The Trip Debrief Habit
Builds guide sites for a livingPublished July 24, 2026Last updated July 26, 202637 min read
- 14 CFR 91.25 bars the FAA from using Aviation Safety Reporting Program submissions in enforcement, because honest reporting is impossible where reporting is punished.
- Keep two records: an operational one that is factual and shareable, and a private one that is unguarded and is the one with the learning in it.
- Write it in the truck before driving, because nobody has ever sustained an evening habit past a fortnight.
- Record conditions as numbers rather than impressions, since three seasons of high and off colour produces nothing usable.
- Record the near misses, which are the free lessons and the entries nobody writes because nothing happened.
- Read nothing during the season and everything once in November, looking for repetition rather than for individual days.
The FAA runs a reporting system on one promise: reports submitted to it will not be used against the person who filed them in any enforcement action, with accidents and criminal offences excluded. The immunity is the mechanism, because nobody reports their own mistakes into a system that punishes them.
A guide debriefing their own day is running a reporting system with an audience of one, and the same rule applies. A record that exists to be defended is a record that will be sanitised, and a sanitised record teaches nothing. What follows is a debrief that takes four minutes, survives a season, and is written in the knowledge that nobody but you will read it. Nothing here is legal advice, and the aviation programme reaches nobody outside aviation. The rest of this cluster is gathered at the ops playbooks hub.
| Line | Why it exists |
|---|---|
| Conditions, with numbers | Three seasons of this beats any forecast |
| What worked and what did not | The only pattern you can act on |
| Something about the person | Everything in the retention cluster runs off it |
| What nearly went wrong | The near miss is the free lesson |
Why does the immunity matter?
Because it is what makes honest reporting possible at all.
Section 91.25 of Title 14 provides that the Administrator will not use reports submitted to the National Aeronautics and Space Administration under the Aviation Safety Reporting Program, or information derived from them, in any enforcement action, excepting information concerning accidents or criminal offences which are wholly excluded from the programme.
Which is a deliberate trade: the regulator gives up the ability to act on what it learns, in exchange for learning it.
The design point is that safety information exists overwhelmingly inside the heads of people who would be exposed by sharing it.
Translated to a one-person operation, the equivalent is that a debrief written for anybody else is written defensively, and a defensive record is worthless as a learning instrument.
Section 91.25 is carried on the eCFR.
Where the day's other records live is set out by the spreadsheet CRM piece.
What does that mean in practice?
Write it as though nobody will read it, because nobody will.
A debrief that says the client was difficult is more useful than one that says the day was challenging, and only the first will be written if you imagine an audience.
Which is worth deciding explicitly, since a guide who has ever been asked to produce a record will write differently forever afterwards.
The practical resolution is to separate the two things: an operational record that is factual and shareable, and a private note that is honest and is not.
The operational record covers conditions, times, what happened and what was offered, and it is the one that would be produced if anybody ever asked.
The private note covers judgment, mistakes and impressions, and its whole value depends on it being unguarded.
Where an incident occurs, the operational record is the one that matters and it should be complete, contemporaneous and factual.
What that record needs to contain is set out by the insurance claims piece.
What four minutes a day compounds into. Four minutes across a hundred trips is under seven hours a season. What it produces is three hundred conditions readings, a hundred client notes and a hundred near-miss entries, which is a data set nobody else on your water has. Three seasons of it answers questions that no amount of recollection can: what flow the lower stretch actually turns on at, which weeks genuinely fish, and which clients said they would return and did not. Every figure here is a stated assumption.
When does it get written?
Before you get home, and that is not negotiable.
Every guide who has promised to do it in the evening has stopped within a fortnight, because the evening after a ten-hour day is not a working period.
Which means it happens at the take-out, in the truck, before driving, and it takes the length of time it takes to drink something.
The specific enemy is the assumption that today was ordinary enough to remember, since ordinary days are exactly the ones that blur.
Voice notes work for people who will not type, provided there is a habit of transcribing them weekly, and that habit is the weak point.
What works best is the shortest thing you will actually do, which for most people is four lines typed into the same note every time.
Consistency of format matters more than completeness, because a consistent record can be read in bulk and a rich one cannot.
Why the trigger matters is set out by the maintenance piece.
The reporting programme reaches nobody outside aviation. 14 CFR 91.25 concerns the FAA's treatment of reports made to NASA under the Aviation Safety Reporting Program and confers no protection on anybody else, in any context. Nothing you write about your own operation is privileged by anything described here, and a record you create may be discoverable depending on the circumstances and the law of your state. Take advice before assuming otherwise, and nothing on this page is legal advice.
