2D codes on packaging and the new SEO: product visibility in the AI era

When a customer asks a search engine or an AI assistant about a specific product, they are no longer looking only for the "best brand". They want the details and specifications of that particular model that matter to them.
The answer to such a question is assembled from information attached to one specific product. What counts is its identifier, variant, composition, specifications, certificates, instructions, availability, and other data that can be connected and verified.
A 2D code on packaging creates a direct connection between the physical product and that information online. Every pack becomes an entry point to data about one specific product. The customer can browse it while shopping in a store, and the same product page increases the product's visibility in search engines and in answers generated by AI models.
Why product-level SEO and GEO matter
Traditional SEO still matters. Domain authority, site quality, links, and category content help a search engine assess how trustworthy a brand is. They are not enough, however, to answer a precise question about a single model, variant, or production batch.
This is where GEO (Generative Engine Optimization) comes in: preparing information so that systems generating answers can find it, interpret it correctly, and connect it to the right product. GEO does not replace SEO. It extends it with a readable data structure, consistency across sources, and answers to detailed user questions.
Product SEO | Product GEO |
|---|---|
| Helps the product page appear in search results | Helps an AI system understand and use the product information |
Relies on content, links, structured data, and the technical quality of the page, among other factors | Relies on unambiguous facts, relationships between data, sources, and consistency of information |
| Directs the user to a page | Increases the chance that the product is described correctly or referenced in a generated answer |
No markup guarantees a citation by a language model. You can, however, remove the biggest obstacles: publish reliable and complete data, and update it consistently across all channels.
The 2D code as a digital entry point to the product
Packaging is one of the few channels that stays physically attached to the product throughout its sale and use. A 2D code can use that space for more than opening a campaign page. It can lead to current information about exactly the product the customer is holding.
A QR Code or Data Matrix is a data carrier. Its value is decided by what it contains:
- a plain URL leads to a single web page,
- GS1 Digital Link can serve different resources, for example a product page, instructions, recycling information, an e-label, or a digital product passport.
A 2D code is not a ranking factor in itself. It does, however, create a durable bridge between the product, the user, and structured data. Every scan can lead to a page that you can expand, localize, and update without reprinting the packaging.
The manufacturer should be the best source of information
A marketplace can describe delivery and purchase terms very well. A reviewer can assess the product in practice. It is the manufacturer, though, who knows the exact composition, specification, instructions, certificates, and change history of the product.
If a brand does not publish this data in a readable form, retailers, catalogs, and users fill the gap. Discrepancies appear: a different variant name, outdated specifications, a missing warning, or a marketing promise presented as a confirmed fact. The search engine, the AI assistant, and the customer all have to decide which version to trust.
A manufacturer does not have to control every opinion online. It should, however, provide a canonical set of facts that other parties can refer to. Where needed, those facts should also be backed by evidence in the form of documents, test results, or certificates.
Which data builds visibility for a specific product?
First you need to answer one question: can a human and a machine tell without any doubt which product is being described?
A basic profile should cover:
- identity: brand, full name, GTIN, model, variant, and images,
- characteristics: dimensions, materials, composition, compatibility, intended use, and limitations,
- life cycle: origin, instructions, repair, spare parts, reuse, and recycling,
- evidence: certificates, declarations of conformity, reports, and their validity dates,
- market context: language, market, currency, availability, and local instructions and warnings,
- relationships: accessories, parts, successor models, and other compatible products.
The data should use the same identifiers and meanings on the brand site, in structured data, in product feeds, and in materials handed to retailers. Search engines rely on markup such as Product on Schema.org, and Google documents its requirements for product structured data separately. Such markup helps a page be understood, but it will not fix contradictions in the source data.
Data distribution: one truth, many channels
Visibility is not built by copying identical text into dozens of catalogs. It is built by consistently sharing the same facts in the places where customers and systems actually look for information.
In practice this means being present in several layers:
- The 2D code on packaging: a direct step from the physical product to its data.
- The manufacturer's site: the most complete and up to date product page.
- Stores and marketplaces: commercial data consistent with the manufacturer's information.
- Comparison sites and industry catalogs: correct identifiers and specifications that make comparison easier.
- Expert material and user sources: instructions, tests, reviews, questions, and experiences.
- Local versions: language, units, warnings, availability, and requirements specific to the market.
