---
title: Compare Tool
group: For donors
order: 6
summary: Put two or three interventions side by side on depth, evidence, durability, feedback, who steers and cost, and see exactly where they differ. There is no overall score, because the trade-offs are yours to weigh.
status: live
conditions: [single-lens, overhead-myth]
help:
  - Add an intervention to the database with a pull request to src/content/interventions/, so people can compare it.
  - Challenge a depth, evidence, feedback or durability rating on an existing intervention with a sourced pull request.
  - Suggest an attribute the comparison is missing, with the evidence that would fill it, in an issue tagged tools.
related: [/tools/impact-calculator/, /tools/values-compass/, /diagnosis/single-lens/, /diagnosis/overhead-myth/]
---

## How it works

Choose two or three interventions from the [Intervention Database](/record/interventions/). The tool lines them up on the same attributes the database records for every entry:

- **Depth:** where the work acts, from relief to source, on the [depth scale](/record/encyclopedia/symptoms-and-sources/)
- **Evidence:** how strong the published evidence is
- **Mechanism:** how it works, in words
- **Time to effect** and **durability:** how soon results appear and how long they last
- **Feedback:** how clearly results travel back to the people steering the work
- **Who steers:** whether the work is community-led, co-designed or top-down
- **Cost data:** a published cost per outcome, if one exists
- **Causes** and **related conditions** from the [Diagnosis](/diagnosis/)

Every rated row is marked “Differs” or “Same”, so you can see at a glance where the choice actually lies, and you can hide the rows that match. Each name links to its full entry and sources. On a phone the comparison stacks by attribute, so the values you are comparing stay next to each other.

## Why there is no score

Adding the rows into one number would mean deciding for you how much durability is worth against evidence, or community control against speed. Those are value judgments, and reasonable people make them differently. A single score would hide that choice inside a formula. Seeing good through one framework without noticing is [the Single Lens](/diagnosis/single-lens/), and a ranking would invite it.

Depth in particular is not a grade. Relief, repair, prevention, structure and source work are all needed. The comparison shows where each intervention acts so you can choose your mix on purpose.

There is no overhead column. How much of its budget a charity spends on staff and systems says little about what the work achieves, which is [the Overhead Myth](/diagnosis/overhead-myth/).

## What it isn’t

It compares interventions, which are kinds of work, not charities. It doesn’t rate any organization. Cost figures appear only where the database has a published, sourced estimate, and figures that count different outcomes can’t be compared directly. Unknown values are shown as unknown.

Your selections live in the page address, so a share link reproduces the same comparison. Nothing is sent to a server.