Quality and integrity: What developers are really thinking about? (Part II)

This article is a part of a series. Here is part 1.

A few days before Christmas, I published Part I of this short series on the words “quality” and “integrity” in the carbon market. I looked at how these terms circulate on the demand side among buyers, standards, rating agencies and intermediaries and how they influence pricing, sentiment and reputational risk.

I received a lot of comments and feedback from different market players, and I realised that the article was very “market-side”. The missing voice was obvious: project developers, the people who actually build and run the projects behind the credits.

So for Part II, I reached out to three developers that I’ve known for many years and whose work I have always admired: Jo Andersen, Co-Founder of Carbon Tanzania; Alessandro Galimberti, Carbon Expert at AVSI Foundation; and Nicola Baggio, Founder of Offgrid Sun. Three different worlds: forest conservation projects in East Africa, clean cooking in African cities and communities, and a mix of off-grid renewables, stoves and water access. What immediately came to my attention during our conversation was how surprisingly aligned their views are on what “quality” and “integrity” mean in practice.

The first thing that becomes clear is that developers and buyers simply do not mean the same thing when they use these words. For buyers, “quality” often means methodologies, baselines, additionality, leakage, ratings and due diligence. It’s a language of accounting. For developers, quality usually starts somewhere else. When Jo talks about quality in conservation, he doesn’t begin with buffer pools or methodology updates. He starts with the land and the relationships: Who owns this forest? How are communities represented? Who benefits from revenue? Will the governance survive political change? As he told me, “Conservation projects start with social contracts, that’s where integrity begins.” Carbon standards and methodologies are considered an important add-on at a later stage. Alessandro sees it in a similar way on the clean cooking side. For him, a good program is one where stoves are actually used, households understand them, and real health and time benefits appear. A project can pass every methodological requirement and still fail on the ground if stoves gather dust in the corner. I first met Alessandro in Mozambique 13 years ago, at the beginning of what became a very successful program with hundreds of thousands of beneficiaries, and the level of care and detail has not changed. Nicola, coming from two decades of trying to make solar work in Africa before entering carbon, uses a more procedural definition but the same spirit: quality means rigour without shortcuts, applying rules consistently, involving national authorities and resisting the temptation to tweak assumptions just to increase credit volume.

In carbon markets, outputs are things we can count: tonnes of CO₂, leakage adjustments, buffer contributions, verification reports, rating scores. Outcomes are things we can observe: forests that remain standing, households that adopt stoves, women who save time, communities that gain income and agency. Markets optimise for outputs because they are tradeable. Developers optimise for outcomes because they determine whether projects actually work in the real world.

A second point that emerged quickly is that projects on the ground do not change as fast as methodologies. All three developers described versions of the same situation: the forest is the same, the stove is the same, the users are the same, the community dynamics are the same, yet the number of credits changes dramatically because a parameter changed or a new methodology came into force. The underlying asset is the same, the impact is the same, but the formula is different. From the outside, this looks like scientific and methodological progress. From the inside, it can feel like the ground moving beneath your feet. One year, a project is considered “good quality” under an older methodology; the next year, the exact same project looks “lower quality” because a new, more conservative methodology exists, even though nothing has changed in the real world. This raises a simple but important question: are we measuring the quality of implementation, or the precision of accounting? Right now, most of the pressure sits on the accounting side because that’s where integrity concerns and reputational risk live.

This would be an academic debate if carbon markets were only about measurement, but they are mainly about finance. And finance determines whether projects exist or disappear. This is brutally clear in clean cooking and energy access. For years, traditional climate and development finance under-funded these sectors. Carbon finance became, for many developers, the only scalable funding mechanism. When methodologies or parameters reduce credit volumes, they don’t just change climate models; they can kill the investment case entirely. One example that came up involved a change in a single parameter (FNRB), which pushed the cost of producing one carbon credit from around USD 10 to above USD 60. The stove didn’t change. The usage didn’t change. The social impacts didn’t change. Only the formula did. At those economics, the project simply would not happen. Nicola, who knows from experience how difficult it is to finance energy access in Africa, put it plainly: carbon finance is the only predictable mechanism he has seen for access-to-energy projects in Africa. If the incentive to build disappears, nothing replaces it.

