Ad creatives and testing

What makes a good Facebook ad headline for lead generation?

A good lead gen headline names one job, one place and one next step, in the words the ad hook used, tested after the concept and the hook.

Published

A good Facebook ad headline for lead generation names one job, one place and one next step, in the same words the ad’s hook used. It is the line the visitor checks to confirm they are in the right place, so its job is recognition rather than persuasion: the creative makes the case, and the headline confirms it.

That also sets its place in the work. A headline is a copy detail, which means it comes after the concept and the hook and before the button colour. Getting the order wrong is how an account ends up with fifteen polished headlines and one weak idea behind all of them.

What the headline has to do

Three jobs, in this order:

  1. Repeat the promise of the ad closely enough that the visitor recognises it.
  2. Say who the offer is for, in the same breath, so the wrong people leave early.
  3. Give the next step, which is usually the form or the call.

Meta’s own testing advice moves from the broad question to the narrow one: test different creatives within the format that won. A headline sits at the narrow end, which is why it is worth testing only once the concept earns budget. A Meta ads creative testing framework for lead gen puts concepts and hooks first for a statistical reason: at lead gen volumes you can only detect large differences, and large differences live in concepts and hooks, not in wording tweaks.

The second job is where most headlines fail. “Quality painting you can trust” repeats nothing and says nothing about who the offer is for. “Exterior repaint in Reading, booked within three weeks” names the job, the town and the timing, and someone outside Reading knows immediately that it is not for them.

Facebook ad headline examples, weak and strong

The weak column is not wrong, it is just unhelpful: the reader has to construct the connection themselves. The strong column restates the promise the ad already made.

SituationWeak headlineStrong headline
Solar installerWelcome to SunCo EnergySolar panels with €0 upfront, installed in Lombardy
Dental clinicSmile with confidenceBook your free 30-minute Invisalign consultation in Milan
Accounting firmAccounting services for small businessesPay your accountant for advice, not data entry
RoofingWe are your local roofing expertsRoof replacement in Plano, installed within three weeks of sign-off
Home insulationSave money on energyCheck if you qualify for the insulation grant
GymJoin a gym todayYour first class in Monfalcone, this week

Three patterns run through the strong column. The subject of the sentence is the service or the appointment, never the reader. The specifics are checkable: a place, a timeframe, a price condition. And the headline sounds like it came from the same ad, because the same phrase appears in the creative.

The roof and insulation rows also show why a headline is a targeting tool. Naming the town filters people before the click, and naming a condition such as a grant check filters them again. On a broad audience the creative does most of the targeting, which in practice means the headline does.

The rules a headline can break

A punchy headline is not worth a rejected ad. Two policies reach lead gen headlines most often.

Personal attributes. Meta does not allow an ad to state or imply something about the viewer, directly or indirectly: health, disability, financial trouble, age, religion, gender identity, criminal record. The words “you” and “your” are fine as long as no attribute follows them. So “Behind on your mortgage payments?” becomes “Mortgage arrears advice from a regulated adviser”, and the sentence keeps the same offer with the subject moved off the reader.

Unrealistic outcomes. Approval presented as certain, income presented as certain, a specific result in a set time for a health or cosmetic offer, and investment returns presented as safe or fast all sit on the wrong side of Meta’s rules. A headline that says “approved in 24 hours, no checks” buys a rejection, and the same space works better with the condition stated: approval depends on eligibility checks.

A third rule is less obvious. The landing page is part of the review, and the offer in the ad has to match the offer on the landing page. A headline promising one thing above a landing page selling another is a rejection risk and a conversion problem at the same time.

How to test headlines without wasting budget

Write the decision rules before launch, then follow them.

StageSpend per creativeWhat you get
Screen3x target cost per leadwhether the concept deserves to continue
Rank10x target cost per leadan order between concepts that are still alive
Confirm25 to 30x target cost per leada winner you can move budget to

Below about ten leads per creative, no verdict. At a €30 target cost per lead, ten leads cost €300 and are still mostly noise, which is why the multiples matter more than the days.

  • Change one variable per variant. A new headline with a new image tests a combination, not a headline.
  • Keep the offer fixed while you test wording, then test the offer.
  • Put the variants in a setup that gives each one spend: a testing ad set that can reach about 50 leads a week, or Meta’s own A/B test, which shows each version to a separate segment of the audience.
  • Re-test the winner on a second landing page. A headline that wins on one landing page can lose on another, and the pair is the unit you scale.

Meta’s A/B test recommends a minimum of seven days and a maximum of 30, with the same budget for each version, and it advises against testing informally by switching ad sets on and off. CBO vs ABO explains why a test needs ad set budgets while scaling belongs on a campaign budget.

Where the headline sits in the account

One concept per ad, with the headline inside it. If several concepts share a single ad as alternative assets, every lead arrives with the same ad ID and you cannot split cost per lead by concept afterwards.

The naming and the log do the rest. An ad name that carries the concept, the hook, the format and the landing page makes the report readable a month later, and a creative log with one row per ad ID turns a series of tests into lessons about which concepts win. Facebook ads account structure has the naming convention and the URL parameters that attach the ad ID to every lead, and lead generation landing page examples shows what the landing page behind each headline should look like.

Two habits keep a winning headline alive. Refresh the creative around it before fatigue sets in, rather than after cost per lead has doubled, and keep the headline fixed while you replace the image or the video, so you learn which element actually changed. Creative fatigue in Meta ads measures the drop per ad and landing page, which is the signal to write new hooks rather than new colours.

Finally, judge headlines on the leads they bring, not on the clicks. A headline that names a price condition attracts fewer clicks and better leads, and the only place that shows up is cost per qualified lead by creative and landing page.

Questions on this topic

All questions on this topic →

FAQ

How long should a Facebook ad headline be?+

Short enough to read in one glance on a phone, which in practice means the headline carries one offer rather than two. The wording matters more than the count: if the reader has to work out how the headline connects to the ad they tapped, it is too long whatever its length.

Should the headline repeat the ad text word for word?+

Close to it. The headline is the line the visitor compares against the ad and then against the landing page headline, so the same promise in near-identical words confirms they are in the right place. A different promise in each place makes the visitor do the work of connecting them.

How many headline variants should I test?+

Three to six distinct concepts in an ad set, each with its own headline, is the practical range. Ten near-identical headlines split the budget so thin that none of them ranks, and at lead gen volumes a gap between two similar headlines is more likely to be noise than a result.

When can I call a headline a winner?+

When it has enough leads to judge, not when it has the lowest cost per lead after two days. Screen a new concept with about three times your target cost per lead, rank it at around ten times, and confirm a winner at 25 to 30 times, then check the winner on a second landing page.

Ready to find out what
converts, for real?

Free to start.

Start free →