Each channel plays a different role, but describes the same product. A stable GTIN, model name, and resource address make it possible to connect these mentions instead of creating several seemingly different products.
How to build product GEO and SEO step by step
Step | What to do | How Thing Identity helps |
|---|---|---|
Establish the product identity | Assign a GTIN, full name, model, and variant. Define the relationships between products and how production batches are identified. | On the platform, every product site is created for a specific GTIN. A batch number can be added to the GS1 Digital Link so that it leads to the right version of the data for that batch. |
Create a data profile | Collect confirmed information in one system. Assign data owners and rules for updates. | The site assigned to a GTIN serves all partners as a single source of truth. You update the data and language versions on the platform. |
Publish the product page | Answer your customers' questions. Add a description, specifications, instructions, and documents. | On the platform you build and publish a product site with dedicated pages, content, and language versions. |
Add structured data | Describe the product in JSON-LD. Include the name, brand, GTIN, variant, manufacturer, image, and description, among other fields. | Thing Identity generates JSON-LD automatically from the published product data. |
Connect the product with the packaging | Put the GS1 Digital Link in the 2D code. Before printing, check its size, contrast, quiet zone, and readability on the finished packaging. | A code generated on the platform leads to the stable Thing Identity resolver and to the site assigned to the right GTIN. |
Update and monitor | Keep the data consistent across all channels. Monitor product visibility, code scans, and the pages people visit. | You update the content without changing the code on the packaging, and Thing Identity analytics shows the traffic generated by 2D codes. |
How to measure the results of product SEO and GEO
To measure GEO results, use dedicated monitoring tools such as Chatbeat. Chatbeat regularly checks how a brand and its products appear in answers from ChatGPT, Gemini, and Perplexity. It shows visibility, position, sentiment, the sources used by the models, and competitor results.
Set up a repeatable set of questions about the product name, model, category, and most important use cases. This lets you see whether the product's visibility is growing, how AI describes it, and which sources it draws information from. You can compare Chatbeat results with scan analytics in Thing Identity to connect the product's presence in AI answers with the traffic generated by the 2D code on the packaging.
The most common mistakes
Mistake | Why it is a problem | How Thing Identity solves it |
|---|---|---|
One generic page for a whole product family | The assistant then does not know which specifications belong to which variant. | Every GTIN can have its own product site, and variants can have dedicated pages. Specifications stay attached to the right product. |
A code that leads only to the home page | After scanning, the user should immediately get information about the right product. | In Thing Identity every product site is created for a specific GTIN, so the code leads to information assigned to the right product. |
A code used as a one off campaign | An address printed on packaging should stay stable longer than an ad or a promotion. | The GS1 Digital Link address stays on the packaging, while the published content and target resources can be updated on the platform without reprinting the code. |
Different data in every channel | Contradictory names, composition, or dimensions make the product harder to recognize and weaken trust. | One managed product site holds the canonical content and its language versions, so it serves all partners as a single source of truth about the product. |
Marketing without evidence | Terms such as "eco friendly" or "completely safe" need a precise meaning and credible confirmation. | On the product page you can publish specific parameters and the documents that confirm them, so that important claims have an accessible source. |
Publishing without maintenance | A product page loses its value if it has no owner and no update process. | The platform lets you prepare a draft, check the preview, and publish updates. Version history and analytics help you maintain and develop the content. |
Summary
In the era of conversational search the brand still matters, but the answer is formed at the level of a specific question and a specific product. The products that win are the ones that can be identified unambiguously, understood well, and verified against consistent sources.
A 2D code on packaging with a GS1 Digital Link creates a physical entry point for this. It connects the product with its identifier, its page, and structured data. The manufacturer supplies the canonical facts, partners distribute them in their own channels, and the customer gets current information after searching for the product or scanning the code.
Read next
- GS1 QR Code, GS1 DataMatrix, or GS1 Digital Link: Which 2D Code Should You Choose?A practical comparison of data encoding, smartphone support, point-of-sale use, and the size of four 2D codes used in the GS1 system.
- Where to print 2D barcodes on packagingHow to position a GS1 Digital Link QR Code or Data Matrix relative to an EAN barcode, preserve quiet zones, and design packaging with POS and consumer codes.
- GS1 Digital LinkGS1 Digital Link - the code that's reshaping product manufacturing.
- E-LabelE-Label - wine labelling is going digital.
- Digital Product PassportWhat is a digital product passport?
