Why Your Facebook Ads Send Low Quality Leads

If your Facebook leads don't answer the phone, ask for the cheapest price before you've quoted, or live outside your service area, the platform usually isn't the problem. Meta delivers what the campaign asks for: if the form takes two taps and the optimisation event is 'a form was submitted', it finds people likely to submit forms, whether or not they want to hire anyone.

Work through the seven steps below in order. The first two change the leads themselves; the rest change who sees the ad, what Meta learns and how many good leads you actually convert.

When you build an Instant Form, Meta asks for a form type. 'More volume' is built to be quick to fill out and submit on mobile, with details pre-filled from the person's profile. 'Higher intent' pre-fills the same way but adds a review screen where the person checks their details and confirms before submitting. (A third type, 'Rich creative', adds landing-page-style sections above the questions.) Qualifying questions are a separate choice and work with any form type.

The review step is the smallest change with the clearest mechanism, so start here. Two things to expect: fewer leads at a higher cost per lead, and possibly fewer placements, since Meta limits where some form types can run (secondary reports say Higher intent runs only in the Facebook and Instagram mobile feeds), so check the placement list after switching. What you're testing is whether more of the leads pick up and book.

A name-and-phone form asks nothing of the person filling it in. A couple of questions that take a moment to answer filter out the least interested without making the form a chore. Use multiple-choice answers where you can: they're quick for genuine prospects and easy to sort in your CRM.

Questions that work for contractor lead forms

• Timeline: Within 30 days / 1–3 months / 3–6 months / Just researching. Route 'just researching' to a slower nurture sequence rather than a sales call.
• Service address or ZIP code: catches out-of-area leads immediately. Asking on the form is allowed even where targeting by ZIP isn't.
• Which service: a drop-down of the jobs you actually do, so wrong-service enquiries are obvious before anyone calls.
• Do you own the home?: Meta no longer offers homeowner status as targeting, so the form is the place to ask.
• Approximate budget (only for large projects): wide bands set expectations; skip it for repairs, where it just adds friction.

An audience wider than the area you serve guarantees leads you can't take. Set the location to your actual service area. Meta counts housing repairs as housing ads, so many home repair and improvement ads fall under its Housing special ad category, and ads built around financing can fall under Financial products and services. In those categories you can't target by age, gender or ZIP code, the minimum radius is 15 miles and lookalike audiences aren't available.

So in those categories, targeting can only get you to the right area. The filtering of who is a real prospect has to happen in the creative and on the form (Steps 1, 2 and 4). A past-customer Custom Audience works in any category; lookalikes built from it are an option only outside the special categories.

The ad is the first filter. An ad that promises the lowest price in town will attract people shopping on price; an ad that shows the work, names the area and sets expectations attracts people who are already closer to hiring.

• Show real jobs: your own before-and-after photos or short phone video of the crew and the owner, not stock images.
• Name the service area and the specific service, so people outside it or looking for something else scroll past.
• Set expectations: if your work is premium, say so; if there's a typical price range or a minimum job size you're comfortable publishing, including it screens out mismatched budgets.
• Only make claims you can back up. Savings, 'best in town' or review-count claims need evidence, and specific financing terms bring Truth in Lending disclosure rules.
• Avoid a bare 'Free quote!' as the whole message: it gives a reason to click and nothing else.

By default a Leads campaign optimises for form submissions, so Meta learns to find people who submit forms. The fix is to feed it what happened next.

For Instant Form leads: CRM stages and the conversion leads goal

Meta's Conversions API for CRM lets you send lead stages (for example contacted, qualified, booked) back to Meta, matched by the lead ID from the original form. With that data flowing, a campaign can use the conversion leads performance goal, which optimises toward leads likely to reach your chosen stage rather than toward any submission. Some CRMs, HighLevel among them, document an integration for sending these events; check whether yours does.

The catch for small contractors is volume. Meta's developer documentation sets criteria for this goal; as summarised by LeadsBridge, they include roughly 200 leads a month (some reports say 250), stage updates sent at least daily, and a target stage that 1% to 40% of leads reach within 28 days. At WordStream/LocalIQ's 2026 Home & Home Improvement average of $42.95 per Facebook lead, 200 leads a month is about $8,600 in ad spend, well above many single-location budgets. Sending the stages is still worth setting up: it costs little once the CRM integration is on, and it's ready when volume grows.

For website leads: Pixel plus Conversions API on a real Lead event

If ads send people to your site, make sure the Lead event fires only on a completed form or booking, not on a page view or a button click, and send it through the Conversions API as well as the Pixel so conversions the browser misses can still reach Meta. Put UTM parameters on every ad URL so your CRM records which campaign each lead came from.

In research published in Harvard Business Review in 2011 (1.25 million leads at 42 US companies), firms that tried to contact a web lead within an hour were nearly 7 times as likely to qualify it as those that waited longer. The MIT and InsideSales.com lead response study found the odds of contacting a lead fell sharply after the first five minutes. A lead that 'never answered' may have been a good lead called too late, or after they'd booked someone else.

• An automatic text the moment the lead arrives: 'Hi [Name], it's [Owner] at [Company]. Got your request about [service]; I'll call you in the next few minutes.'
• A real call within minutes during business hours, with the lead's form answers in front of whoever calls.
• A short follow-up sequence (for example, texts at one hour, the next day and later that week) for leads who don't answer.

Spend 15 minutes a week on the last seven days of leads and mark each one: in the area or not, right service or not, reached or not, booked or not. The pattern tells you which step is leaking. The thresholds below are starting points to compare against, not published benchmarks; your own history matters more.

Scroll sideways to see the full table →

The last row is the one that decides whether the ads pay. A higher cost per lead is fine if booked jobs cost less; a lower cost per lead is no win if nobody books.