splide

You print once. We answer for years.

The striped barcode on your packaging is being replaced worldwide by a QR-style code that a till and a phone can both read.

That code is not a link to a page. It is an address that can be asked a question, and the answer can change every day for the rest of the product's life without the pack ever being reprinted. Someone has to run that. It is what Splide does.

Talk to us about your packaging

What is actually changing

For fifty years your barcode has been a number. A till reads it, looks up a price, and that is the end of it.

The code replacing it is a web address that still carries the number inside it. Tills keep working exactly as before. The difference is that a customer holding your product, or a retailer's system, or a customs officer, can now ask that code a question and get an answer from you.

Tesco replaced the barcode at its checkouts in April 2026. Walmart is pushing supplier categories to follow. GS1, the organisation that issues barcode numbers, has set the end of 2027 as the point by which retailers should be able to accept the new codes.

Two things you will hear overstated, and one that is usually left out entirely:

In South Africa, nothing is illegal and nothing expires. Your existing barcodes keep working. There is no fine for doing nothing. Anyone selling you urgency on that basis is selling you something.

The deadline that matters is not GS1's. It is the morning a retailer you supply sends a letter with a date on it. That is when this stops being a good idea and becomes a project with a deadline, a scramble and a cost. Being ready before that letter arrives is cheap. Being ready after it arrives is not.

Unless you export to Europe, in which case there is a real law with real dates. The EU's Digital Product Passport requires a product to carry a machine-readable link to its own compliance, materials and end-of-life information. Batteries are first, in February 2027, with textiles, electronics and tyres following. The passport has to be reachable from the product itself, and this is the mechanism it is carried on. If you sell into Europe in one of those categories, this is not an advantage over your competitors. It is a condition of being on the shelf at all.

One code. As many answers as you need.

This is the part that gets missed, and it is the whole point.

The same printed code answers differently depending on four things: what is being asked for, who is asking, when they ask, and how precisely the code identifies the thing in their hand. Nothing on the pack changes. You print once.

It can be asked for something specific

A pack does not have one destination. It has a set of them, and each one has a standard name that any system in the world can ask for: ingredients, allergens, nutrition, instructions, safety data, certification, warranty, sustainability, recycling, the manual, the recipe, where to buy it.

So the same code on the same jar gives a customer at the stove the recipe, and gives a retailer's onboarding system the allergen data in a format it can read, without either of them knowing the other exists.

The answer knows who is asking

Language and country, from the phone itself. A pack sold in three markets carries one printed code and answers in three languages, with the ingredient declaration each market legally requires. You stop printing market-specific packaging to solve a problem that is no longer a printing problem.

The answer knows when

A code can answer one way before the expiry date and another way after it. A discontinued product can route to its replacement instead of a dead page. A seasonal promotion can end without leaving a broken link on stock that is still in the field two years later.

Three levels of precision, on the same pack

This is the one people underestimate, and it is where the value stops being about marketing.

The product. This is a 250g jar of chilli sauce. The same answer for every jar ever made.

The batch. This is batch AB2143, made on a date, with a certificate of analysis, an expiry and a recall status of its own. Batch-level answers are how quality, compliance and food safety stop being a spreadsheet somebody emails around.

The individual unit. This exact jar. This exact fork. This exact battery. Once units are serialised, the code stops describing a product and starts identifying a thing, which makes four hard problems tractable at once: warranty registered by scanning instead of filling in a form, a service history attached to the item rather than the model, ownership that can transfer when it is resold, and authentication, because a unit that has been scanned in four countries this month is not four units.

You do not have to set all three up. If someone scans a code more specific than anything you have configured, they still get an answer: it falls back to the batch, and then to the product, rather than failing. You can start at product level and add depth later, on packaging that is already printed and already in the field.

It is not only retail packaging

Retail products are one of fifteen kinds of thing this system can identify. The same infrastructure addresses shipping cases and pallets, physical locations, equipment and returnable assets, product models, individual components and spare parts, and service relationships.

That matters if your problem is not a supermarket shelf. A code on a shipping carton that your dealers scan into your own system. A code on a machine in the field that opens its service record. A code on a spare part that proves it is the right part and that it is genuine.

Your product data gets one address

Every retailer, marketplace and distributor you supply wants your product information in their format, on their portal, on their schedule. You fill in spreadsheets. Every one of them starts drifting out of date the moment something changes, and you find out when a listing is wrong.

A code that answers questions changes the shape of that problem. Anyone who needs your product data asks the code, in a standard format, and gets back every link the product has, current as of that second.

You maintain one set of answers. Everyone who needs them asks the same address. Nobody has to be told when something changes, because there is nothing cached on their side to go stale.

What makes someone scan twice

Utility beats marketing, every time, and it is not close.

Scan-to-win gets scanned once and never again. A washing instruction, a torque spec, a dosage, a fitting guide or a recipe gets scanned at the moment of need, by a different person, for years.

The person holding a part they cannot fit is already committed and already frustrated. Answer that, and the code has earned a permanent place on the pack.

Why this is a service, not a project

Packaging is printed months before it ships and sits in the world for years after. Once a code is on a pack, it is permanent. You cannot recall it to change a link.

So the address on that pack has to keep answering long after the campaign ended, the website was redesigned, the agency was changed and the marketing manager moved on. When one of those things breaks a link, the pack in a customer's hand is broken, and you find out from the customer.

That is the job Splide takes: run the answering service behind your codes, and keep it correct while everything around it changes.

Printers put the code on the pack. We run everything it points to.

What you get to see

Most vendors are vague here. We would rather be exact, including about the parts that are not ours to give you.

The till never calls us. Checkout scanning reads the number locally, by design, and no request reaches us. Retail sell-through data is the retailer's own product and they sell it. Anyone promising you shopper data out of this is describing something that does not happen.

Phone scans are a sample, not a census. Only some customers scan. What survives that kind of sparse data is not tracking, it is anomalies, and those are worth real money:

Inside your own channel, the picture is complete. Where your own people do the scanning, your dealers, your reps, your own stores, the scanning app is ours and every scan is real data. That is a full view of your channel, honestly obtained.

When a batch goes wrong

You will not think about this until the day you need it.

If a batch has to be pulled, one switch makes every unit of that batch, anywhere in the world, answer with your safety notice instead of the usual page. Not the product line. That batch. The units beside it on the same shelf carry on as normal, because the code can tell them apart.

The notice is written and sitting ready before anything happens, because nobody should be drafting a recall page under pressure.

How you would start

Start here. Your codes point at our infrastructure and you manage what they answer with. Nothing changes on your side except what gets printed.

On your own domain. The codes carry your web address and point at our infrastructure behind it. This matters more than it sounds. A printed code is permanent, so if the codes carry our name you are tied to us forever. On your own domain you can walk away and take your packaging with you. The fee buys operations, never permission.

Wired to your catalogue. We take your product feed, generate the links and keep the destinations correct as your website changes. You stop maintaining a list of links entirely.

No pricing here yet. It depends on how many products, how deep you go and which of the three above, and we would rather scope it against your actual range.

Talk to us about your packaging

Is this real, or a pitch?

It is running. You can check it yourself right now, without talking to anyone:

That is a demonstration brand on numbers reserved for demonstrations, running on the same service a real one would use.

The technical detail, the standard it conforms to and the conformance test result are on the technical page.