What goes in the conditions line?
Numbers, because impressions do not compound.
Flow, water temperature where you measure it, air temperature, wind and sky, plus whatever your water's own indicator is.
The reason numbers rather than words is that three seasons of high and off colour tells you nothing and three seasons of nine hundred and forty-two does.
Which is the single highest-value line in the whole debrief, because it is the only one that produces knowledge unavailable anywhere else.
Adding what the fishing actually did alongside it is what turns the readings into a model, and the pairing is the point.
Over enough seasons the file answers the question every client asks and no guide can currently answer with evidence, being when their water genuinely fishes best.
Which is also marketable, since a guide who can show three years of data about their own river is doing something nobody else on it is doing.
How that becomes publishable is set out by the blogging piece.
What is the near-miss line for?
The free lesson, and it is the one nobody records.
Aviation reporting exists overwhelmingly for events where nothing happened, because those are the ones that carry the information without the cost.
The guiding equivalents are the fish hook that nearly went in somebody, the wading step that nearly went wrong, the rock you saw late, the client who nearly stood up at the wrong moment.
All of which are forgotten within a day precisely because nothing happened, and all of which are the cheapest safety information available.
Writing them down is uncomfortable in a way the other lines are not, since a near miss is usually a small failure of attention.
Which is exactly why the record has to be private, and it is the clearest application of the immunity principle at the top of this page.
Three seasons of near misses reveals a pattern about your own operation that no amount of general safety advice will.
What the pattern feeds into is set out by the contingency piece.
What about the client line?
One line, about the person, not the fishing.
Who they came with, what the occasion was, and one detail that is not about angling.
Which is the line that makes every retention practice possible, since a message a year later is only worth sending if it contains something specific.
Twenty seconds, written the same day, and it is the difference between a follow-up that lands and one that reads as a template.
The rule worth applying is to write it as though the client might one day read it, because occasionally somebody does and because it costs nothing to be decent.
Which is a different standard from the near-miss line and the separation is deliberate: facts about people get the shareable standard, judgments about yourself get the private one.
Where that distinction feels awkward, keeping them in two different places resolves it entirely.
What the follow-up does with it is set out by the client touches piece.
What is read, and when?
Nothing during the season, everything in November.
A record read continuously is a record that produces anxiety rather than knowledge, since a single bad day looks like a trend when read in isolation.
Which means the discipline is to write during the season and read after it, once, in a single sitting with the whole file open.
What appears at that point is repetition, which is the only signal available at this sample size and is invisible day by day.
Three near misses in the same place, four clients mentioning the same thing, two cancellations for a reason you had not registered as a category.
None of those are visible without the file and all of them are obvious with it.
Acting on two of them is the entire return, and attempting to act on all of them is how the exercise gets abandoned.
Where that reading sits is set out by the winterising piece.
What does the record answer that memory cannot?
Anything about frequency, and memory is systematically wrong about it.
Ask a guide in November how many days were lost to weather and the answer will be shaped by the most recent loss and by the worst one.
Ask the file and the answer is a number, and the number is usually well below the impression, which changes what gets decided about next season.
The same distortion runs through every operational question: how often clients are late, how often a stretch fishes badly, how often a piece of gear fails.
All of them feel frequent because the bad instances are memorable and the ordinary ones are not, and all of them are countable if anybody wrote them down.
Which is the actual return on the habit, since decisions made from a corrected frequency are different decisions.
An operation that believes it loses fifteen days a year to weather prices and schedules differently from one that knows it loses six.
The measurement discipline this belongs to is set out by the numbers piece.
Should anything be shared with clients?
The conditions data, eventually, and nothing else.
Three seasons of readings is genuinely interesting to anybody who fishes your water, and publishing a version of it is unusual enough to be noticed.
Which is a use for the record that nobody anticipates when starting it, and it is the one that eventually justifies the whole habit commercially.
What can be published is the water rather than the people: flows, temperatures, when things happened, and what the fishing did.
What cannot is anything identifying a client, anything about another guide, and anything from the private half of the record.
The line is easy to hold provided the two halves were kept separately from the start, which is the argument for the separation made above.
Where a client's day is genuinely the story, asking them is a two-line message and the answer is almost always yes.
Does a second guide keep one?
They should, and it is harder to make happen than for yourself.
