Lookalikes

Meta says a lookalike suggestion gets 11% lower cost per quality lead. Should my B2B Leads campaign add one?

By The Tofu team ·

Add one as a test in a single ad set, if you can seed it from qualified leads or closed-won customers. Jon Loomer reported on November 2, 2026 that Meta now shows a prompt in Leads campaigns: "You could get 11% lower cost per quality lead by using a lookalike." With Advantage+ audience on, a lookalike is only a suggestion, so Meta can still reach people outside it. Meta hasn't said how it got the 11% or what counts as a quality lead. Skip it if your seed is small or is raw form fills.

As of November 2026.

Quick answer by situation

If you…UseWhy
Have a list of qualified leads or closed-won contactsA lookalike of that list as a suggestion, in one ad setCheap to test, and the seed shows Meta what a good lead looks like
Only have raw form fills or every contactNo lookalike yetMeta says a seed of your best customers may work better than all of them
Need every impression inside your ICPDon't count on a suggestionMeta can deliver beyond suggestions and can expand lookalikes past the size you pick
Mostly get junk leadsA better optimization goal firstMeta's CRM-based goals aim at quality; a suggestion is only a hint
Run several ad sets in one campaignThe suggestion in one ad set onlyLoomer warns that suggestions in several ad sets create overlap in the auction

What is Meta's new lookalike prompt?

Loomer found it while building a Leads campaign. The message reads: "You could get 11% lower cost per quality lead by using a lookalike." It then asks you to choose a list of converted customers or a lookalike. Loomer says he sees it only in Leads campaigns, so he doesn't read it as a wider change in Meta's advice.

Meta hasn't published a help article or announcement about the prompt. We found nothing from Meta on how it measured the 11%, across which advertisers or dates, or what "quality lead" means in this message.

Elsewhere, Meta ties "quality" to CRM data. Its developer docs list a QUALITY_LEAD optimization goal for lead ads and explain how to use it with a CRM data source. They say the Conversion Leads goal "may yield higher quality leads that are more likely to convert," and that it works only with Instant Forms. So treat the 11% as Meta's claim, not a forecast for your account.

What does a lookalike suggestion actually do?

In a Leads campaign, the Audience section shows Advantage+ on, and most audience settings are suggestions. Meta says ads may also reach other people when that's likely to improve performance, but never outside your audience controls: location, minimum age, custom audience exclusions and languages. Meta lists "lookalike or custom audiences to include" among the suggestions you can add.

Two more things widen delivery:

  • Advantage+ lookalike. Meta turns it on by default for new, duplicated and draft campaigns and ad sets that use a lookalike built from a customer list, website or app audience. When Meta finds better results outside the lookalike percentage you chose, it may expand past it.
  • Website leads. Meta's developer docs list optimization goals where lookalike expansion is always on, and offsite (website) conversions is one of them.

So a suggestion tells Meta where to start. It isn't a fence. Loomer says it's very hard to prove that suggestions change delivery, and that where he could test it, Meta repeatedly ignored them.

If you have a list of qualified leads or closed-won contacts

Use the people who matter to sales, not everyone who filled in a form. For B2B, that means leads your team marked qualified, or the contacts on closed-won deals. Meta's help says a lookalike made from your best customers may do better than one made from all your customers. Loomer suggests a lookalike of qualified leads when lead quality is the problem.

To test it:

  1. Build the lookalike from that list. The steps, the seed size Meta needs and the value-based option are in how to build a Meta lookalike from closed-won customers.
  2. Add it as a suggestion in one ad set. Compare it with a broad ad set on the same budget, form and creative.
  3. Keep your customer and open-opportunity exclusions. Meta always treats custom audience exclusions as controls, so they hold even with Advantage+ audience on.
  4. Judge it on the share of leads that become qualified, and on pipeline in your CRM, not on Meta's cost per lead.

If you only have raw form fills or every contact

Skip the lookalike for now. A seed of every form fill points Meta at more people like your average lead, which isn't what the prompt promises. B2B customer lists can also match poorly, so check your match rate first; see why Meta match rates are low with work emails. Come back to it once your CRM can export a clean list of qualified leads.

If you need every impression inside your ICP

A suggestion won't do that. To make an audience a hard limit, open Further limit the reach of your ads in the ad set's Audience section and uncheck Use as a suggestion for the custom audience you include. Meta warns that removing suggestions can turn Advantage+ audience off. And Advantage+ lookalike can still take delivery past the percentage you chose. Whether giving up Advantage+ audience is worth it for B2B is covered in whether to turn off Advantage+ audience.

If you mostly get junk leads

Change what Meta optimizes for before you change the audience. Send lead stages from HubSpot or Salesforce back to Meta and optimize for qualified leads; see how to send CRM lead stages to Meta. If you only get a few leads a week, see what to optimize for with low B2B lead volume. For website forms, sequenced events are another way to point Meta at qualified leads. Meta's lead ads guide also lists form settings for higher quality leads, such as a higher intent form, which adds a step where people review and confirm their answers.

If you run several ad sets in one campaign

Add the suggestion to one ad set only. Loomer warns that suggestions spread across several ad sets create needless overlap in the auction. If you want to compare audiences properly, run a planned test with one change at a time.

What to avoid

  • Taking the 11% at face value. Meta hasn't said how it measured it, and cheaper leads aren't the same as more pipeline.
  • Seeding the lookalike from every lead or every contact.
  • Assuming the lookalike limits who sees your ads.
  • Changing the audience, the form and the optimization goal at once, so you can't tell which change worked.

Sources

Ask Tofu next

Pick a question and Tofu Ads Agent answers it in chat.

Related answers

More on

Let an agent build your next campaign.

Tofu Ads Agent reads your CRM, plans the funnel, builds the audience and creative, and stages full campaigns across LinkedIn, Google and Meta for your approval.

Try it free Get a demo