Bluetooth for Green Products - Time to broadcast the data

We see eco-friendly marks on products all the time. Ticks, leaves, letter grades, little globes with something growing out of them. It is almost strange now to find a product making no environmental claim at all.

It is a bit like trying to find a critic worth reading. Somewhere in the flood of influencers and casual reviewers there are a few people with proven expertise and judgement to lift the standard of the industry, rather than feeding it empty calories and clicks. Good luck telling them apart from the thumbnail.

In product sustainability, Type 1 ecolabels are those rare and serious critics. They look at products across the full life cycle and multiple criteria that matter, and they are independent. This type of environmental label is in the minority and covers well under 1% of what the world buys.

I have supported and helped set up schemes, sat on committees and contributed to some of the standards, so this is slightly uncomfortable to write.

How it was supposed to work

The original idea came out of West Germany in 1978. The federal interior ministry created The Blue Angel on the premise that instead of only banning bad products, you could help identify the best performers in a category, perhaps the top 20%, and make that visible at the point of purchase.

This then starts a self-reinforcing loop of recognition that creates trust and demand. More demand rewards further improvement, and with improvement in the category the bar can be raised. The mark could be trusted as the government had no commercial stake in who could be certified.

The Blue Angel is today recognised by 90% of Germans, and it can be found on 70,000 products from over 1,800 companies.

The concept has spread, and the Ecolabel Index now tracks 456 schemes across 199 countries. The schemes now range from Fairtrade to BIO Hotels and WindMade™. However, proper product ecolabels are the minority. The landscape is also still filling up. I recently read Maëlys Courtat's doctoral work, and she found new schemes in Europe arriving at more than one a month through 2021 and 2022. She concluded that none of them were fit for purpose.

It is tempting to solve a new problem with a new label, but we are now seeing the side effects of a flooded and uncoordinated market.

Why I don't see it working

If this self-reinforcing loop had worked, proper environmental certifications would be woven into a substantial share of the economy. Thresholds would be tightening regularly, as leaders would keep being caught up with, and consumers would reward and demand proper ecolabels.

Globally, a very small number of products carry a proper ecolabel. Type I only operates in a limited set of product categories and penetration within those categories is patchy. Full coverage isn't the goal, because Type I only certifies the best in a category, which puts the ceiling around 20 to 25%. There is no published figure of exact market coverage, but it's almost certainly less than 1% and not even close to critical mass.

And it is the global economy that must get greener, not just leading countries like Germany, Sweden and Switzerland. So the question is not whether ecolabelling works, which we can see it can, but why it has barely scratched the surface of what it could do.

Why Type 1 hasn't taken over

For the self-reinforcing loop to close, the proof must first reach the buyer at the moment of the decision, and the signal needs to be clear:

• A shopper faces a barrage of ticks and leaves and grades, most of them unfamiliar, and no way to tell which one can be trusted and proves true category leadership. I struggle when I come to a supermarket shelf where eight marks mean eight different things.
• A business buyer has to learn that not every mark is an ecolabel, and it's hard to see the difference from a self-declared claim, industry best practice or a single attribute. Also, not all ecolabels are the same, because two Type 1 schemes can still measure different impacts against different thresholds, especially across countries and regions.

The proof gets diluted at the shelf for consumers and devalued in the tender document for businesses.

What other industries did when they hit the same problem

Sustainability is a young discipline and some of this is teething. Older disciplines have hit comparable walls, and some got past them.

• No two shipping lines' boxes fitted each other. Pressure from the US military buying at Vietnam scale, converged the shipping container industry to ISO dimensions around 1968. Previously McLean's SeaLand ran 35-foot boxes, Matson ran 24-foot, and operators in Latin America wanted 17-foot for mountain roads.
• A bank card was only reliable where the shop banked with the issuer. Dee Hock turned a collapsing card scheme into a member-owned cooperative in 1970, which became Visa.
• Check-out scanners were useless until everyone printed the same symbol. In 1973 a committee met in a hotel near Grand Central to choose one barcode design from several competitors. It took only fifteen months before a packet of Wrigley's Juicy Fruit was scanned in Troy, Ohio.
• Wireless links were useless if they only spoke to one maker's devices. Ericsson, IBM, Intel, Nokia and Toshiba formed the Bluetooth Special Interest Group in 1998, named after a tenth-century king who unified warring clans, which tells you what they thought the problem was.

Six things they had in common:

1. A decision or transaction depended on the answer. No barcode, no checkout. No agreed container size, no military contract.
2. Every participant traded a large share of a small market for a small share of a large one.
3. The standard described what a thing was, rather than whether it was any good.
4. Nobody was allowed to own the standard, so it became neutral ground. Sometimes a dominant customer forced it, sometimes the participants did it themselves.
5. Getting there cost somebody a proprietary advantage. McLean handed over his patent on the container corner castings so the standard could exist at all. Ericsson could have kept Bluetooth to themselves.
6. Each of them took about four years.

Our problem is not absence of standards and cooperation. We have ISO 14024 for Type 1 ecolabels, 14025 for environmental product declarations and 14021 for self-declared claims, and the Global Ecolabelling Network members have been working on this since 1994.

ISO 14024 did what it set out to do, to govern a business-to-consumer trust mark, and leave the criteria and the underlying evidence to the programme operator. The standard is tried and tested, but it doesn't cover what we need to accelerate type 1 ecolabels today.

The biggest attempt so far

In 2005 Walmart launched a new business strategy built around three goals set by CEO Lee Scott: 100% renewable energy, zero waste, and selling products that sustain people and the environment.

