How to adapt a decision tree template
- Rewrite the root as your actual question. One decision, phrased as a question with a real answer: "Buy or build the feature?" not "Feature strategy".
- Make the options mutually exclusive. Each first-level branch should be something you could do instead of the others. If two branches could both happen, they are not options, they are tasks.
- Put the uncertainty on the second level. What happens after you choose usually depends on something outside your control. The templates here all split a second time for exactly that reason.
- Name every leaf as an outcome, not a verdict. "Live in 3 months, team blocked" is a decision tree leaf. "Bad" is not.
- Keep the bracket labels. Text in
[square brackets]at the start of a line becomes a label on the line leading into that node, which is what makes a decision tree readable without a legend.
Which template to start from
Buy vs. build is the right shape for any make-or-acquire question, including hiring an agency versus training a team. Hire now or wait a quarter generalises to every timing decision — the structure is "act now with a risk" against "wait with a different risk". Accept the job offer is the personal-decision template, and its useful feature is that staying put is drawn as a branch with its own outcomes rather than as the absence of a decision. Ship or hold for the fix is the template for a decision that hinges on one probability estimate, which is the cue to switch modes and let the tool do the arithmetic.
Adding numbers to a template
A decision tree earns its keep when the branches carry values. Switch the mode selector to probability and write each branch's likelihood in parentheses — (0.6), (60%), or (3/5) — and the tool multiplies along every path and warns you when a set of sibling branches does not sum to 1. Multiply each path probability by that outcome's value and you have the expected value of the branch. The probability tree diagram maker covers that mode in full.
Often the useful result is not a winner but a discovery: the decision turns entirely on one number nobody has measured. That is worth more than the diagram, and it is the reason to draw the tree before the meeting rather than during it.
Frequently asked questions
What should a decision tree template include?
A single decision at the root, one branch per option, a labelled outcome at the end of every branch, and — where the outcome depends on something you cannot control — a second level of branches for the uncertainty. A template that only shows the options you like is not a decision tree; every branch needs its bad case written down.
Are these decision tree templates free?
Yes. Every template opens in the editor with no signup and no limit on how much you change. PNG and PDF exports are free and carry a small footer watermark; the copyable plain-text version never carries one.
Can I use a decision tree template in Word or PowerPoint?
Export the finished tree as a PNG and paste it into the slide or document. For a version you can keep editing later, use the share link — it encodes the whole outline in the URL, so pasting that link into your notes gives you an editable copy of the diagram rather than a flat picture.
How is a decision tree template different from a flowchart template?
A flowchart can loop and merge; branches rejoin. A decision tree never rejoins — every path from the root is a distinct scenario, which is what lets you compare them and attach probabilities or values to each. If your paths need to come back together, you want a flowchart, not a decision tree.
For the full guide to the mode — expected value, when a tree is the wrong picture, and the keyboard shortcuts — see the decision tree maker. The tree diagram maker handles plain hierarchies, hierarchy charts handle reporting lines, and the examples gallery has a starting point for every other mode.