---
title: paying for a tree that is still alive
canonical_url: https://ensurance.app/guide/paying-for-a-tree-that-is-still-alive
markdown_url: https://ensurance.app/guide/paying-for-a-tree-that-is-still-alive.md
subtitle: planting is an activity. survival is a result. one model pays for the second.
category: nature-finance
---

# paying for a tree that is still alive

*planting is an activity. survival is a result. one model pays for the second.*

To pay for results in restoration means a buyer pays only for a result that has been checked, not for the work that went in. In the model World Resources Institute (WRI) has published, one party funds the planting up front, and an outcome buyer pays for each tree that is verified alive, every year it survives. No surviving tree, no payment.

A forest, a creek, or a herd already has a condition, whether or not anyone pays for it. Outcome-based financing pays only after an agreed result at that place is checked by someone independent. The result is a reading of the place, not the place itself.

Picture a cocoa farm in Ghana's Cocoa Belt, two years after a local group helped plant native shade trees between the cocoa. What matters now is how many of those trees are still standing. In Africa's semi-arid regions, WRI wrote in March 2026, fewer than half of planted trees survive, while the best projects keep more than 80% alive. That gap is the whole reason to pay for the second number instead of the first.

:::johnson
**a tree planted is an activity. a tree that is still alive next year is a result.** — Paying for results moves the money from the first sentence to the second. The hard parts are who counts the trees and who has a reason to pay for them.

[next: holds or improves, outcome payments into the place →](/guide/holds-or-improves-outcome-payments-into-the-place?from=guide)
:::

## how pay for results works in restoration

WRI's insight "Could AI Unlock Finance for Nature Restoration?" (August 2025, updated March 2026) names three actors:

| who | what they do | when money moves |
|---|---|---|
| Local organization or enterprise | Grows the trees with farmers and communities | Funded up front, so the work can start |
| Upfront investor or grantor | Covers the early costs and takes the early risk: a bank, a development finance institution, a fund, a philanthropy or a government | Pays first; is reimbursed as results are verified |
| Outcome buyer | Pays for surviving trees for every year they can be verified | Pays only after the count |

WRI's monitoring page, published with its January 2026 guidebook, puts the rule in four words: "No impact, no payment." Philanthropies and other flexible funders pay for the first critical years. WRI or another independent institution verifies the results. Then outcome payers reimburse the upfront funders for the project's costs and claim the impact, and the capital is recycled into more restoration.

A buyer who will pay for survival may not want the risk that half the seedlings die in the first dry season. An upfront funder can carry that risk if it expects to be reimbursed. The surviving tree is the hinge.

## a cocoa landscape, year two

WRI's own first example is a cocoa one, and WRI presents it as an illustration, not a signed deal. A group of companies that source cocoa from Ghana want trees on the farms in their supply chain but do not have the upfront capital. A grantor pays a local organization to help farmers plant native trees. Every year that independent AI data verifies the trees survive, the cocoa companies pay the grantor. After a few years the grantor is reimbursed and can revolve the money into the next project.

In WRI's telling, survival payments also make agroforestry more viable for farmers. The setting is real: the Ghana Cocoa Belt is a Restore Local landscape, where WRI reports TerraFund-backed groups have planted 4.1 million trees, with 9,300 hectares under restoration (October 2025). The deal is hypothetical. The trees are not.

## who buys restoration outcomes, and why

An outcome buyer is not a donor with a new label. It pays because it depends on the landscape. The buyers WRI names in the same piece:

- **Cocoa companies and international grocers** that source cocoa and fruit and need the farms in their supply chains to stay productive.
- **Local water and energy companies** whose supply runs through the catchment. [Who pays for forest restoration](/guide/who-pays-for-forest-restoration?from=guide) makes that case for water utilities in full.
- **Tour operators**, in WRI's second illustration: "fewer trees, fewer animals, few visitors, lower profits." Rwanda's Forestry Authority and the Wildlife Conservation Society did identify 10 priority sites for chimpanzee habitat that could not raise the money. The development-bank bond WRI sketches for them, with a tour operator paying for verified trees, is hypothetical.
- **Government agencies**, alongside those private buyers.

