Most operators can tell you how many cups they bought. Far fewer can tell you what happened to them afterwards.

That isn’t carelessness. Plenty of reusable cup schemes run perfectly well on counting, so many cups out, so many back, and if the gap looks wrong somebody goes and has a look. At a small scale it’s workable, and it’s where most schemes start.

But counting only ever tells you how many. It never tells you which. And “which” is what nearly every question that matters later depends on: did this customer return the cup they were charged for, where are the losses actually happening, have these cups been used enough times to have been worth buying in the first place.

Why a cup needs an identity of its own

Give every cup a unique digital identity and you can follow individual cups through their whole working life: sold, returned, collected, washed, put back into circulation. I’ve written before about what actually happens to a cup after you return it, and the identity is what makes that chain visible rather than assumed.

It matters most for digital deposits. When a customer pays a deposit, or we place a temporary hold on their card, the system has to know which cup they walked away with. When that specific cup comes back, the refund or the release follows automatically. Without a unique identifier you’re relying on someone’s judgement at a counter, and at any real volume that falls apart.

Two technologies do this job in practice: QR codes and RFID. Neither one is the right answer everywhere, which I’ll come back to.

QR codes: the cheapest way to give a cup a name

A unique QR code printed on each cup, read by a phone or a camera-based reader. No tags, no antennas, very little to install. For a smaller reusable cup scheme with a manageable number of cups and a couple of return points, it’s a sensible place to start, and we do use QR codes in simpler deployments.

It’s worth being clear about what a QR code can do, because it’s usually undersold. The identity it gives a cup is every bit as unique as an RFID tag’s, and in principle you can record exactly the same events against it: sold, returned, collected, washed, back into stock. There is nothing a well-built QR scheme can’t know about its cups.

The limitation is visibility. A camera has to be able to see the code. In some systems that means one cup at a time, and in others it doesn’t, depending on how the cups are designed and what reading technology is in place. Reading several at once is possible. It’s just harder, and it puts constraints on the cup, on how cups are stacked and on the process around them.

RFID asks for none of that. A tagged cup can sit at the bottom of a crate, under a hundred others, and still be read.

A venue we’ve been analysing hires around 50,000 cups for a single event. You can build a QR-based process at that scale. You’ll just spend the effort somewhere: in the cup design, in how rigidly cups have to be handled, in staff time, or in all three.

So both can get you to the same place. One route takes a good deal more engineering than the other.

The job QR codes are genuinely good at

There’s a second use for QR codes that has nothing to do with tracking, and I think it’s underrated.

A code on the cup can take a customer to a page explaining how the scheme works, where the nearest return points are, and how long they have before a deposit or payment hold expires. No app download, no account. Just a web page.

I’m firmly against asking customers to download anything. Reuse has to be as easy as single-use, and an app is a wall between a person and their coffee. A QR code that answers a question in five seconds is the opposite of that.

RFID: reading a crate instead of a cup

RFID gives each cup a small electronic tag with a unique identity. Same idea, different physics: nothing has to see the cup.

Multiple cups can be read at once, including inside collection bags, crates and bins. Readers can be placed at fixed points: a bar, a back-of-house door, the entrance to a washing facility. Cups are picked up as they pass, with nobody handling them individually.

That changes the economics of the whole operation. Instead of a person scanning cups one by one, a crate goes past a reader and the system knows exactly what was in it. This is the technology behind what we call Invisible: the customer drops a cup into a return point, the cup identifies itself, and the refund follows without a code to scan or a machine to queue for.

It’s also worth saying plainly: across our RFID deployments we haven’t had return fraud. The cup either is the cup it claims to be, or the system doesn’t accept it.

The tags are the easy part. The complexity sits in the software behind them: giving every cup a unique identity, recording each event in its working life as it’s sold, returned, collected, washed and put back into stock, and keeping that record accurate across thousands of cups moving at once. Done properly, the system knows which cup is where and what stage it’s at, at any moment. That’s the hard engineering, and we’ve spent years refining ours on live schemes.

It’s part of why not every provider offers RFID. Some have looked at it and decided it isn’t right for the customers they serve, which is a perfectly reasonable position. Others don’t offer it because they haven’t been able to build it. Those are two very different answers to the same question, and it’s worth knowing which one you’re being given.