An interesting dynamic that all market players have noticed over the last two years is the conservative methodology review cycle across all the most used methodologies. More conservative parameters reduce credit volumes and partly for scientific accuracy, partly (in my view) to protect standards from criticism after the scandals we’ve seen. This is understandable and, honestly, necessary. But it only works if markets recognise the value of those more conservative credits. Right now, I don’t see that happening. Developers talk about future Article 6 and CORSIA markets with prices above EUR 20, but we are not there yet, and the voluntary market alone cannot carry the burden. And in the Voluntary Carbon Market, price is still the main driver when a company needs to buy credits.

By coincidence, while I was organising my notes from these conversations, a friend forwarded me a recent Financial Times article on Koko Networks, once considered one of Kenya’s most promising startups. Beyond the specific case, which I don’t claim to know in all its operational details, what struck me is how clearly it illustrates the same point raised by the developers when discussing finance. Koko was a large-scale, data-rich clean cooking project, serving more than a million households and built on the assumption that carbon revenues could function as a long-term energy subsidy. Its business model, however, depended on access to higher-value compliance demand and on the issuance of a Letter of Authorisation by the host government. Decisions around authorisation are, understandably, linked to national policy priorities and NDC considerations, and governments have legitimate reasons to proceed cautiously. At the same time, when such decisions determine whether credits can access different markets, the financial exposure for projects becomes significant. In practice, the very same credit, identical in terms of technology, users and impact, can be valued several times higher with an LoA than in the voluntary market. While this price differential may be rational from a regulatory perspective, from a project-finance standpoint, it concentrates a large share of risk on developers. If this risk is not explicitly managed or shared, it can undermine project viability and, in some cases, lead to business closures.

Another dimension of the quality and integrity discussion with my three guests concerns the role of digital MRV and the expectations increasingly placed on data and monitoring technologies. The market often turns to digital MRV as a potential bridge between developers and buyers: better data, better monitoring, and greater comfort for demand. Developers welcome the tools because they can simplify data collection, think IoT in clean cooking and remote sensing models in NBS, but are realistic about their limits. Digital data still flows into human assumptions: baselines, sensor gaps, mixed fuels, user behaviour, projection of emissions over time. That last layer of judgment still drives credit outcomes. And there is also the issue of opportunity cost. If a project was designed well from the beginning, “quality” does not necessarily translate into more costs; this was a consistent point among Alessandro, Jo and Nicola. But the demands of a more sophisticated market do shift resources. Every euro dedicated to new hardware, independent consultants and additional audits an extra layer of due diligence is a euro not spent on the “soft infrastructure” of long-term success: community education, women’s groups, health benefits, local governance.

In the last part of our conversation, I asked about their view on rating agencies, the new actors sitting between developers and buyers, translating complexity into simple scores. In Part I, I looked at them from the buyer’s side, where they act as filters: “A vs BBB”, “high vs low risk”, “good vs bad”. From the developer side, the view is more cautious. Nicola understands why buyers want a second opinion, but sees a structural issue when developers are expected to pay for their own rating: be too strict and you lose clients, be too generous, and you lose credibility. A general point that emerges is whether a single organisation, often a startup, can meaningfully evaluate Forest Conservation, clean cooking, renewables and Direct Air Capture, given how different the skills and experience required are across these project types. Jo’s experience in forest conservation reinforces that point. High-quality Forest conservation work depends on land tenure, political realities and long-term relationships. Compressing that into a single comparative score next to a renewable energy plant or a tech-removal facility is not straightforward. This also raises a question on standardisation: can two projects of completely different nature really be considered “equal quality” because they have the same score? None of the three are anti-rating. What they want is simple: better alignment of incentives (for example, ratings requested and paid by buyers, not developers), more specialisation (teams focused deeply on specific technologies and contexts), and more direct buyer engagement (companies and investors visiting projects and understanding the work on the ground instead of outsourcing all judgement).

A functioning carbon market needs both truths at the same time. From the buyer’s side, if the reason to buy is unclear, demand stagnates and this is clear in the last 2025 MSCI report. From the developer side, if the reason to build disappears because they cannot even cover the costs, supply collapses. Integrity and quality are not only about reducing downside risk for buyers. They are also about ensuring the system rewards/support the people who do the work on the ground. If we optimise the entire market for accounting precision and reputational safety, we may end up with perfect projects on paper and no forests, no clean stoves, no water kiosks, and no community benefits. That would be a very strange definition of integrity and quality.

This article is also published on the Climate Playbook. illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.

Track the real‑world impact behind the sustainability headlines. illuminem’s Data Hub™ offers transparent performance data and climate targets of companies driving the transition.