Facebook lead ads vs website bookings

The longest standard attribution window Meta offers is seven days after a link click. Most other settings measure a single day. A guided fishing trip is frequently decided over months.

Which means the account cannot see most of what a guide's advertising produces, and the gap is not a flaw to be fixed but a structural mismatch between the measurement period and the buying period. It also decides the lead form question, because a lead form captures somebody inside the window while a website booking usually happens outside it. The account will therefore flatter one and undercount the other regardless of which actually fills more days. What follows is the machinery and what to do about it. Companion pages are at the getting-booked hub.

What are the attribution models?

Three, and one of them is asking a different question entirely.

Meta describes an attribution model chosen at ad set level which informs delivery and determines how conversions are credited, offering standard and incremental models.

Standard attribution, it says, optimises delivery for selected time windows and user behaviours, and lets an advertiser choose whether to credit conversions based on impressions, clicks or video plays.

Incremental attribution optimises delivery for incremental conversions, using models that predict whether a conversion was caused by an advertisement, and enables optimisation and reporting on that basis.

Custom attribution lets an advertiser share click-level data from an external analytics tool so that delivery optimises toward outcomes based on that logic.

The page sits at Meta's Business Help Center.

The delivery system these settings feed is described by the boosted posts piece.

Williamson, FL

Why is the incremental distinction so important?

Because it separates what happened from what was caused.

A standard model credits a conversion that occurred within a window after somebody saw or clicked an advertisement, which is a statement about sequence rather than about cause.

Meta's own description of incremental attribution names the difference directly, framing it as predicting whether a conversion was caused by an advertisement.

Which matters enormously to a guide, because a returning client who was always going to book will happily click an advertisement on the way to doing so.

That booking then appears in the account as a result the advertising produced, when the advertising produced a click on a decision already made.

The effect is largest for operations with strong retention, which is to say for the best-run ones.

Which is the specific mechanism by which a good operation can conclude its advertising is working better than it is.

How to measure retention separately is set out by the repeat rate piece.

South Texas Outcast, TX

What exactly do the settings count?

Days, and only ever a small number of them.

Meta lists the standard settings currently supported for website and in-store conversions.

Click-through counts events occurring within one day or seven days after a link click on the advertisement.

View-through counts events occurring within one day after an impression.

Engage-through counts events within one day after a non-link click action, including any clicks other than link clicks, and for video advertisements includes a video played for five seconds, or for ninety-seven per cent of its length where the video runs under five seconds.

It notes that some accounts may still use prior versions of click-through and engage-through attribution while the feature rolls out, and that not all settings are available for certain campaign types.

None of those windows approaches the period over which a destination trip is decided.

The evidence on that period is examined by the lead times piece.

So does the lead form win?

On visibility, always. On bookings, that is a separate question.

A lead form completes inside the platform, immediately, which places it comfortably within every attribution window on offer.

A website booking happens off the platform and frequently weeks later, which places it outside all of them.

Which produces an account in which one route appears to work and the other appears to do nothing, on the basis of when the event occurred rather than what it was worth.

The honest comparison therefore cannot be made inside the advertising account at all, and has to be made against the calendar.

Which requires knowing which bookings came from which route, and that requires a mechanism rather than the account's own reporting.

Building that mechanism is set out by the marketing report piece.

What does view-through credit actually claim?

That an impression alone moved somebody, within a day.

The view-through setting counts events occurring within one day after an impression, which credits conversions to advertisements that were never clicked.

Which is defensible in principle, since being reminded of something can prompt an action without a click, and indefensible in practice at small volumes.

The reason is that a guide's audience is small and local, so somebody who books today has probably seen an impression at some point regardless of whether it influenced anything.

Including view-through therefore inflates apparent results in exactly the situation where the inflation is hardest to detect.

For an operation trying to establish whether advertising works at all, excluding it produces a more conservative and more trustworthy number.

Which is a setting worth checking rather than accepting, since the default is not chosen with a small seasonal business in mind.

How does this affect two campaigns running together?

The later touch absorbs credit the earlier one earned.

An operation running a quiet winter presence and a concentrated pre-season campaign has two things acting on the same person months apart.

Whichever is nearer the booking falls inside a window and the other does not, so the account credits the second and reports nothing for the first.

Which will, run for a season, produce a confident conclusion that the winter spending achieved nothing and should be cut.

That conclusion is an artefact of the measurement period rather than a finding, and acting on it removes the part of the programme doing the slow work.

The protection is deciding in advance that the winter window will be judged on a different measure, and writing down what that measure is.

Enquiries mentioning you unprompted, and the share of new clients who say they had known about you for months, are both usable and neither appears in an account.

How to collect that second one is set out by the channel share piece.

What about the data the form collects?

A responsibility rather than an asset, and it should be treated as one.

A lead form hands you contact details for somebody who filled in a field on a phone, which creates obligations that vary considerably by jurisdiction.

Which is worth resolving before running one rather than afterwards, since the rules on what may be collected, stored and used differ and change.

Confirm the current position with a lawyer in your state before building a list from advertising, and before uploading any list back to a platform.

The practical hygiene is straightforward regardless: collect the minimum you need, say what it is for, and delete what you stop using.

An operation that cannot say where its contact list came from has a problem waiting rather than an asset.

Where the client record should sit is described by the spreadsheet CRM piece.

Is there a version that uses both?

Yes, and it is probably the right answer for most operations.

Run the website as the destination for the main campaign, where the page can work unattended and the booking is the outcome.

