Two Experts on Why Compliance in Fashion Is a Purchasing Decision

Last year, facilities using SLCP's converged assessment framework went through 10,700 social audits, covering 7.5 million garment workers across 122 countries. They found a great deal: 91 percent of the facilities assessed had at least one legal non-compliance, and the average assessment turned up ten. Wages and benefits accounted for the largest share, followed by health and safety, then working hours (SLCP Impact Report 2025). You could read that as a failure, but it is more useful as proof that fashion has become good at discovering what is wrong inside a factory. Of course, ten findings per assessment is only a good number if something happens to the findings afterwards. 

Daniele Seiffert, ESG Legislation and Policy Expert at Retraced, follows human rights and environmental legislation for a living at Retraced. She describes what a closed finding can look like on paper:

"It’s very easy that during an audit you say, oh yeah, we have three people in the room and then we have one fan. Then check mark. They are trying to address the heat stress. But then you try to stay in a room locked with a fan for ten hours at fifty degrees." 

While somebody took the effort to check whether heat stress was addressed, nobody checked whether the room got cooler. 

600 Suppliers and No Way to Visit Them 

Saqib Sohail spent his career on the other end of that file. For many years he was responsible for ESG, Social Sustainability & CSR at one of Pakistan’s biggest denim suppliers, Artistic Milliners, and now consults for producers with UpFront Textiles. Ask him how the industry ended up here and he will tell you that nobody set out to build a box-ticking system:

"It’s easier if you have one or two suppliers. But what if you have 600 suppliers? Then the brand cannot go and evaluate each and every one like that. And then the third parties came, and then all these certifications came." 

Certification solved a problem of scale, and it solved it well enough to become the default. No brand can visit 600 factories a season. Audits are how you find things out at that size, and the numbers above show they find plenty. Nobody built the second half, because only the first half had a market. Saqib is blunt about what that produced: 

"That check mark does not say improve your facility. It’s assuming that you are taking that certification, which most times is just the bare minimum that you can do to get it. Because not all of these certifications are so strict." 

The Price of a T-Shirt Has Not Moved 

Look again at what those audits find most often. Wages and benefits, at 40 percent of all non-compliances (SLCP). The single most common finding in fashion’s audit system is a pay problem, and pay is the one thing a factory cannot fix on its own. 

"On one hand, the cost of selling is not increasing, but the cost of making goods has been increasing over time. The wages are increasing as everyone demands that fair wages be paid. But as soon as you start paying fair wages and add them to the cost, the question is who will pay for it," says Saqib Sohail 

The same pattern shows up in energy. A brand asks a factory to move off cheap fuel, and the request arrives without money attached. Saqib has watched that request land in places where the alternative barely exists:

"Whatever is happening in the south of Pakistan is different from the central region. The central region has good access to biomass, so they can always opt for that for decarbonization. But if somebody says, hey, you are in the south of Pakistan, do the same thing because it’s working very well for the central, for me the cost will be much higher." 

Then there is the question of who collects the benefit when a supplier does invest. 

"All these organizations think, I’m doing this from my own pocket, but this benefit is mostly going to the retailer. They get to report good numbers to their regulators, and they might get some benefit from the customers or regulators. The supplier is not getting enough ROI out of it for them to keep investing in ESG." 

The imbalance starts earlier than that, in the ordinary business of placing an order. In the 2025 Better Buying Purchasing Practices Index lets factories rate the companies buying from them, anonymously, across seven areas of purchasing behavior. Its 2025 edition drew on 1,340 responses from garment and textile producers. What they most want brands to fix is planning and forecasting, named the top priority by 37 percent of them. A late or wrong forecast gets paid for on the factory floor, in overtime, idle machines or air freight. That cost arrives long before anyone asks about a corrective action. 

And the cost of proving compliance keeps climbing. Sapphire Group in Pakistan held four certifications fifteen years ago and now needs around forty, at roughly €130,000 a year in fees, with implementation costing three to five times that (Vogue Business). Over the same period the price of a t-shirt did not move. 

