Most businesses complaining about the quality of their Facebook leads have never touched their form. They kept the one Meta suggests by default: name, email, phone number, three fields filled in automatically, submitted in two seconds from a sofa.
That form does exactly what it was designed to do, which is maximise the number of submissions. The problem is that this is almost never the goal of a business selling jobs worth several thousand euros.
The form is the only place in your setup where you decide who gets through and who does not. It is also the cheapest thing to change: a few minutes in the ads manager, no extra budget.
More volume or higher intent: a choice, not a setting
Meta offers two form types, and this single choice matters more than all the others put together.
The more volume type validates the submission in one tap. The prospect sees their details already filled in, presses the button, and it is done. They have read nothing, thought about nothing, and often forget they contacted anyone within the hour.
The higher intent type adds a confirmation screen: the prospect has to review their details and confirm a second time. That one extra gesture removes accidental submissions and the ones made without thinking.
Switching over mechanically lowers the number of contacts you receive. That is the point. A form that brings in forty enquiries of which three are workable costs more in call-back time than a form that brings in twelve of which six are real.
Qualifying questions, and the order to ask them
Meta lets you add custom questions. This is where the quality of what you receive is decided.
Ask closed questions, not open ones. A free text field saying "describe your project" gets filled in with "quote" eight times out of ten. A list of options forces the prospect to place themselves, and hands you usable information before you even pick up the phone.
Only list the services you actually want to sell. This is the most effective filter, and the most neglected. If your list includes an "other" option or a "small jobs" option, you have just reopened the door you were trying to close. Anyone who does not recognise themselves in any of the boxes drops out, and that is precisely the result you wanted.
Ask about timing. "When are you looking to start?" with three options: as soon as possible, within three months, later. This question scares nobody off and completely changes how you handle the contact. Someone answering "later" is not waste, they go into a follow-up sequence instead of eating into your calling time.
Budget, if your average job justifies it. The question makes people nervous, yet it is the one that sorts best above a few thousand euros. Phrased as ranges rather than an exact figure, it goes down well. A prospect who refuses every range usually has no project yet, only curiosity.
Order them from easiest to most committing. Project type first, timing second, budget last. A prospect who has already answered two questions is far less likely to abandon at the third than if they had run into it straight away.
Conditional questions
Meta lets you show a question depending on the previous answer. It is underused and very effective.
An installer can show the budget question only if the prospect has ticked a major job type. Someone coming in for a small service never sees that question, so they are not discouraged for nothing, and you keep a contact you will simply handle differently.
The idea is not to apply the same level of scrutiny to everyone, but to raise the bar on the projects that deserve it.
What happens before the form
Two elements decide whether the prospect opens the form in the right frame of mind.
The form cover image. It is separate from the ad creative, and plenty of people leave it empty. It is the last image the prospect sees before handing over their details: it should show the concrete result, not your logo.
The headline and the intro paragraph. This is where you announce what is about to happen. "Get your free quote" says nothing. "Describe your project in three questions and I will call you back within 24 hours to arrange a visit" sets the frame, commits you, and pushes away anyone who only wanted a price by email.
The ending screen, the part everyone forgets
The thank-you message can be edited, and it is free selling space.
Use it for three things: restate what happens next and when, add a direct call button for people in a hurry, and give a reason to wait for your call rather than carry on searching. A prospect who knows they will be called back today shops around far less actively.
How many fields, really
The rule you hear everywhere, "fewer fields, more conversions", is true and that is exactly what makes it dangerous. It optimises the wrong number.
Every field you add lowers the completion rate. That is not up for debate, it is mechanical. The real question is what each drop-off costs you. If you sell a 30 euro product, every drop-off is a straight loss and you should indeed cut to the bone. If you sell a 25,000 euro job, one drop-off in two saves you a pointless call-back.
In practice, three to five fields in total, contact details included, is the zone that works for a high-value service. Below three, you qualify nothing. Above six, you also lose serious prospects, who are often the ones in the biggest hurry.
One detail matters: the fields Meta pre-fills, name, email and phone number, cost almost nothing in friction since they are already filled in. Your custom questions, on the other hand, demand real effort. Count them separately: two custom questions maximum on a first form, three if your average job genuinely justifies it.
The phone number, a special case
For this kind of service, the phone is the only channel that converts. A lead without a usable number is a lost lead, however good it looks otherwise.
Meta pre-fills that field with the number on the account, which is often old, sometimes wrong, sometimes a landline nobody uses any more. You end up with details that look valid but are unreachable.
Two settings limit the damage. Make the field editable, and explicitly ask, in the intro, for a number where the person can be reached during the day. And if your volume allows it, add a question about the preferred call window: it noticeably improves your pick-up rate, because you call when the person expects to be called.
Getting the leads out, where everything falls apart
This is the most expensive mistake, and it has nothing to do with the form itself.
By default, enquiries pile up inside Meta's lead manager. Nobody goes to fetch them. A business owner on a job site does not log into Business Manager three times a day, and Meta does not send a warning. Companies pay for campaigns for weeks with dozens of contacts sleeping in an interface nobody opens.
Three ways to fix it, in order of robustness: a direct connection to a CRM or a spreadsheet, an automatic email or SMS notification for every new enquiry, or failing that a daily reminder to go and download the file by hand.
The test is simple: between the moment someone submits the form and the moment a human being can call them back, how much time passes and who gets told? If the answer contains "I have to go and check", the setup is broken, and no form setting will make up for it.
Testing two forms without drawing the wrong conclusion
A form cannot be judged on one week or on ten enquiries.
The right protocol is straightforward: run two versions in parallel on the same campaign, not one after the other. Periods are never comparable, a week of school holidays or a change in the weather is enough to distort a sequential comparison.
Let it run until you have at least thirty or so enquiries per version, which often takes two to three weeks on a local budget. Below that, the gap you are looking at is noise.
And above all, compare the right number. Not the number of enquiries, not the cost per enquiry, but the number of appointments actually held against the spend. A form that produces half as many contacts and the same number of appointments is twice as good, even if Meta's manager shows a cost per lead twice as high.
What to measure next
A well-configured form pushes two numbers down and one number up.
The number of enquiries drops, often by half. The cost per enquiry rises mechanically, and that is normal: you are paying more for a contact worth a great deal more. The only number that really counts is the cost of an appointment actually held, not the cost of a form being filled in.
If you only track cost per lead in Meta's ads manager, you will conclude that the simple form was better. It was not, it was just cheaper to fill in.
Where to start
If you change only one thing this week: switch your form to higher intent and add one closed question about project type, with no "other" option. Those two settings take ten minutes and are enough to filter out most of the enquiries that waste your time.
The rest, conditional questions, budget, ending screen, comes afterwards, once you can see what the enquiries getting through the first filter actually look like.