An employed or contracted guide has no obvious incentive to record anything, since the compounding value accrues to the operation rather than to them.
Which means the requirement has to be small, explicit and part of the day rather than an aspiration expressed in February.
Four lines, sent as a message at the end of the day, is the version that gets done, and anything requiring access to a system does not.
The near-miss line is the one that matters most and the one least likely to be filled in honestly by somebody who is worried about the work.
Which brings the immunity principle back with force: a second guide reports honestly only if reporting has visibly never been used against anybody.
Saying so explicitly, once, and then demonstrating it, is the only thing that produces useful reporting from anybody other than yourself.
The hiring question underneath it is set out by the capacity signal piece.
What about the trips that went well?
Recorded too, and they are the half everybody skips.
The instinct is to write when something went wrong, which produces a file describing an operation that is failing continuously.
Reading that file in November is demoralising and misleading in equal measure, since the good days are the population and the bad ones are the sample.
Recording what worked matters practically as well, because a technique that produced a good day is exactly the thing forgotten by the following season.
Which is the difference between a guide who improves and one who merely accumulates experience: the first can say what they changed and the second cannot.
The entry for a good day is usually shorter, being conditions and one line about what was different, and it takes ninety seconds.
Skipping it is why most guides can describe their worst days in detail and cannot reconstruct their best ones at all.
Where does it live?
With the trip record, in one file, in plain text.
A separate debrief app, notebook or system is a second place for the truth to live and it will diverge from the first within a season.
Which means the entry attaches to the booking it describes, in whatever holds the bookings, rather than existing independently.
Plain text matters more than it sounds, because a format you can search, copy and read in ten years is worth more than one that is convenient this year.
Anything requiring an application to open is a format that will eventually be unreadable, which is a real consideration for a record whose value is cumulative.
Backing it up is the other half, and it belongs in the same routine as backing up photographs.
What that routine looks like is set out by the photography workflow piece.
Does it survive a bad season?
Only if it is short enough, which is the whole design constraint.
Every record-keeping habit dies the same way, which is that it grew and a tired person stopped doing it.
Which argues for four lines rather than a template, and for accepting an incomplete entry over no entry at all.
The version that survives is one where a day recorded as blown out, nothing to say counts as a complete entry, because it preserves the habit.
Guides who build elaborate debrief systems in February have all abandoned them by August, and the survivors are always people doing something almost trivially small.
Which is worth designing for deliberately rather than discovering: pick the smallest version, and add to it only after two full seasons.
Adding a fifth line in year three is a reasonable decision and adding it in week three is how the whole thing fails.
The same principle applied to checklists is set out by the daily checklist piece.
Where do debriefs fail?
Six ways, and the evening is the first.
Deferring it to the evening, when nobody has ever sustained it past a fortnight.
Writing impressions instead of numbers, so three seasons of conditions data produce nothing usable.
Writing defensively, which sanitises exactly the entries that carried the value.
Skipping the near misses, because nothing happened and it is uncomfortable to record a lapse.
Reading it continuously during the season, which converts a single bad day into an apparent trend.
And building something elaborate, which is abandoned by August every time.
What the record feeds is set out by the repeat rate piece.
What is the working habit?
Four lines, in the truck, before you drive.
Conditions with numbers, and what the fishing actually did alongside them.
What worked and what did not, in your own words, without softening.
One line about the person, written to a standard you would be content for them to read.
One line about anything that nearly went wrong, written privately and honestly, because that is the entry with the most information in it.
Attach it to the booking rather than to a separate system, keep it in plain text, and back it up with the photographs.
Read nothing until November, then read all of it once and act on two things.
Keep it small enough to survive August, and add to it only after two complete seasons.
The statutory basis for the aviation rules is 49 U.S.C. 44701, with the section mirrored on govinfo.
The morning half of the same discipline is set out by the daily checklist piece.
How this was checked. The enforcement prohibition comes from 14 CFR 91.25, which provides that the Administrator of the FAA will not use reports submitted to the National Aeronautics and Space Administration under the Aviation Safety Reporting Program, or information derived therefrom, in any enforcement action, except information concerning accidents or criminal offenses which are wholly excluded from the Program. The section was read on the Electronic Code of Federal Regulations on 26 July 2026. It governs the FAA's treatment of reports made to NASA under that programme and confers no protection on anybody outside it, in any context. Nothing a guiding business writes about its own operations is privileged by anything described on this page, and a record so created may be discoverable depending on the circumstances and the law of the relevant state, neither of which was researched here. Nothing on this page is legal advice. No industry figure for record-keeping practice, incident rates or near-miss frequency in guided fishing is asserted, because no consulted source publishes one; the arithmetic panel uses stated illustrative assumptions.
