Guide Income by Trips Per Year: The Model

Federal price analysis has a warning attached to comparison. The prior price must be a valid basis for comparison. If there has been a significant time lapse, or the terms and conditions differ significantly, or the reasonableness of the prior price is itself uncertain, the comparison does not do the work.

Which is the whole problem with asking what guides earn. Every available comparison fails at least one of those three tests, and usually all three. What follows is not a survey and does not contain one. It is a model, built openly from stated inputs, that produces a number for your own operation and refuses to produce one for anybody else's. Nothing here is a benchmark and no industry figure is asserted. Everything adjacent lives under the guide industry data hub.

What does the analysis rule say about comparison?

That a prior price has to earn its status as a benchmark.

Section 15.404-1(b)(2)(ii) of Title 48 lists comparison of proposed prices to historical prices paid, whether by the government or otherwise, for the same or similar items, among the recognised price analysis techniques.

It then attaches the qualification directly: the prior price must be a valid basis for comparison, and it is not where there has been a significant time lapse, where the terms and conditions differ significantly, or where the reasonableness of the prior price is itself uncertain.

Paragraph (a) frames the whole exercise, stating that the objective of proposal analysis is to ensure that the final agreed price is fair and reasonable, and that the complexity and circumstances of each acquisition should determine the level of detail required.

Paragraph (b)(1) defines price analysis as examining a proposed price without evaluating its separate cost elements and proposed profit, which is precisely what a benchmark comparison does.

Section 15.404-1 is carried on the eCFR.

The reason no benchmark exists to compare against is unpacked in the methodology piece.

SC Guide Service, CA

Why does every guide comparison fail those tests?

Time, terms and provenance, in that order of severity.

Time, because any figure circulating about guide income predates the cost movements of recent years and is quoted without a date.

Terms, because a saltwater charter carrying six, a technical trout day for one, and a lodge week are not the same item and are not similar items in any sense the rule would accept.

And provenance, because the reasonableness of the prior figure is not merely uncertain but usually unknowable, since its trail terminates in another article.

Which is three failures against a test that requires none, and it is why this page contains a model rather than a number.

The model has the property that a benchmark lacks, being that its inputs are visible and can be replaced with your own.

What replaces them is your own records rather than anybody's survey.

Where those records come from is set out by the debrief piece.

All Out, MO

Which input matters most?

Trip days, by a distance, and it is the one guides misremember.

The output moves almost linearly with days sold, and days sold is the input people are most wrong about, because a season feels longer than it was.

Counting the calendar rather than estimating produces a number that is frequently ten to twenty per cent below the guide's impression, and the correction changes every conclusion downstream.

Which is worth doing before anything else, because a model run on a remembered day count is a model with a systematic error at its largest term.

The second most important input is realised rate rather than list rate, since discounts, group pricing and shortened days all reduce the former without touching the latter.

Realised rate comes from dividing what actually arrived by the number of days, which takes two minutes and is almost never done.

Most operations find the gap between list and realised is larger than expected, and the gap is where a season's margin quietly goes.

What that gap does to pricing is set out by the margin piece.

What does the model leave out?

Four things, and all four matter.

Tax, which is substantial for a self-employed person and varies with entity, state and circumstances that no model can carry.

Equipment replacement, which is a real annual cost even in years nothing is bought, and which the fixed-cost line above deliberately excludes to avoid pretending otherwise.

The owner's own labour, which is the whole thing being measured and which has no cost line, so the output is a return to labour and capital combined rather than a wage.

And any income from other sources, which for a great many guides is the difference between a viable year and a difficult one.

None of those omissions is an oversight; each is a place where a general model has to stop and a particular business has to take over.

Which is the honest limit of any model of this kind and it should be stated rather than hidden.

The replacement question is examined by the gear resale piece.

How sensitive is it?

Enormously to days, moderately to rate, barely to anything else.

On the stated inputs, ten additional trip days adds roughly seven thousand two hundred dollars before costs and about six thousand three hundred after direct cost, which is a large movement for a small operational change.

A forty dollar increase in realised rate across the same ninety-five days adds about three thousand eight hundred, plus a little on gratuities.

A ten dollar reduction in direct cost per day adds nine hundred and fifty, which is real and is an order of magnitude smaller.

Which sorts the operational priorities directly: sell more days, then raise the realised rate, then economise, in that order.

It also explains why cost-cutting feels unrewarding in this trade, since the cost base is small relative to the revenue and there is little to cut.

All of which follows from the structure rather than from the specific numbers, and it holds when the numbers are replaced.

The extra days have to come from somewhere, which is the subject of the channel share piece.

What is a realistic day count?

A question about your own water, and nobody else can answer it.

The ceiling on trip days is set by the length of the fishable season, the weather, the demand and the guide's own capacity, and those four differ so much between operations that a general figure would be meaningless.

Which is why this page does not offer one, and why any figure offered elsewhere should be checked against the three questions in the methodology piece.

What is generalisable is the shape: almost every operation has more capacity than demand in the shoulders and more demand than capacity in the peak.

Which means the achievable day count is usually limited by the shoulder rather than by the peak, and the marginal day is a Tuesday in October rather than a Saturday in July.

Selling that Tuesday is a completely different marketing problem from selling the Saturday, and most operations work only on the second.

What the shoulder day is worth is exactly the sensitivity above.

The shape of that calendar is examined in the lead times piece.