What that visibility is actually worth

Faster scanning is the obvious benefit. The operational picture is the valuable one.

Once every cup is identified at every stage, you stop guessing about things you’re currently taking on trust.

Losses come first. Every scheme loses cups. Very few can say where. With tracked cups you can see whether they’re walking out of the venue, going missing between the bar and the collection point, or not coming back from the washing operator, and it’s rarely where people assume. That last one is especially worth knowing about, because once cups leave the site the trail normally goes cold. “We sent out four thousand and got three and a half thousand back” is an argument nobody can win. If every crate is read on its way out and on its way back, there’s nothing left to argue about.

The same record lets you check what you’re being charged for. If you hire cups, you’re paying for the number you take and usually for the ones that don’t come back, and until now you’ve had to take those figures on trust, because the only record of them belonged to your provider. Now you have one of your own. I’d encourage any venue to use it, ours included. A provider confident in their numbers won’t mind being checked.

Washing is next. A cup can be scanned as it’s washed and that wash recorded against its identity, so you can tell the difference between a cup that has simply come back and one that’s ready to be served again. For anyone carrying responsibility for food safety, that isn’t a reporting nicety.

Then there’s stock, during the event rather than after it. You can see which bars are holding cups and which are running low, in time to move some across, rather than finding out because a bar has run dry at half-time. And before the next event, you know how many clean cups you actually have rather than estimating.

For a visitor attraction working through roughly 1,400 cups a day, that’s useful. For a stadium turning over tens of thousands in a few hours, it’s the difference between a scheme you can run and one you’re constantly firefighting.

Proof, not estimates

Most reuse reporting I’ve seen is modelled. Someone takes the number of cups ordered, multiplies by an assumed number of uses, applies a factor and presents the result as an outcome. It’s an estimate wearing the clothes of a measurement, and anyone reading it closely can take it apart.

Per-cup identity gives you something better. You’re not estimating how many times your cups were used. You’re counting actual uses of actual cups.

That matters because a reusable cup starts out worse than a paper one. It takes more to make, and it has to be washed and moved for its whole working life. It only beats single-use once it’s been used enough times to pay that back. In some schemes that break-even point is as low as four uses, depending on the cup, how it’s washed and what you’re comparing it against. That’s genuinely encouraging, because four is not a demanding target for a cup that stays in circulation.

The problem is that most schemes are guessing at whether they’ve crossed it.

With tracked cups you don’t have to guess. You can see how many times each cup has been used, when it passed its break-even point, and what the fleet looks like as a whole. If a batch is underperforming, you’ll see it in the data rather than discovering it at the end of the year.

Which means when a board, a tender panel, a sustainability report or a local authority contract asks what your scheme achieved, you can hand over a number built from what actually happened. I’d rather give someone a figure they can challenge and have it hold.

What it costs, and which one is right for you

QR codes cost very little in hardware. The cost shows up in the process: designing the cups and the handling around a camera needing to see the code, and the staff time that goes with it. RFID costs more upfront, a tag on every cup plus readers, antennas, connectivity and integration work, and then takes that handling cost out of the operation permanently.

The crossover point isn’t a number I’d give you without looking at your venue, because it depends on how many cups you’re moving, how fast, and who’s currently touching them. That’s the same logic behind choosing between Two-Tap, Integrated and Invisible: start with the operation, then work backwards to the technology. If you want a rough sense of the scale you’re dealing with first, our reusable cup calculator is a reasonable starting point.

Most schemes should use both

This isn’t really a contest.

In most of the schemes we run, RFID does the identification and tracking behind the scenes, where speed and bulk reading matter, and QR codes face the customer, where a human needs an answer to a question. Each one does the job it’s good at.

There is no one-size-fits-all in reuse. I’ve said that about payment models and it’s just as true about tracking. A café and a stadium are not the same operation, and anything sold to you as the answer for both is being sold rather than recommended.

So be wary of anyone who only ever arrives at one conclusion. If a provider can offer a single technology, that technology will turn out to be what you need, every time. Ask them what they would do if your venue needed the other thing. The answer tells you whether you’re getting advice or a catalogue.

What matters more than how a cup gets identified is whether that identity does any work for you afterwards: whether it tells you where your stock is, where it’s leaking, and whether the scheme you’re paying for is doing what you said it would.

If it doesn’t, all you’ve really bought is a scanner.