Facebook ad targeting for fishing guides

The audience number the platform shows while you build an advertisement is not a count of people. Meta says so directly: the estimate should not be interpreted as the number of accounts who will actually view your ads, and it is not designed to match population or census figures.

Which matters more in a fishing town than almost anywhere else, because one of the factors Meta names as making its estimate diverge from population is how many temporary visitors are in a particular geographic location at a given time. In a destination fishery that is not noise. It is either exactly the audience you want or exactly the one you do not, and the number cannot tell you which. What follows is what the figure is, what it is not, and how to target around it. The rest of this cluster sits under the getting-booked hub.

What is the number actually estimating?

Accounts meeting your criteria, on a definition with edges.

Meta describes estimated audience size as generally estimating how many accounts meet the targeting criteria an advertiser selects while creating an advertisement.

It says the estimate updates in real time during ad creation, considering factors including how many accounts were shown ads across its technologies in the past thirty days, self-reported demographics such as age and gender, and where accounts see ads.

It also lists the settings that shape the figure: location targeting, audience targeting including age and gender, language, custom audiences and detailed targeting, placement and device targeting, and brand suitability settings.

Which makes it a function of your own configuration as much as of any underlying population.

The page sits on Meta's help site.

The auction that number feeds into is described by the complete guide.

Panacea, FL

Why is the visitor factor decisive here?

Because a fishing destination is mostly visitors in season.

Among the reasons Meta gives for its estimates differing from population figures is how many temporary visitors are in a particular geographic location at a given time.

Which in an ordinary market is a small distortion and in a destination fishery during the season is a large part of the number.

The consequence cuts both ways and the direction depends entirely on your business.

An operation selling to travelling anglers wants precisely those accounts and should be delighted the figure includes them.

An operation selling to a local market is looking at a number substantially inflated by people who will be gone in a fortnight.

Neither can tell which from the figure, which is why the targeting has to be built from your own client records rather than from the estimate.

Where those records come from is covered by the channel share piece.

Reel Obsession, TX

What does an account mean?

Not one person, and the rule is stated precisely.

Meta notes that where somebody has added their Facebook, Instagram or WhatsApp profiles to the same account, those profiles are counted collectively as a single account for advertising estimation.

Where those profiles have not been linked, they are counted as multiple accounts.

Which means the same person may be one unit or several depending on a setting they chose years ago and have never revisited.

It also means comparing an audience figure between two regions is comparing populations with different linking behaviour as well as different sizes.

None of that makes the estimate useless; it makes it a measure of addressable accounts rather than of people, which is a different quantity with a different meaning.

The page also states plainly that the figure is not a proxy for the company's monthly or daily active user measures.

How platform figures should be read generally is set out by the methodology piece.

How should a guide build the audience?

From the two facts your own records already hold.

The first is how far your clients genuinely travel, which every trip record knows and almost no operator has looked at.

Setting a radius from that rather than from a guess is the single largest improvement available, because the wrong radius wastes the budget on people who were never going to drive that far.

The second is what your clients have in common besides fishing, which is frequently age, frequently a second interest, and occasionally an occupation.

Both are countable from a season of bookings and neither is available from any estimate the platform provides.

Which is the recurring theme: the platform supplies the machinery and your records supply the inputs, and the operators who do well are the ones with inputs.

Where those records live is described by the spreadsheet CRM piece.

Does the estimate predict reach?

No, and Meta says which two things do.

The page states that the number of accounts a campaign actually ends up reaching depends on the budget and on performance.

Which follows directly from the auction, since reaching somebody requires winning the auction for them and winning depends on total value rather than on eligibility.

An audience of forty thousand with a twenty dollar daily budget will reach a small fraction of it, and the fraction is decided by how well the advertisement competes.

The practical error this prevents is choosing a large audience because the number looks reassuring, when the budget was never going to address it.

A smaller, better-matched audience with the same budget concentrates the spending where the advertisement competes best.

What decides that competition is set out by the complete guide.

How narrow can it safely go?

Not so narrow that delivery has nothing to learn from.

There is a floor, and it is imposed by the delivery system rather than by anything about your offer.

Optimisation improves by observing outcomes, so an audience too small to produce many of them leaves the system with nothing to work from.

A few thousand accounts inside your travel radius gives it room; a few hundred generally does not.

Which pushes most operations towards a middle position: enough breadth for the machinery, with the filtering done by copy and image rather than by settings.

That arrangement also feeds the auction well, because a sharply written advertisement lifts exactly the components a small budget can influence.

Comparing the two approaches is a fair experiment provided only one variable moves at a time.

The seasonal timing of those tests is set out by the seasonal timing piece.

What does location targeting actually mean?

Several different things, and the default is rarely the one you want.

A radius around a point can mean people who live there, people currently there, or people who have recently been there, and those are materially different populations for a guide.

An operation selling to travelling anglers wants people who live elsewhere and are planning to come, which is not the same as people currently inside the radius.

An operation selling locally wants residents specifically, and including recent visitors wastes budget on people who have already gone home.

Which is the same visitor distortion that affects the audience estimate, appearing again as a targeting decision rather than as a measurement problem.

Getting it wrong is invisible in the reporting, since the advertisement delivers either way and the difference shows up only in whether anybody books.

Checking which setting is actually applied is a two-minute audit that a surprising number of accounts fail.

What the records say about travel distance is covered by the channel share piece.

Does the placement setting matter?

Meta lists it as shaping the estimate, and it shapes the result more.

Placement and device targeting appear among the settings the platform names as determining the audience figure, which means restricting placements narrows the pool before anything else happens.