Should gratuities be in the model at all?

Yes, and separately, because they behave differently.

Gratuities are a material share of guiding income and excluding them understates the position substantially, which is why the line exists.

Separating them matters because they do not scale with rate in the way the model implies, and because they are the input most operators have never measured.

Which makes the gratuity line the least reliable number in the whole calculation and the one most worth replacing with a real figure.

Replacing it requires a season of recording, which is the same conclusion every page in this cluster reaches.

Where no record exists, running the model with the line at zero produces a defensible floor, which is more useful than a guess.

A floor you trust beats a total you do not.

Producing that real figure takes one season, as the tips piece describes.

What about a second boat?

A different model, not a doubled one.

Adding a second guide does not double the output, because the second boat carries its own direct costs and the guide takes a substantial share of the revenue.

What it does add is capacity in the peak, which is exactly where demand exceeds capacity, so the incremental days are the highest-value ones.

Against that, the fixed cost line grows, the administrative load grows, and the operation acquires a set of obligations it did not previously have.

Whether any of it is worth doing is answerable only with your own five inputs and a realistic estimate of the second boat's day count, which is usually lower than the first's in year one.

Modelling it as a separate calculation rather than as a multiplier is the only version that produces a usable answer.

Whether the operation is anywhere near that point is argued in the capacity signal piece.

What does the output actually mean?

A number to compare against alternatives, not a verdict on the business.

The figure that falls out is a return to the owner's labour and capital together, before tax, before replacement and before anything is set aside.

Which makes it comparable to what the same person could earn doing something else, plus what the same capital could earn elsewhere, and that is the comparison worth making.

Most guides never make it explicitly, and the ones who do frequently find the arithmetic uncomfortable and the conclusion unchanged, because the trade is not chosen on that basis.

Which is a legitimate position and it is better held knowingly than by accident.

The version that causes damage is neither comparison nor acceptance but avoidance, where an operator runs for a decade without ever computing the number.

Computing it once removes the anxiety in either direction, since a known figure can be acted on and an unknown one only worried about.

Where the alternatives sit is a personal question this page cannot help with.

Does the model work for a part-time operation?

Better, and the fixed cost line is the reason.

A guide running twenty days a year carries almost the same insurance, permits and storage as one running ninety, which means fixed cost per day is four to five times higher.

Which the model shows immediately and which part-time operators frequently have not registered, because the annual figures are small enough to feel manageable.

On the same fixed cost of fourteen thousand, twenty days carries seven hundred dollars of fixed cost per day and ninety-five days carries about a hundred and forty-seven.

That difference is larger than any variable cost in the business, and it is invisible without running the calculation.

The practical consequences are real: a part-time operation either needs a materially higher rate, a genuinely lower fixed base, or an acceptance that it is a supplemented hobby.

All three are defensible and only the third is defensible unknowingly.

What the fixed base actually consists of is set out by the insurance costs piece.

What should be modelled before a change?

The change alone, not the whole business.

The temptation before buying a boat or adding a fishery is to rebuild the entire model with new assumptions, which produces a number nobody can trace.

What answers the question is the incremental calculation: what does this change add in days and revenue, and what does it add in fixed and direct cost.

Which is a much smaller and much more honest exercise, and it isolates the decision from every other assumption in the picture.

An incremental model with three lines is more useful than a rebuilt one with fifteen, because its errors are visible.

The specific figure worth being pessimistic about is the additional days, since new capacity almost never fills at the rate the existing capacity does.

Halving the first-year estimate is a crude correction and it is closer to observed outcomes than the original number usually is.

The purchase side of it is set out by the gear resale piece.

How often should it be run?

Once a year, in November, with last season's actuals.

A model run on projections is a planning exercise and a model run on actuals is a measurement, and only the second improves anything.

Which means the November sitting is where this belongs, alongside the cost-per-day work that supplies one of its inputs.

Running it three years in a row is where the value appears, since the level tells you little and the direction tells you a great deal.

Keeping the format identical between years is what makes the comparison possible, which is the same discipline the whole cluster keeps returning to.

Changing the model to improve it each year is how three years of data become three unrelated calculations.

Where that sitting fits is set out by the winterising piece.

Where do income models go wrong?

Six ways, and the remembered day count is the first.

Estimating days sold rather than counting them, which introduces a systematic error at the largest term.

Using list rate rather than realised rate, so discounts and shortened days disappear from the calculation.

Omitting gratuities entirely, which understates a material share of the income.

Treating the output as a wage, when it is a return to labour and capital with no tax deducted.

Quoting somebody else's inputs, which is exactly the comparison the analysis rule warns against.

And changing the model between years, so the only genuinely valuable output disappears.

What the surrounding numbers should be is set out by the numbers piece.

What is the working model?

Five inputs, all yours, run in November, kept unchanged.

Count trip days from the calendar rather than estimating, because that term dominates the output.

Compute realised rate by dividing what arrived by the days, not by reading the price list.

Include gratuities as a separate line, from a record if you have one and at zero if you do not.

Take direct cost per day from receipts rather than from impression, and keep fixed annual costs separate.

Read the output as a return to labour and capital before tax and before equipment replacement, and not as a wage.

Run it once a year on actuals, keep the format identical, and compare only against your own prior years.

The statutory basis for the acquisition rules is 41 U.S.C. 1303, with the section mirrored on govinfo.

Why no published figure will do the job is set out by the methodology piece.