The common thread is dependence: each buyer pays to keep something it already relies on, which makes the payment a business cost, not a gift.

## how surviving trees are counted

None of this works without a count both sides trust. WRI describes a system, tested over four years and applied to more than 200 projects, with four parts:

1. **Targets set before the contract:** trees, hectares, jobs, people.
2. **Site boundaries,** mapped by local staff with a phone, so satellites and field crews know where to look.
3. **Reports every six months** through WRI's TerraMatch platform.
4. **Independent checks.** WRI counts trees with AI on satellite imagery at baseline, partway through, and at year five to assess survival. For the 20% of projects where imagery falls short, field crews count sample plots. WRI visits each project three times in five years.

The satellite does not replace the restoration lead's reports. It checks them, and WRI noted in January 2024 that each TerraFund organization is paid to monitor its own sites for six years. [What MRV actually is](/guide/what-mrv-actually-is?from=guide) covers the general case.

### two cost and accuracy figures

WRI's two pages describe the AI count differently. Read both:

| WRI page | what it says about the AI tree count |
|---|---|
| "Could AI Unlock Finance for Nature Restoration?" (updated March 2026) | "Preliminary results are 80% as accurate as traditional forestry methods that measure trees in the field, at 3% of the cost." |
| "Monitoring Locally Led Land Restoration at Scale" (with the January 2026 guidebook) | For most projects, with open ecosystems and available imagery, the AI count "is as accurate as traditional field measurements – at only 1% of the cost." |

They may describe different stages or project types; neither page reconciles them. Both support a narrower claim: remote counts are far cheaper than field inventories. WRI's January 2026 guide estimates the approach could cut independent verification costs by 98% across the first 198 TerraFund projects.

## where it stands

The verification side is running. WRI reports that 198 TerraFund projects across Africa have attracted more than $32 million, planted more than 21.2 million trees and brought 43,500 hectares under restoration. Restore Local, with $100 million over four years through The Audacious Project (announced April 2023, launched February 2025), works in Kenya's Greater Rift Valley, the Ghana Cocoa Belt, and the Lake Kivu and Rusizi River Basin.

The payer side is not yet public. As of November 2026 we found no signed outcome purchase agreement, no named outcome buyer and no price per tree on WRI's public pages. So there is no price to quote here, and none should be inferred. This post cites WRI's public pages only; it does not speak for WRI, and there is no partnership between WRI and **[ensurance](/natural-assets?from=guide&topic=outcome-based-financing)**.

## the limit wri names

WRI says it plainly: "counting trees is an incomplete stand-in for the vast amount of ecosystem services and economic benefits that holistic land restoration projects provide." A surviving tree says little alone about carbon, soil, water or species. WRI says it is exploring low-cost ways to use tree-count data to estimate water quality, biodiversity and carbon storage.

The other half is people: WRI sets targets for jobs and people alongside trees. [How to earn a living in ecosystem restoration](/guide/how-to-earn-a-living-in-ecosystem-restoration?from=guide) covers that side; [a planted tree is not a forest](/guide/a-planted-tree-is-not-a-forest?from=guide) explains why a count of stems is not a stand.

None of this dismisses the model. It asks for precision about what is being bought.

## what to ask before you pay for results

Five questions sort a structure from a slogan:

1. What exactly is the result, and what was the baseline before the season started?
2. Who calculates it, and who verifies it? Are they different parties?
3. How often is it read, and what happens to the payment if a drought kills trees in year three?
4. What is the price per unit, for how many years, and is it written down before anyone plants?
5. When the buyer pays, where does the money go?

The last question is where this series turns. In WRI's model, the outcome payment reimburses the upfront funder. The next post asks what changes if a rule written before the season pays into the place itself, into an account the place holds in its own name, when its condition holds and when it improves. That is being wired on our side. No reading has moved money yet.

If you depend on a restoration landscape, or fund one, tell us the place, the result you would pay for, who would check it, and who else depends on it: [start the conversation](/contact?from=guide&topic=outcome-based-financing). A reply tells us where to look first; it does not switch a program on.