Run a lead form only for a narrow, time-limited purpose where speed is genuinely available, such as a specific week you are trying to fill and can respond to that day.

Which uses each instrument for what it is good at rather than choosing between them as a general policy.

The condition is that the two are measured separately and never compared on cost per contact, since the contacts are not the same thing.

Keeping them in separate campaigns makes that separation automatic rather than a matter of discipline.

Which is also what the platform's own structure implies, since a campaign carries a single objective.

What is a lead actually worth?

Whatever share of them book, which nobody measures.

A lead form produces a name and a contact detail from somebody who spent four seconds on a form, which is a materially weaker signal than somebody who navigated to a booking page.

Which means leads and enquiries are not comparable units, and a cost per lead cannot be set against a cost per website enquiry without adjusting for that.

The adjustment is measurable: record what proportion of leads from each route eventually book, over a season.

Most operations who do this find the conversion rates differ by a large multiple, which reverses the apparent cost comparison entirely.

Until that is measured, the cheaper unit is simply the cheaper unit rather than the better one.

What the record needs to hold is described by the spreadsheet CRM piece.

What does a lead form demand of you?

A response fast enough to matter.

Somebody who completed a form in a feed was not intending to start a conversation, and their interest decays quickly once the moment passes.

Which makes response time the variable that decides whether lead forms work at all for a given operation, more than any setting in the account.

An operation that cannot reply within a few hours during the season is buying contacts it will not convert, and that is a real operational constraint rather than a marketing one.

The same operation may do considerably better sending people to a page that answers the questions without needing a reply at all.

Which reframes the choice as a question about your own availability rather than about the platform.

What that reply should contain is set out by the enquiry replies piece.

Should the website be the destination instead?

Usually, for an operation whose page can answer the questions.

A page stating the price, the water, the length, what is included and what dates remain does the work a reply would otherwise do, at any hour.

Which suits a seasonal business whose operator is on the water precisely when enquiries arrive.

The cost is measurement, since the booking then happens off-platform and outside the attribution windows, which is where this article started.

That cost is worth paying because it is fixable with a mechanism, whereas an unanswered lead is not fixable at all.

The mechanism can be as simple as a booking page reachable only from the advertisement, which converts an invisible conversion into a countable one.

What that page has to contain is set out by the booking page piece.

Does the attribution setting change delivery?

Yes, and that is the part people miss.

Meta states that the attribution model informs ad delivery as well as determining how conversions are credited.

Which means the setting is not merely a reporting preference; it changes who the system tries to reach.

A short window pushes delivery toward people likely to act immediately, and a longer one toward people likely to act within it.

For a business whose clients decide slowly, that biases delivery toward the least representative part of the audience.

None of the available windows spans the period a guide actually cares about, so the practical move is to choose the longest available and accept the limitation knowingly.

The seasonal consequence is examined by the seasonal timing piece.

What should be measured instead?

Days sold, against a mechanism you built.

Every measurement problem in this article resolves the same way, which is that the account measures its own events and the business measures its calendar.

A dedicated page, a distinct number, or a question asked at booking and recorded, are the three available mechanisms and one of them is enough.

Which produces a figure the account cannot produce, being how many days were sold and what they cost to sell.

The account's own numbers then become diagnostics about delivery rather than claims about business outcomes, which is what they are.

Operations that make that separation stop arguing with their reporting and start reading it.

The whole record is described by the marketing report piece.

Does any of this apply to a small local operation?

Less than it applies to a destination one, and the reason is the decision period.

A local client booking on Wednesday for Saturday makes the whole decision inside a window the account can actually see, which makes the reporting broadly honest for that population.

Which is the one case where a click-through figure means roughly what it appears to mean, and where lead forms and website bookings can be compared with some confidence.

The distinction matters because published advice is written for businesses with short decision cycles and transfers badly to a destination fishery.

An operation serving both populations is running two different measurement problems at once, and averaging them produces a number describing neither.

Separating the two into different campaigns is therefore a measurement decision as much as a targeting one.

Which population you actually serve is answerable from your own booking records rather than from an assumption.

Reading that distribution is set out by the lead times piece.

Where does this decision go wrong?

Six ways, and comparing across the window is the first.

Comparing lead form results to website results inside an account that can only see one of them properly.

Treating a form completion and a booking page visit as the same unit of interest, when they are not.

Crediting a returning client's booking to the advertisement they clicked on the way to a decision already made.

Choosing lead forms without the ability to respond quickly enough for them to convert.

Sending advertising traffic to a page that cannot answer the questions, which forces a reply anyway.

And treating the attribution setting as a reporting preference when it also shapes delivery.

What the destination page must do is set out by the booking page piece.

What is the working choice?

Website, a page that answers everything, and a mechanism to count it.

Send traffic to a page that states the price, the water, the length, the inclusions and the remaining dates, because that page works while you are on the water.

Use lead forms only where you can genuinely reply within hours during the season, since an unanswered lead converts at nothing.

Choose the longest attribution window available and understand that it still falls short of how your clients decide.

Measure days sold through a dedicated page or a distinct number, and treat the account's figures as delivery diagnostics.

Record what proportion of contacts from each route eventually book, over a season, before concluding anything about cost.

Exclude returning clients from acquisition campaigns so their bookings stop being credited to advertising.

Deceptive practices in commerce are addressed at 15 U.S.C. 45, reproduced at govinfo.

The mistakes this produces are catalogued by the ad mistakes piece.