Where the 30% Goes

The 30% of GreenSweep revenue not directed to environmental projects covers platform operations, fraud prevention, technology infrastructure, GDPR and multi-jurisdiction compliance, localisation into seven languages, and project verification against Gold Standard, Verra, and Plan Vivo. It protects the integrity of the 70% that flows to verified climate projects, and a ratchet clause in our statute commits us to shrinking it over time.

Seventy percent of GreenSweep’s revenue is committed to environmental projects our community votes for. That number is enforced by Maltese law, not by a pledge or a press release.

Today we want to talk about the other number.

Most organisations publish a pie chart, label one slice “programme costs” and the other “administrative,” and call it transparency. That tells you almost nothing. Administrative costs at one organisation might include a compliance team preventing fraud. At another, a headquarters lease nobody has renegotiated since 2014. The label is the same. The substance is entirely different. Watchdogs like

exist precisely because that single “admin” label hides too much.

Here is what the 30% actually buys, line by line.

Fraud Prevention and Lead Quality

Every registration on GreenSweep passes through a scoring system that evaluates twenty-seven signals — device fingerprint, IP reputation, behavioural timing, email domain validity, and more. This system exists because our revenue depends on the quality of our user base, and the projects we fund depend on our revenue. A single batch of fraudulent leads could get our commercial partnerships suspended and cut off funding to every project on the platform. Fraud prevention is not overhead. It is the immune system.

Regulatory Compliance

GreenSweep operates across multiple jurisdictions — the EU, the Philippines, India, Nigeria, the United Kingdom. Each has its own data protection regime, advertising disclosure requirements, and consumer protection standards. Our consent architecture follows

(lawfulness, purpose limitation, data minimisation) with market-specific configurations (double opt-in where required, single opt-in where permitted). We employ a Data Protection Officer, file with the

, and build compliance into the platform rather than bolting it on afterward. This costs money. The alternative — cutting corners and hoping regulators don’t notice — is not a strategy we’re willing to adopt.

Infrastructure and Platform Operations

The website, the database, the authentication system, the voting mechanism, the real-time transparency dashboard, the email infrastructure, the content delivery network. These are not glamorous. They are also not optional. A platform that goes down costs more in lost trust than it saves in server fees. We host on European infrastructure, encrypt data in transit and at rest, and run monitoring systems that check site health every fifteen minutes.

Localisation and Translation

GreenSweep serves communities in seven languages across fifteen countries. Each locale requires translated content, culturally appropriate messaging, and market-specific legal language. A registration flow that works in Germany does not work unchanged in the Philippines — not because of language alone, but because consent norms, data expectations, and community trust patterns differ. Getting this right is not a luxury. It is the difference between a platform people use and one they abandon.

Project Verification and Due Diligence

Before a project appears on GreenSweep, it goes through a structured evaluation: verification status (Gold Standard, Verra, or equivalent), additionality, permanence, measurability, co-benefits, geographic balance. We maintain a catalogue of twenty-five projects across five markets, each with documented funding pathways and impact metrics. This due diligence is what allows us to say “ verified ” and mean it. For how the registries themselves work, see Inside Gold Standard.

Team Costs

GreenSweep is a small team. The work described above — fraud prevention, compliance, infrastructure, localisation, verification — is done by humans and AI working together. We are honest about this: we use artificial intelligence extensively across our operations, not to replace judgement but to compress the cost of functions that would otherwise require departments we cannot afford.

That is the 30%. It is not a slush fund. It is the operating cost of doing this properly.

The 30% exists to protect the 70%. Every quarter, we intend to make it smaller — or make it work harder.

The Ratchet

Now, the commitment. Our founding documents include a ratchet clause: the 70% allocation to projects is a floor, not a ceiling. Our stated target is 85%, and we intend to reach it. Not by cutting corners on compliance or verification — those are non-negotiable — but by making every operational function more efficient over time.

This is where the AI opportunity becomes real. Automated compliance monitoring across jurisdictions, fraud detection that improves with every registration, translation workflows that reduce per-language costs by an order of magnitude, due diligence processes that surface red flags in minutes rather than weeks. Each of these capabilities exists today. We are building them into our operations from day one, which means our operating costs should decline as a percentage of revenue even as the platform grows.

We should be transparent about something else: these are our current estimates, built before launch, based on the best modelling we can do with the information available. They will be wrong in places. Some costs will come in lower than projected. Some will surprise us. What will not change is the commitment to publish actuals alongside projections — so you can see where we were right, where we were wrong, and what we did about it.

In the early period, the 30% will sometimes generate a working capital buffer — months where operational costs come in below the allocation. That buffer is not profit. It is not distributed. It is reinvested to expand GreenSweep’s geographic footprint: new markets, new languages, new communities brought into the voting system. Every new market we enter multiplies the number of people directing funding to the environmental projects that matter to them. The buffer funds growth, and growth funds impact.

At some point, the model could reach a steady state — operational costs stabilised, geographic footprint mature, revenue predictable. Frankly, we hope it doesn’t. We hope that by the time the current engine reaches equilibrium, we’ve developed a phase two play — a new layer of capability or reach that we can stack on top of what exists, driving more impact, reaching more communities, channelling more capital toward restoration and protection. The ambition is not to build a comfortable machine. It is to build a machine that keeps finding new problems to solve.

We don’t claim to be lean enough yet. We claim to be honest about what things cost, and structured so that every efficiency gain flows to projects or to the expansion that brings more communities into the system.

You can see the current allocation breakdown at Transparency. For the legal machinery that binds the split, read Why We’re Not a Charity. For how the audit timing problem plays out in conventional charities, see The Twelve-Month Lag.