FIRE Model: Separate Facts from Interpretations
The FIRE Model helps you process criticism by walking four stages (Facts, Interpretations, Reactions, Ends) so you can answer what is checkable instead of reacting to the story you attached to it.
Framework Card
- Name:
- FIRE Model
- Goal:
- Separate observable facts from interpretations, reactions, and desired ends so criticism can be answered without treating the whole story as evidence.
- Flow:
- Facts → Interpretations → Reactions → Ends
- Best For:
- Receiving performance reviews without swallowing every opinion; Slowing down during a conflict about what actually happened; Checking the story you are telling yourself when anxiety spikes after feedback
Why it matters
Most people treat a harsh comment as one lump: the words, the tone, the implication, and the urge to defend or disappear.
That lump is rarely one thing. A reviewer says "the deck used too much jargon." The fact is the wording on the slides. The interpretation might be "they think I am not ready for this audience." The reaction might be shame or anger. The end you are pulled toward might be never volunteering to present again, or rewriting the whole deck tonight to prove them wrong.
That matters in performance reviews, tense 1:1s, and the hour after a meeting when your stomach is still tight. If you reply to the interpretation, you argue about respect. If you reply to the reaction, you apologize for existing. If you reply to the facts, you can check whether the jargon was real and what, if anything, to change.
FIRE is a way to slow that sequence down. It does not make criticism pleasant. It keeps the layers from collapsing into one verdict about you.
What it is
The FIRE Model is a four-stage sequence for evaluating what just happened, described by Mark Murphy (Leadership IQ; *Truth at Work*) as the way people already process events:
| Stage | Question it asks |
|---|---|
| Facts | What can I see or hear that is checkable? |
| Interpretations | What story did I add from my history, habits, and personality? |
| Reactions | What did I feel once I believed that story? |
| Ends | What am I now trying to protect, prove, avoid, or get? |
Facts are objective: wording, timestamps, who spoke, what was on the screen. Interpretations are not. Two people can hear the same sentence and read competence, contempt, or care into it. Reactions follow the reading, not the raw observation. Ends are the goal the emotion is driving toward: safety, status, fairness, or getting the work right.
This FIRE is Facts, Interpretations, Reactions, Ends. It is not the other workplace acronym that uses Facts, Impact, Request, and Expected outcome as a script for *giving* feedback. If you came looking for a delivery template, the methods in Alternatives are closer.
How it works
Walk the four stages in order. The point is to keep them from merging.
1. Facts
Write only what a camera or a transcript could confirm. "You used technical words on three slides" can be a fact. "You didn't appeal to the audience's self-interest" is one person's opinion unless other people, or a pattern of similar comments, give you reason to look closer.
Concentrate here first. If the fact is thin, the rest of FIRE will be a novel.
2. Interpretations
Name the reading you added: history, previous reviews, personality, the relationship with this person. Interpretations vary. Try not to spend the whole session here. The job is to label the story, not to litigate whether you are a good person.
3. Reactions
Once the interpretation is named, the feeling is easier to see: defensiveness, shame, relief, anger, freeze. FIRE does not ask you to suppress it. It asks you not to treat the feeling as extra evidence that the interpretation was true.
4. Ends
Emotions pull toward a goal: get them to take it back, look competent, leave the role, or actually fix the slides. Naming the end lets you choose whether that goal still makes sense once you return to the facts.
A usable pass through FIRE ends with a small, fact-based next step (ask a clarifying question, change a specific behavior, or set the opinion aside), not with a new identity.
How it compares
When another lens fits better, these methods do different jobs. They are not interchangeable labels for "feedback."
Related communication lenses
| Framework | What it helps you see | How it differs from FIRE |
|---|---|---|
| SBI Model | A message built from Situation, Behavior, and Impact | For *giving* behavior-based feedback. FIRE is for sorting a comment you already received. |
| FFC Technique | Feeling, Fact, and Comparison against an expectation | Another delivery pattern. FIRE's "Facts then interpretation" is inward; FFC is outward. |
| COIN Model | Context, Observation, Impact, and Next steps | A conversation spine for a hard message you will deliver, including a joint next step. |
| Radical Candor | Care personally and challenge directly (and the failure modes when one axis is missing) | A stance map for how you give feedback, not a four-stage receive sequence. |
FIRE is the lens when the comment has already landed and you need to keep facts, story, feeling, and goal from collapsing. Other methods help when the job is to *say* something specific, or to choose how direct to be while still caring.