## frequently asked

### how does pay for results work in restoration?

An upfront funder pays a local organization to grow the trees. After an independent check, an outcome buyer pays for the result (in WRI's model, per tree verified alive each year it survives), which reimburses the upfront funder.

### who buys restoration outcomes?

Organizations that depend on the landscape. WRI names cocoa companies, international grocers, local water and energy companies, tour operators and government agencies. As of November 2026, no named buyer or signed outcome purchase agreement appears on WRI's public pages.

### how are surviving trees counted?

Projects report every six months; WRI counts trees with AI on satellite imagery at baseline, partway through and at year five, with field plots for about 20% of projects. WRI's two pages give different accuracy and cost figures for the AI count.

### is this a carbon credit?

No. The unit is a surviving tree, not a tonne of carbon, and the buyer is someone who depends on the landscape. A buyer may treat trees as a proxy for carbon or habitat, as in WRI's chimpanzee illustration, but WRI's illustrations describe no credit being issued. WRI also notes that carbon and biodiversity certification is often too complex for smallholder landscapes. See also [alternatives to carbon credits](/guide/alternatives-to-carbon-credits?from=guide).

This is research, not investment, legal or tax advice.

## sources

[WRI, "Could AI Unlock Finance for Nature Restoration?"](https://www.wri.org/insights/ai-unlocking-nature-restoration-finance), by Nienke Stam, John Brandt, Radhika Rao and Will Anderson; first published 14 August 2025, updated 10 March 2026 — the three actors, per-tree-per-year survival payments, the cocoa, chimpanzee and mango illustrations, survival rates, the 80% at 3% claim, "incomplete stand-in."

[WRI, "Monitoring Locally Led Land Restoration at Scale"](https://www.wri.org/monitoring-locally-led-land-restoration-scale) — "No impact, no payment," capital recycling, six-monthly TerraMatch reports, AI counts at baseline, partway through and year five, the "as accurate at 1% of the cost" claim, field plots for 20% of projects, three visits in five years.

[WRI, *Land Restoration From Planting to Proof: A Guide to Monitoring, Reporting, and Verification*](https://www.wri.org/research/land-restoration-planting-proof-guide-monitoring-reporting-and-verification), 12 January 2026 — up to 98% lower verification cost across the first 198 TerraFund projects; nine indicator categories.

[WRI, "Locally Led Projects Restore 50,000 Hectares of Degraded Land"](https://www.wri.org/outcomes/locally-led-projects-restore-50000-hectares-degraded-land) — 198 projects, more than $32 million, 21.2 million trees, 43,500 hectares (read November 2026).

[WRI, "From Degraded Land to Thriving Farms: A Regeneration Story in Ghana's Cocoa Belt"](https://www.wri.org/insights/restoring-land-ghana-cocoa-belt), editor's note 16 October 2025 — TerraFund figures for the Ghana Cocoa Belt.

[WRI, TerraFund for AFR100 anniversary update](https://www.wri.org/update/terrafund-afr100-anniversary-learnings), January 2024 — organizations paid to monitor their sites for six years.

[WRI Africa, Audacious Project release](https://africa.wri.org/news/release-world-resources-institute-receives-audacious-project-investment-accelerate-restoration), April 2023, and [Restore Local launch release](https://africa.wri.org/news/release-restore-local-officially-launches-boost-locally-led-land-restoration-across-africa), 25 February 2025 — $100 million over four years; the three landscapes.

## the series

1. [what outcome-based financing actually is](/guide/what-outcome-based-financing-actually-is?from=guide)
2. [what an outcome bond investor actually holds](/guide/what-an-outcome-bond-investor-actually-holds?from=guide)
3. [pledged is not paid](/guide/pledged-is-not-paid?from=guide)
4. [paying for a tree that is still alive](/guide/paying-for-a-tree-that-is-still-alive?from=guide)
5. [holds or improves: outcome payments into the place](/guide/holds-or-improves-outcome-payments-into-the-place?from=guide)
6. [the buyer pays for what it depends on. the holder earns from what it holds](/guide/the-buyer-pays-the-holder-earns?from=guide)