If your booking calendar has more open weeks than you’d like, I’ll build you a free preview of your booking site before you pay a cent.
Get a free website previewThe four lines, when they get written, and why nothing should be read until November
Why does the immunity matter?
14 CFR 91.25 provides that the Administrator will not use reports submitted to NASA under the Aviation Safety Reporting Program, or information derived from them, in any enforcement action, excepting accidents and criminal offences which are wholly excluded. The regulator gives up the ability to act on what it learns in exchange for learning it, because safety information sits inside the heads of people who would be exposed by sharing it. A debrief written defensively is worthless as a learning instrument.
What does that mean for a guide?
Separate the two records. An operational record covering conditions, times, what happened and what was offered, which is factual, contemporaneous and would be produced if anybody ever asked. And a private note covering judgment, mistakes and impressions, whose whole value depends on being unguarded. Nothing you write is privileged by anything described here, and a record may be discoverable depending on your state; take advice before assuming otherwise.
When should it be written?
At the take-out, in the truck, before driving. Every guide who has promised to do it in the evening has stopped within a fortnight, because the evening after a ten-hour day is not a working period. The specific enemy is assuming today was ordinary enough to remember, since ordinary days are exactly the ones that blur. Four lines typed into the same note every time is what survives.
What goes in the conditions line?
Numbers, because impressions do not compound. Flow, water temperature, air temperature, wind and sky, plus your water's own indicator, paired with what the fishing actually did. Three seasons of high and off colour tells you nothing and three seasons of nine hundred and forty-two does. It is the only line that produces knowledge unavailable anywhere else, and eventually it is publishable.
Why record near misses?
Because they are the free lessons. Aviation reporting exists overwhelmingly for events where nothing happened, since those carry the information without the cost. The guiding equivalents are the hook that nearly went in somebody, the wading step that nearly went wrong, the client who nearly stood at the wrong moment. All forgotten within a day precisely because nothing happened, and all cheap safety information.
When should the file be read?
Nothing during the season, everything once in November. A record read continuously produces anxiety rather than knowledge, since one bad day looks like a trend in isolation. What appears in a single sitting is repetition, which is the only signal available at this sample size: three near misses in the same place, four clients mentioning the same thing. Act on two of them and no more.
How does it survive a bad season?
By being short enough that a tired person still does it. Every record-keeping habit dies because it grew. Four lines, and an entry reading blown out, nothing to say counts as complete, because it preserves the habit. Guides who build elaborate systems in February have abandoned them by August; the survivors are always doing something almost trivially small. Add a fifth line in year three, never in week three.
Sources & methods
- 14 CFR 91.25 on the Electronic Code of Federal Regulations, read for the provision that the Administrator of the FAA will not use reports submitted to the National Aeronautics and Space Administration under the Aviation Safety Reporting Program, or information derived therefrom, in any enforcement action, except information concerning accidents or criminal offenses which are wholly excluded from the Program. The section governs the FAA's treatment of reports made under that programme and confers no protection on anybody outside it, in any context.
- 49 U.S.C. 44701 at the Office of the Law Revision Counsel, cited as the general regulatory authority under which the aviation safety rules are made. Nothing a guiding business writes about its own operations is privileged by anything described on this page, and a record so created may be discoverable depending on the circumstances and the law of the relevant state, neither of which was researched here.
- The 2024 annual edition of 14 CFR 91.25 published on govinfo, used as an independent copy of the provision quoted above. Nothing here is legal advice, and no industry figure for record-keeping practice, incident rates or near-miss frequency in guided fishing is asserted because no consulted source publishes one.
Every figure here is traced to a named public source and checked against it. Licensing, tax, and fee rules change. Verify your state’s current rules with the agency directly before you count on any number here.
Evan Knox
I build booking websites and run the ads and search for owner-run fishing guides, one operation per stretch of water. My first guide client, Bowman Fly Fishing, grew its revenue 4x in a year from that work. Field Notes is where I put the straight numbers on the business of guiding.
More field notes
Three seasons of your own river beats anybody's opinion.
I'm Evan. Data about your own water is publishable, and publishable is bookable. I build guides the booking site and run the ads that fill it. Free preview before you pay a cent.