When to Use This Framework
- Performance reviews: A manager named errors or gaps. You need to stay with what was observed instead of swallowing every adjective as identity.
- Arguments that stall: Two people are fighting about meaning. Asking "are we arguing about facts or interpretations?" can show whether the disagreement is evidence or reading.
- The spike after feedback: Sudden anxiety, rumination, or a plan to over-correct. Asking "what story am I telling myself right now?" is the Interpretation check.
Example
A concrete example makes the structure easier to reuse when you are under uncertainty.
Example: Written review after a client demo
Your manager writes: "In yesterday's client demo you talked over the engineer twice. The client left confused about who owns the technical answers."
Facts: Two interruptions during the demo; the client asked a follow-up that mixed product and engineering. Those can be checked against a recording or against people who were in the room.
Interpretations: "They think I am a poor teammate." "I am going to be taken off client work." Those are readings. They may or may not be what the manager meant.
Reactions: Heat in the chest; an urge to write a long reply explaining that the engineer was slow.
Ends: The emotion wants to restore status (prove you were helping) more than it wants a cleaner demo.
What FIRE does here: You can agree with the checkable interruptions, ask whether talking over the engineer is the pattern they want changed, and leave the identity story unproven. You do not need a speech about your intent.
Example: A peer says the design "wasn't ready"
In standup, a peer says the mockups "weren't ready" before engineering picked them up.
Facts: The files were shared Tuesday; engineering started Wednesday; two states were still marked "draft."
Interpretations: "They are saying I blocked the sprint." Maybe. The sentence did not say that.
Reactions: Irritation; a plan to list every late request they made last month.
Ends: Win the fairness argument.
Implication: FIRE would send you back to the draft states and the handoff time. If those facts are real, you can change the handoff rule. The fairness ledger is a different conversation, and it is not required to use the facts.
Production CMS has no example block. These scenes are educational illustrations, not Apply templates.
Takeaway
What FIRE can help with
- Splitting a painful comment into checkable facts, added story, feeling, and the goal you are being pulled toward
- Staying usable in a review or conflict long enough to answer the observation
- Noticing when you are about to reply to an interpretation as if it were evidence
What FIRE cannot replace
- A structure for giving feedback: Situation, Behavior, Impact (SBI), Feeling / Fact / Comparison (FFC), and Context / Observation / Impact / Next steps (COIN) are for composing a message to someone else.
- A stance about care and challenge: Radical Candor maps how directly you challenge while still caring. FIRE does not tell you how to show up as the giver.
- A verdict on whether the feedback is fair: Separating layers is not the same as deciding the substance. You still need evidence, and sometimes a second source.
- Conflict resolution as a whole: FIRE can de-escalate a facts-versus-story tangle. It does not negotiate interests, roles, or process design.
- Therapy or a diagnosis of the other person: Do not use the stages to infer motives or personality.
Honest scope: FIRE organizes how you *receive*. It does not write the other person's message, and it should not be sold as a way to stop taking things personally by force of will.
Frequently asked questions
Facts, Interpretations, Reactions, and Ends. It is Mark Murphy's sequence for how people evaluate events, used here to process criticism. It is not Facts, Impact, Request, Expected outcome.
Receiving. You use it on a comment, review, or argument that already happened. If you need to write what you will say to someone else, look at SBI, FFC, or COIN.
Sticking to facts is the first stage, not the whole method. FIRE also names the interpretation you added, the reaction that followed, and the end that reaction is chasing, so those layers do not silently drive the reply.
No. Reactions are a stage, not a defect. The discipline is to stop treating the feeling as extra proof that your interpretation was correct.
When more than one person concurs, or you have heard a similar comment before. FIRE still keeps that comment in Interpretation until you find a checkable fact. It does not ask you to dismiss a pattern just because the wording was opinionated.
A short list of checkable facts, a labeled story you are not treating as evidence, a named feeling, a named end, and one response that answers the facts (clarify, change a specific behavior, or set an unsupported opinion aside). If the output is only a new theory about the other person's motives, the stages were used as rumination.