For a small budget that is usually the wrong direction, since restricting where an advertisement can appear removes cheap opportunities without improving relevance.

The instinct to exclude placements comes from a belief that some surfaces produce worse traffic, which may be true and is testable rather than assumable.

What is not in doubt is that a restricted account competes in fewer auctions on the same budget, which is a real cost.

The defensible reason to restrict is a creative that genuinely does not work in a given format, which is a production problem rather than a targeting one.

Fixing the creative is usually cheaper than narrowing the pool.

How many audiences should run at once?

One, for anybody at guide scale.

Splitting a small budget across several audiences divides the volume each has to learn from, which is the same problem as narrowing too far arriving by a different route.

Two audiences on twenty dollars a day are two underfunded tests rather than one useful one, and neither will produce a readable result.

Which means comparison has to happen sequentially across seasons rather than simultaneously within one, and that is a slower way to learn.

It is also the only honest way at this scale, and accepting it prevents a year of parallel tests that all conclude nothing.

Where an operation genuinely has two distinct products, those belong in separate campaigns with separate budgets rather than as audiences inside one.

The campaign structure that follows is covered by the ad cost piece.

What about detailed interest targeting?

Useful, and it describes stated interests rather than behaviour.

Detailed targeting is one of the settings Meta lists as shaping the estimate, and the underlying signals include self-reported demographics.

Which means an interest category captures people who signalled an interest, not necessarily people who act on it, and the two overlap imperfectly.

For fishing that distinction is real: a great many people interested in fishing content have never booked a guided day and never will.

The interest that predicts a booking is usually not fishing but travel, since the clients who spend on a guided day are frequently the ones who spend on trips generally.

Testing that against the obvious targeting is worth a season, and it is the kind of test almost nobody runs because the obvious option feels correct.

Your own client records will suggest which to try first.

The record that supplies the suggestion is described by the debrief piece.

Should past clients be targeted?

Only deliberately, and usually not with acquisition money.

A custom audience built from your own client list is available and it is the most responsive audience you will ever build.

Which is exactly why spending acquisition budget on it is frequently a mistake, since those people can be reached by a message that costs nothing.

The legitimate uses are narrow: reminding a lapsed group that a season is opening, or reaching people whose contact details you no longer have.

The other legitimate use is exclusion, keeping existing clients out of an acquisition campaign so the budget addresses people who do not know you.

Which is the setting most guides never touch and which quietly improves every campaign it is applied to.

Rules on using client data for advertising vary by jurisdiction; confirm the current position with a lawyer in your state before uploading anything.

The free version of reaching them is set out by the text scripts piece.

What does the estimate move with?

Your settings, and time, and neither means the market changed.

Meta notes that estimates may vary over time, and that audience expansion features mean the figure will not reflect the total number of accounts meeting the criteria.

Which means a figure that shifted between two sittings tells you nothing about your market and should not be read as though it did.

The number is a planning aid for whether a configuration is plausibly sized, and it is not a measurement of anything.

Using it as a measurement is the specific error, and it shows up as operators tracking audience size across weeks as though it were a market indicator.

What is worth tracking instead is the result, which is bookings, measured through a mechanism you built.

How to build that mechanism is set out by the marketing report piece.

What should be excluded?

Three groups, and excluding them costs nothing.

Existing clients, because they can be reached free and their inclusion inflates results that look like acquisition and are not.

People who already enquired and were answered, since an advertisement reaching them again is spending to repeat a conversation you already had.

And, for a destination operation, the immediate local area, where the audience is largest and the booking intent is lowest.

That last exclusion is counterintuitive and it is frequently the largest single improvement available to an operation in a tourist market.

The budget stops being consumed by people who live twenty minutes away and have never considered paying for a guide on water they fish for nothing.

Testing it is straightforward and the result is usually decisive within a season.

What does a good audience look like written down?

Four lines, and it should fit on an index card.

The geography, stated as a named area and a rule about residents or visitors rather than as a radius alone.

The age band, taken from your own client records rather than from an assumption about who fishes.

The interest or behaviour you are testing this season, named specifically so it can be judged at the end.

And the exclusions, listed explicitly so they survive somebody rebuilding the campaign next year.

Written down, that definition can be reproduced, compared and improved; held in an account interface, it will be forgotten and rebuilt differently.

Which is the same discipline that runs through every record in this trade, arriving at a marketing account instead of a spreadsheet.

Where does targeting go wrong?

Six ways, and treating the estimate as a population is the first.

Reading the audience figure as a count of people, when it counts accounts and is not designed to match population sources.

Ignoring the visitor factor in a destination market, where it can be a large part of the number.

Setting a radius from instinct rather than from how far clients demonstrably travel.

Choosing a large audience because the number is reassuring, when the budget was never going to address it.

Narrowing below the volume the delivery system needs, which starves optimisation.

And spending acquisition budget on past clients who could be reached by a message for nothing.

The budget arithmetic underneath is set out by the ad cost piece.

What is the working approach?

Radius from records, breadth from the system, specificity from the creative.

Set the geography from how far your clients actually travel, which your trip records already know.

Treat the audience estimate as a plausibility check on the configuration rather than as a measurement of a market.

Keep the audience broad enough for delivery to optimise and let an extremely specific advertisement do the filtering.

Test travel interest against fishing interest, since the second is crowded with people who will never book.

Exclude existing clients from acquisition campaigns, and reach them by message instead.

Track bookings rather than audience size, because the estimate moves for reasons that have nothing to do with your market.

The general prohibition on deceptive practices is codified at 15 U.S.C. 45, with a parallel text at govinfo.

What the advertisement itself should say is set out by the ad examples piece.