The one buyer with enough leverage to move global supply chains examined 30 years of environmental labelling, and went and built something else.

They did not build a Walmart system, but partnered with Arizona State and the University of Arkansas to establish The Sustainability Consortium, with over a hundred corporate and civil society members. I worked on it, and we built the first global version of the Walmart Sustainability Index for the Chilean market.

They also tackled that sustainability is not one question the way a container dimension is one question. They went category by category to work out where the main impacts sit in a product's life cycle, then defined key performance indicators against those hotspots. Category dossiers, sustainability profiles, KPIs were developed using scientific methodology, product type by product type. The European Union is doing something similar now with its Product Environmental Footprint work.

The reach was remarkable and went beyond the membership. Several thousand companies used the tools, and by the Consortium's own account the system was informing performance management across more than $200 billion of annual consumer purchasing at Walmart, Kroger, Walgreens and others.

Two things it did not do.

1. Jon Johnson, who co-founded the Consortium, has since said Walmart "never used that information to make procurement decisions at any scale that had the effect we were hoping it would." I do not know why, and I would be speculating if I offered you a confident reason. What is clear is that the measurement existed and the decisions did not visibly change.
2. And it never closed the loop. The apparatus measured suppliers and never told them that doing well would mean selling more. There was no club to be in, no reward for beating the field, nothing that turned a good score into volume. A supplier who improved got a better number on a scorecard and the same purchase order. The proof never reached a shopper, so it created no consumer demand, and without demand or reward the self-reinforcing loop breaks.

Walmart built the system down to the level where it worked inside an organisation, but never handed that power to the people actually buying the products in their stores.

One sector has started

The major green building rating systems around the world now have a common way to describe product sustainability data, and they are reorienting their own tools around it. That agreement is the Common Materials Framework, curated by the non-profit mindful MATERIALS, which organises more than 100 building product certifications into one shared structure, without defining what constitutes good and bad. In 2025 they shut down their own technology platform so that it would not compete with the ecosystem they were aligning.

Forums in 2022, agreement in 2023, an implementation toolkit in 2025. This is barcode pace, with no regulator behind it. Nothing equivalent exists for supermarkets or general consumer goods.

You might say the European Union is already solving this with the digital product passport, and the intent is right, but the timeline has already slipped. If it lands it will be the biggest step forward in fifty years.

Your AI assistant is already learning to shop

Meanwhile, in February 2026 OpenAI turned on Instant Checkout in ChatGPT, so people in the United States can buy things without leaving the chat. It runs on the open Agentic Commerce Protocol, developed with Stripe, which lets ChatGPT ingest structured catalogue data and present products in context. Google announced its own coalition-backed protocol in January, heading into Search AI Mode and Gemini.

Next time you ask Claude or ChatGPT or Gemini to find you a pair of work boots it probably already knows where you live, roughly what you will pay, your size, and the brands you have liked before. From your conversations it may also have a fair idea of how much you care about global warming, labour conditions, animal welfare and sustainability in general.

It will then weigh whatever it can compare in terms of price, delivery date and reviews based on structured fields it can line up side by side. If environmental performance is not in the protocol, your preferences on it are invisible. Ultra-fast fashion is perfectly comparable on everything a machine can weigh and entirely opaque on everything that counts against it.

Imagine if your priorities could be factored into a purchase you were making anyway. You could get better choices without researching anything. And the manufacturers who did the work get a clear commercial signal that it was worth doing.

We risk the same disconnect as with Walmart, where the proof never reached the consumer, and the consumer never had the choice to reward the producer. We also can make agentic shopping work for sustainability rather than against it.

What I am asking for

A ubiquitous system where a shopper in an aisle, a procurement officer with a tender to award, or an AI agent buying on your behalf can tell which option is better on the things you care about, without anyone first having to research the certification scheme behind the mark.

Type 1 labels need to stand out and be universally recognisable. Every Bluetooth device on earth carries the same small rune, whatever brand is on the box. Sony, Apple, a fifteen-dollar speaker from a service station. Type 1 needs its rune to signal it is compatible and verified.

We have the data and the certification, but like Bluetooth, we need a shared structure underneath the labels, where schemes can keep their brands and their criteria, and to publish their outputs into one common and comparable format. The scope could be bounded to what can be agreed today and expand over time. The early versions of Bluetooth were unreliable and the interoperability was poor, which is worth remembering as we cannot wait until a standard is perfect.

I know it's asking a lot, because most of these programmes are not-for-profits or government bodies on modest budgets, and an organisation in that position cannot easily invest heavily or vote for a short-term reduction in revenue keeping it alive. But the answer has to be a bigger and connected market.

Do that, and the engine the Germans designed in 1978 properly connects. The leaders can prove what they have done, instantly, to a consumer, procurement system, a rating tool or a buying agent. The market will reward them with volume, which gives the schemes the resources to keep raising the bar. And once the top of the market has demonstrated that low-impact solutions are practical, safe and profitable at scale, regulators can raise the mandatory floor with some confidence, pulling everyone else up behind it.

That was always the design, but we got lost in the crowd and never built the protocol that lets the signal travel globally.

I am not arguing this solves everything. It is one step, and it feels like a necessary one to take real environmental labelling from the distant margins into the main conversation.

Someone will write the history of why we did or did not clean up our consumption and production. I would rather this was not the chapter about how the self-reinforcing loop was left open, while the people we built it for went about their day, never knowing there was a right choice to make.