products, or a single product? These are three different levels of identity. Choose the wrong one and everything downstream — traceability, recall, anti-counterfeiting — bends out of shape.

三個識別層級，先講清楚Three levels of identity, defined

在 GS1 的世界裡，產品識別並不是只有一種碼，而是由「精細程度」不同的幾個層級疊起來的。對製造商來說，最常用到的是以下三層，由粗到細：In the GS1 world, product identification is not a single code but a stack of levels, each more precise than the last. For manufacturers, three matter most, from broad to specific:

• GTIN（全球貿易品項識別代碼）——識別「一種產品」，也就是一個品項 / SKU。同一款、同一規格、同一包裝的商品，共用同一個 GTIN，就像型錄上的一個條目。GTIN (Global Trade Item Number) — identifies a kind of product, i.e. a product class or SKU. Every unit of the same product, same specification, same pack size shares one GTIN, like a single line in a catalogue.
• 批號 / 批次（Lot / Batch）——識別「一批產品」，也就是同一次生產所做出來的那一群。批號不會單獨存在，它是掛在 GTIN 底下，用來把同款產品再切成一批一批。Lot / batch — identifies one production run: the group of units made together. A lot number never stands alone; it sits under a GTIN and subdivides the same product into runs.
• 序號（Serial）——識別「一個產品」，也就是單一個體。同一個 GTIN 之下，每一件都有自己獨一無二的序號，真正做到「一物一碼」。Serial — identifies one individual unit. Under the same GTIN, every single piece carries its own unique serial — this is what truly delivers "one item, one code."

關鍵觀念：這三層是「包含」關係，不是互相取代。序號之下隱含批號，批號之下隱含 GTIN。你愈往細的層級走，能回答的問題就愈精準，但要管理的資料量與紀律也愈高。The key idea: these levels nest, they don't replace one another. A serial implies a lot; a lot implies a GTIN. The finer you go, the more precise the questions you can answer — but the more data and discipline you have to maintain.

用一個比喻記住差別An analogy that sticks

把它想成一本書。GTIN 像是這本書的書名與版本——市面上每一本《同一書名、同一版》指的都是同一個東西。批號 像是「第幾刷」——同一版可能印過好幾刷，哪一刷出了印刷瑕疵，你可以只回收那一刷，不必動到整個版。序號 則像是圖書館在每一本實體書上貼的那個唯一館藏編號——它指向的不是「一種書」或「一刷書」，而是「就是手上這一本」。Think of a book. The GTIN is the title and edition — every copy of the same title, same edition refers to the same thing. The lot is the print run — one edition may be printed several times, and if one run has a printing defect you can pull just that run, not the whole edition. The serial is like the unique accession number a library sticks on each physical copy — it points not to "a kind of book" or "a print run," but to "this exact copy in your hand."

製造現場也是同樣的邏輯：型號是 GTIN，當天那一鍋 / 那一批是批號，產線上每一件流下來的個體則對應序號。The shop floor works the same way: the model is the GTIN, today's batch or kettle is the lot, and each individual unit coming off the line maps to a serial.

每一層能讓你做什麼What each level lets you do

分層的意義，在於每一層各自解鎖不同的能力。值得注意的是，GS1 用「應用識別碼（Application Identifier，AI）」這套標準前綴來標示資料的意義——一般而言，(01) 代表 GTIN、(10) 代表批號、(21) 代表序號。同一組條碼或 QR Code 裡，靠這些前綴就能讓相容系統正確讀出哪一段是哪一層。The point of layering is that each level unlocks different capabilities. Worth knowing: GS1 uses Application Identifiers (AIs) — standard prefixes that tell a system what a piece of data means. In general, (01) marks the GTIN, (10) the lot, and (21) the serial. Within one barcode or QR code, these prefixes let a compliant system read out which segment is which level.

• GTIN——讓產品「被找到、被賣」。上架、報價、通路對接、電商上架、結帳掃描，靠的都是 GTIN。沒有乾淨一致的 GTIN，產品在通路與系統間就會「對不上號」。GTIN — makes a product findable and sellable. Listing, pricing, channel onboarding, e-commerce, checkout scanning all rely on the GTIN. Without a clean, consistent GTIN, a product simply won't reconcile across channels and systems.
• 批號——讓召回「精準縮小範圍」。當某一批原料或某一段製程出問題，批號讓你只圈出受影響的那一批，而不是全品項下架。對食品、保健品這類受監管的品類，批號往往是衛福部食藥署（TFDA）相關追溯與通報要求下的基本單位。Lot — lets a recall stay precise. When one raw-material batch or one stretch of the process goes wrong, the lot lets you isolate just the affected run instead of pulling the entire SKU. For regulated categories such as food and supplements, the lot is often the basic unit of traceability and reporting under the requirements overseen by the TFDA.
• 序號——讓你掌握「單一個體的真偽與履歷」。每件獨一無二，才談得上防偽驗證、保固綁定、灰色市場追查，以及精確到「這一件」的流向與歷程。Serial — gives you authenticity and history at the unit level. Only when each piece is unique can you do authenticity checks, warranty binding, grey-market tracing, and a movement history precise down to "this one unit."

那我到底該做到哪一層？So which level do I actually need?

不是愈細愈好——層級愈細，編碼、列印、資料維護的成本與紀律都跟著上升。實務上可以這樣判斷：Finer is not automatically better — each added level raises the cost and discipline of encoding, marking, and data upkeep. In practice, judge it like this:

一個常見且穩健的起點，是先把 GTIN 與批號做扎實——這已能支撐合規追溯與大多數召回情境；等到防偽或單件履歷成為明確需求時，再往序號層擴展。重點是讓現場列印、資料系統與標準三者對齊，而不是一次把所有層級都塞上去。A common, solid starting point is to get GTIN and lot right first — that already supports compliant traceability and most recall scenarios — then extend to serial once authenticity or per-unit history becomes a clear requirement. The goal is to keep your marking on the line, your data systems, and the standard aligned, rather than forcing every level on at once.

無論你停在哪一層，一個好的追溯平台都應該做到：忠實依循 GS1 的識別與 AI 規範、能隨業務需要在層級間順暢擴充、並且讓現場人員「掃一下就懂」，而不是把標準的複雜度丟回給操作端。這正是 Animfy 想為台灣製造業解決的事。Wherever you stop, a good traceability platform should do three things: follow the GS1 identification and AI conventions faithfully, scale smoothly between levels as your business grows, and let people on the floor scan and immediately understand instead of pushing the standard's complexity back onto operators. That is exactly what Animfy sets out to solve for Taiwan's manufacturers.