Article
Writing a Report: The NATO Method, Usable Tonight
"Things are kicking off downtown" isn't information. Here's how serious organisations pass on what they observe: a six-line format and a two-character grading, both usable by a civilian.
Someone close to you calls: "things are kicking off near the station, looks serious, be careful."
What do you do with that? You don't know how many people, what they're doing, exactly where, when (ten minutes ago or two hours?), or whether your caller saw it themselves or read it somewhere.
That message contains emotion and zero usable information. It's the default format of nearly every civilian report, and it's fixable in ten minutes of learning.
Organisations that handle information under time pressure solved this problem long ago, with two simple tools: a reporting format and a grading system. Both are public, both transpose directly, and this article shows how to use them without ever having worn a uniform.
The principle that comes first
A report is not a story.
A story mixes what you saw, what you concluded from it and what you felt. A report separates them, in that order of priority, and the first category takes up most of the space.
Three lines that illustrate the difference:
- "They're planning something" → interpretation.
- "I was frightened" → feeling.
- "Around fifteen people gathered outside the station, banners, 14:20" → observation.
Only the third one gets transmitted. The other two may follow, clearly labelled as such, because your interpretation has value, provided the recipient can distinguish it from what you actually saw.
It's the direct extension of what we wrote about confirmation bias: whoever receives a story inherits your conclusions; whoever receives a report can form their own.
The format: six questions in a fixed order
The format NATO forces use to report an observation fits in an acronym, SALUTE, each letter matching a question. Here is its civilian transposition:
| Question | What you note |
|---|---|
| How many? | Number of people, vehicles, items |
| What? | What they're doing, factually, without qualifying |
| Where? | The most precise location possible, with visible landmarks |
| Who? | What identifies them: clothing, markings, language |
| When? | Time of observation, and how long it has lasted |
| With what? | Visible equipment, vehicles, notable gear |
Always in that order, always all six, even to say "not observed". An empty field is information: it tells the recipient what you couldn't see, and that's often as useful as the rest.
Take the opening call again, reformatted:
"Fifteen to twenty people, gathered and stationary, on the north forecourt of the station on the avenue side, civilian clothing, banners with no text legible from where I was, observed at 14:20 for about five minutes, no vehicles noticed. I'm a hundred metres away, I did not approach."
The same time to say as the original message. Incomparably more useful.
Three rules that make the difference:
Describe, don't qualify. "Man, dark jacket, backpack" is an observation. "Dodgy-looking guy" isn't: that describes your impression, not the person.
Give the time of observation, not the time of sending. Information without a timestamp is unusable: nobody knows whether it's five minutes or three hours old.
Locate by visible landmarks. "Next to the pharmacy, corner of Market Street" beats an approximate address, and vastly beats "somewhere in the centre".
The grading: two characters that change everything
Here is the most interesting tool, and the least known to the public.
When information travels through a NATO organisation, it carries two characters: a letter and a number. This system, called the Admiralty Code or NATO system, rests on an idea that ought to be obvious and never is: the reliability of the source and the credibility of the information are two separate questions, to be assessed separately.
The letter (the source, based on its track record):
| A | Completely reliable |
| B | Usually reliable |
| C | Fairly reliable |
| D | Usually unreliable |
| E | Unreliable |
| F | Reliability cannot be judged |
The number (this specific piece of information, here and now):
| 1 | Confirmed by other independent sources |
| 2 | Probably true |
| 3 | Possibly true |
| 4 | Doubtful |
| 5 | Improbable |
| 6 | Truth cannot be judged |
You put them together: B2 means a usually reliable source reporting probably true information.
Why this beats "true / false": it allows the combinations reality produces all the time. E1: an unreliable source whose specific piece of information has nonetheless been independently confirmed. A4: an excellent source that, this time, is reporting something doubtful. A single scale would flatten both cases; two scales preserve them.
And two values are deliberate admissions of ignorance: F means you don't know the source, 6 that you have no way to judge the content. These aren't judgements about the quality of the information but about the state of your knowledge. Being able to write that down is valuable: it stops you turning uncertainty into assertion.
Civilian use is immediate. A message from a neighbour who is often wrong, about something nothing corroborates: D4. An official statement cross-checked by two outlets: A1. An anonymous screenshot: F6, and you then know exactly what to do with it, namely wait.
An honest caveat: this grading remains subjective, and it was designed for human intelligence chains rather than for social media feeds. Its value isn't in the precision of the grade, it's in forcing you to separate two questions the mind spontaneously merges.
The cycle: why you collect at all
One last element, which puts the rest in its place. Intelligence works as a loop, in four stages:
A question asked: what you're trying to find out, and for which decision. Collection directed by that question. Processing: sorting, cross-checking, grading, discarding. Dissemination to whoever must decide, in a format they can use.
The point that matters: the loop starts with a question, not with collection. Accumulating information without a prior question produces noise, never intelligence. It's the same sentence as in our OSINT article, and it bears repeating: intelligence is an answer to a question you took the trouble to formulate before starting.
The limits, and the classic mistake
What these tools give you: reports that are usable by the people receiving them, a timestamped trace that remains usable days later, and the mental habit of separating what you know from what you assume.
What they don't give you: the truth. A perfectly formatted report can be entirely wrong. The format improves transmission, not the quality of the observation.
The classic mistake is playing soldier. No jargon, no acronyms, no invented call signs: it impresses a group and confuses everyone outside it. What you take from the NATO model is two mechanisms: the fixed order and the source/content separation, not its vocabulary.
The second mistake is believing this is a subject for major crises. These tools serve everyday purposes first: reporting a problem to a building manager, describing an incident to an insurer, passing an observation to a colleague. A well-built report is a professional skill before it's a preparedness one.
Where to actually start
- Learn the six questions in order: how many, what, where, who, when, with what. Ten minutes.
- Practise on the mundane. Describe an unremarkable street scene in six lines, today. On stakes-free material, never during a crisis.
- Always note the time of observation, separately from the time of sending.
- Grade what you pass on, if only mentally. Two characters before sharing.
- Formulate the question before searching: which decision should this information inform?
- Keep notebook and pencil accessible. A report written on the spot beats ten memories reconstructed in the evening.
Conclusion
Intelligence is neither a shadowy profession nor a matter of equipment. It's a writing discipline: saying what you saw, in an agreed order, flagging what you don't know, and separating the source from the content.
Two tools are enough, both public and free: six questions in a fixed order, and two characters of grading. They turn "things are kicking off near the station" into something someone can act on.
Six questions. In order. With the time. And two characters to say what it's worth.
Companion sheet: Download the printable A4 PDF
Sources and further reading
- Intelligence grading (Wikipedia FR) (in French). NATO doctrine: separate assessment of source reliability and content credibility.
- The Admiralty Code / NATO 6×6 system: detail of both scales, exact meaning of the extreme values and example combinations.
- Enterprise Open Systems: The Admiralty Code for information verification: uses outside defence (journalism, cybersecurity) and acknowledged limits of the system.
- SRM: What is the Admiralty Scale?: official definitions of the six reliability levels and six credibility levels.
- ICAO: Cyber information sharing (in French): an international civil organisation recommending the Admiralty Code to assess sources and content.
- Standards for evaluating source reliability in intelligence production: comparison with other grading systems and the associated research.