Nobody Believes the Regulator Anymore 

European law was going to settle this argument but then the editing started. 

"Sometimes they move too fast. The laws get adopted and made effective, and then along the way they realize that what they adopted is not effective, or not easy to do. So they soften it. And then you lose all the credibility as a regulator. This is why we are seeing so many brands that end up taking greater risks now. Okay, we will not go after compliance, let's just take the risk and see what happens. Or, why am I going to take all this effort now if I know that this regulation is going to afterwards be diluted?" Daniele Seiffert observes.

Daniele does not draw the obvious conclusion from her own observation. She wants better law rather than less of it, and the EU has spent the last few years building something closer to it:

"It’s more about the European Union taking the responsibility and putting a regulation across. It gives companies more time as well. They are announced, this is the regulation, we adopted it, it’s going to become effective in three years. This gives time for companies to start settling with that idea, preparing the internal systems. And it gives time to the regulators to understand the industries they are talking to, and how they will apply the text in a practical manner."

The EU Forced Labour Regulation is the working example. It applies from 14 December 2027, to every company and every product, whatever the industry, with no size threshold and no sectoral exemption. There is nothing to file. The Commission states plainly that the regulation introduces no audit or reporting obligations (European Commission). Enforcement runs through investigation instead. Once a case opens, a company has 30 to 60 working days to hand over what it already holds. There is no version of that timeline in which evidence gets assembled from scratch. 

Daniele has one more observation that cuts against the romance of the whole subject, and it is the reason she keeps arguing for legislation:

"It’s never, or hardly ever, going to be something where all brands just feel inspired. They do the work because they have to. But the hope is that because they have to, it gradually becomes implemented as a core part of the business."

Where she loses patience is with how shallow the law still reaches. The CSDDD caps supply chain visibility at the first tier, with substantiated knowledge of a violation as the trigger to look further. Most brands are still learning to map that far, a gap we have written about in more detail here. 

"We know that the biggest, the strongest, the hardest cases happen down the tiers. It’s not on your tier one. That one is always easy to cover," Daniele Seiffert says.

16 Percent of Factories Sit Above the Danger Line 

For the first time, SLCP’s 2025 assessment framework collected climate data, and the results describe the fan problem at industry scale. Eighty percent of facilities continuously monitor indoor temperatures. 16 percent hold temperatures at 31 degrees or above, which is the point at which heat stress becomes dangerous. 31 percent have a climate adaptation plan, and that figure splits sharply by size: 46 percent of large facilities against 15 percent of small and medium ones (SLCP Impact Report 2025). 

So the industry is measuring the heat. Four out of five factories know exactly how hot their floors are. Most of them have no plan for what to do about it, and the smallest suppliers, who have the least room in their margins, have the fewest plans of all. 

Saqib puts the missing question simply: 

"How do you improve the life of your worker? It’s not always about forced labor, it’s not always about child labor. There are things beyond that. How do you deal with your workers with heat stress? What can we provide them inside and outside the workspace to deal with heat stress?"

Daniele goes further, and connects the two halves of the category name:

"One stems from the other. If you’re a climate refugee you have to leave your country, and you’re going to take any job you can in the place you found, because you need it. So it also makes you more prone to accept unethical conditions of working, forced labor, getting barely paid, having no rest."

Heat displaces people, and displaced people accept worse work. 

The Plan Gets Written Without the People Who Run It 

Saqib is candid that compliance requirements carry commercial motives alongside ethical ones, and that trade advantage is part of the picture. His interest is in the part that survives that. "In there, in between, there is some good to be had. So we take that." What holds the good back, in his account, is distance. Rules get written by people who have not seen how the work happens. 

His example is traceability, and it lands close to home for anyone selling supply chain software:

"Let's just collect traceability data from tier 2 to tier 4. It should be easy, right? Assuming that everyone has the data. That's the first wrong assumption." 

