A decision tree is the simplest honest way to think a choice through on paper. You start from a single Key question, you fan out the options you could take, and you follow each option to the outcomes it actually leads to. Nothing about the structure is complicated — a question, some branches, some ends — and that plainness is exactly why it works. It forces you to name the choice, to name the alternatives, and to be specific about where each one lands, instead of holding the whole tangle in your head and pretending you have compared it fairly.
This template is built for anyone facing a decision with more than one reasonable answer: a founder choosing between two go-to-market motions, a project lead picking a vendor, an engineer weighing build versus buy, a hiring manager deciding on an offer. The cards in this graph are deliberately generic — Key question, Option A, Option B, Outcome A1, Outcome A2, Outcome B1 — because the shape is the reusable part. You rename them to your own decision and the tree becomes yours.
What sets this apart from a whiteboard sketch is that it opens as a living graph in FlowGraph. Each branch is a real object you can carry evidence on, assign an owner to, and connect to the rest of your work — and every change you make carries a receipt, so the reasoning behind a decision survives the meeting it was made in.
What's in this decision tree template
The graph has six cards and five edges, arranged as a classic root-and-branch tree that reads left to right and then down. Here is what each part represents and why it is placed where it is.
Key question. This is the root of the tree — the one decision everything else hangs from. On the graph it is the single card with two edges leaving it. A good root is phrased as an actual question with a yes-or-no or either-or answer at stake: not "marketing strategy" but "Do we lead with the free template or the paid vault?" The sharper the question, the cleaner the branches.
Option A and Option B. These are the two paths you could take, drawn as the two branches leaving the root. On the graph, the Key question card connects to each option with a depends edge — the option you can pursue depends on which way you decide. Most real decisions start as a fork of two live candidates; if you have three or four options, you simply add more branch cards leaving the root. Keep each option to one clear course of action so its consequences stay traceable.
Outcome A1 and Outcome A2. Every option leads somewhere, and rarely to a single place. Option A branches into two outcomes — a good case and a not-so-good case — connected by feeds edges, because the option feeds into the result. This is the honest part of a decision tree that a pros-and-cons list skips: an option is not one outcome, it is a spread of them. Naming both the upside (Outcome A1) and the downside (Outcome A2) of Option A stops you from comparing your best case for one path against your worst case for another.
Outcome B1. Option B feeds forward to its own outcome. In this starter tree Option B carries a single modelled outcome, but the pattern is identical — you add as many outcome cards as the option realistically produces, each on its own feeds edge. The asymmetry in the starter graph is intentional: real decisions are rarely symmetric, and the template does not pretend they are.
Read together, the two edge types tell the story. A depends edge means "which path we can take depends on this choice." A feeds edge means "this path feeds into that result." That typed distinction — choice versus consequence — is what turns a pile of boxes into a tree you can actually reason down.
How to use it in FlowGraph
- Open the template. Click Open in FlowGraph to load the decision tree as a live, editable graph. Viewing, editing, and rearranging are free with no account — you can reshape the whole tree before you decide to save anything. You only reach the signup moment when you want to save the tree to a vault or plan with AI.
- Frame the question. Rename the root card to the exact decision you face. Phrase it as a question with a real fork in it. This single card sets the frame for everything below, so it is worth getting sharp before you branch — a vague root produces a vague tree.
- Branch the options. Rename Option A and Option B to your real alternatives, and add more branch cards if you have more than two. Then extend each option to its outcomes: keep the honest downside next to the hoped-for upside so the comparison stays fair. Every card you add and every edge you draw is a governed edit that carries a receipt of who changed what and when.
- Attach the evidence. Put the cost, the risk, the estimated probability, or the note behind each branch directly on its card. In FlowGraph a card is a knowledge atom — it carries provenance and a confidence level — so a branch can hold not just a label but the reasoning and the source that justify it. When you revisit the decision later, the evidence is still attached to the branch it supported.
- Cross-link to the real work. A decision does not live in isolation. Connect an outcome card to the project, document, or task it would trigger, using the typed relationship that fits. The tree stops being a thought experiment and becomes wired into the graph of everything else you are tracking.
- Ask AI to weigh the branches. With your own AI key you can ask FlowGraph to surface the risks on each option, propose outcomes you may have missed, or estimate the trade-offs — grounded in the evidence already on the cards. The AI proposes; you verify and decide. Nothing is written to the tree until you accept it, and when you do, the change carries a receipt like any other.
Why a living graph beats a static decision tree
A decision tree drawn once in a slide is a snapshot of your thinking on the day you drew it. It cannot hold the cost figures that justified a branch, it cannot tell you who owned the choice, and it certainly cannot connect to the work that flowed from the outcome you picked. Six months later, when someone asks why you went with Option B, the picture is mute — the reasoning evaporated the moment the meeting ended.
Because every card in FlowGraph is a real object rather than a shape, each branch can carry its evidence, its confidence, and its provenance, and each edge states a typed relationship — depends for a choice, feeds for a consequence. The tree becomes a durable record of not just what you decided but why, and it stays connected to everything the decision touches. When the situation changes you update the branch in place; the tree and the reasoning never drift apart.
And it stays honest. Every edit is governed and reversible, so the history of how a decision evolved is right there, timestamped, rather than reconstructed after the fact. That is the difference between a diagram you throw away and a decision you can stand behind.
Frequently asked questions
What is a decision tree?
A decision tree is a diagram that maps a choice and everything that follows from it. It starts at a single question, branches into the options you could take, and follows each option to its likely outcomes — so a decision can be reasoned through visually instead of argued in the abstract. The tree structure makes the alternatives and their consequences sit side by side, where they can be compared fairly.
The value is discipline. By forcing every option down to its concrete outcomes, a decision tree exposes the branch you were quietly avoiding and the downside you were quietly discounting. It turns a gut call into a structure you can inspect, challenge, and revisit.
How do I use this template?
Start from the Key question and rename it to the decision you actually face, phrased as a real fork. Add each option as a branch leaving the root, then follow every option to the outcomes it leads to — including the outcome you would rather not think about. In FlowGraph you can attach the cost, the risk, or the supporting note to each branch, so the tree carries the evidence and not just the labels.
Because everything is a live card, you are never locked into the starting shape. Add options, split outcomes, and cross-link a branch to the project it would set in motion. The template is a scaffold; the tree you end with is entirely yours.
Can AI help evaluate the branches?
Yes. With your own AI key you can ask FlowGraph to weigh the options, surface risks you have not named, and propose outcomes on each branch — always as a proposal you review, never as an automatic answer. Because the AI reasons over the evidence already attached to your cards, its suggestions are grounded in your graph rather than pulled from thin air.
The governing rule never changes: AI proposes, you decide. Every suggestion arrives as a change you accept or reject, and if you accept it, it lands as a governed edit with a receipt — so even the parts of the tree the AI helped with stay fully attributable to a human decision.