The alternative he describes is not slower:

"If they would come to the supplier side and say, this is our intent, these are the outcomes we need to achieve, let's discuss and see how do we do this, the product then would be entirely different. The operational people on the shop floor, at the farm, at the ginning, would tell them that if you go this way it will be a lot easier for everyone."

The same gap shows up in how targets get set. A factory that decarbonized early is penalized by a late baseline, since every remaining point costs more than the first ones did. 

It shows up in enforcement too. Saqib supports the intent behind forced labor rules and doubts what a ban achieves alone. Ban a practice without changing the conditions that produce it, he argues, and it continues out of sight. He would rather see support and phased improvement come first, with the ban arriving at the end of that sequence. On child labor, he puts it as a question rather than a position: "You have to find out why he has to go and work." 

How Working Together Pays Off 

Ask Saqib what a brand should change first and he does not ask for more audits or fewer. He goes after the scoreboard:

"The first thing that comes to my mind is changing the way they measure outcomes. Outcomes should not be measured by check boxes."

Decide what you are trying to improve. Benchmark where a community stands today, then track livelihoods, health and conditions over time. And then attach it to the money: 

"Once you start to make those your success factors and link it with business, if a certain percentage of that is weighted against your business scoring, then automatically you’re giving an incentive to the suppliers and yourself to work on it," Saqib says.

A supplier is already scored on price, quality and delivery. Putting social and environmental improvement into that same scorecard, with weight behind it, changes what both sides gain from the work. It also stops compliance being a cost center that reports to nobody. 

There is evidence this pays. Facilities that share one converged assessment rather than repeating proprietary audits saved an estimated $35 million last year, sharing each assessment 2.7 times on average. 41 percent report fewer social audits since adopting the framework, 66 percent cut the staff time they spend on auditing, and 53 percent saved money (SLCP Impact Report 2025). 

Which answers a question Saqib asks and does not expect a good reply to: Is there any incentive for a factory to spend money on its people? 

Give a factory back the time and the money and most of them spend it exactly there. Of the facilities that saved through harmonization, 76.9 percent put it into workplace improvements, 70.1 percent into new programs or services for workers, 64.3 percent into technology and 41.2 percent into environmental certifications (SLCP). 

And the conditions improve. Facilities that have used the same framework for five consecutive years record around 20 percent fewer legal non-compliances than first-time users. Staying with a supplier long enough for the work to compound produces measurably better factories. Switching every season does not. 

That is the argument for the thing Saqib keeps returning to, which is being in the room at the beginning:

"The brands I’m quoting have helped in co-creating projects from day zero, where they would listen to the suppliers and say, these are our areas of work. What do you have in mind? When we discussed it from day zero, the success rate was much higher. The farmers benefited, and the money actually flowed back."

None of this scales through a compliance department of two people, which is where it currently sits at most brands. We have argued before that transparency works when it is built collaboratively, and Saqib takes that one step further into how a company is organized:

"Why should I always have a separate team working on ESG? Why can’t I have sourcing making those decisions based on the principles of ESG? Everyone in their own position should be well aware that this is important. Whenever I design, whenever I buy, whenever I sell, the product has to comply with all these things."

Which is where the two of them arrive at the same place from opposite ends of the chain. Daniele wants the law to reach further down the tiers and to stop being watered down before it takes effect Saqib wants the people who will implement a requirement to help design it. Both are describing a system where the finding is the start of the conversation. 

Ten Findings, and What Happens Next 

Only 41.5 percent of buyers currently accept a converged assessment (SLCP). The rest still send their own auditors to ask questions that have been answered. Every one of those visits costs a supplier money that could have gone into the floor and produces another set of findings for a file. 

The looking is done. Ten findings per SLCP assessment, 7.5 million workers, 122 countries, and four out of five factories already measuring their own temperature. Fashion knows what is wrong inside its supply chain in more detail than at any point in its history. 

Somebody checked whether the heat stress was addressed. The next decade belongs to whoever comes back and asks whether the room got cooler.