The Complete Guide to Facebook and Instagram Ad Targeting (That Actually Converts)

You boosted a post, watched the budget disappear, and got twelve likes from people three states away. Sound familiar? Most owners who come to us have tried to target advertising on Facebook at least once, and most of them got burned by the same thing: they paid Meta to show ads to the wrong people.

Here’s the good news. Facebook and Instagram targeted ads still reach buyers more precisely than almost any other channel, and the targeting tools got easier to use, not harder. The catch is that the rules changed. The interest-stacking tricks that worked five years ago now waste money, and the advertisers winning in 2026 are running broader audiences with smarter inputs.

We manage Meta ads for roofers, builders, orthodontic practices, and other owner-run businesses, and this is the same playbook we use on real accounts. Let’s get into it.

How Facebook Ad Targeting Works Now (The Short Version)

Meta’s ad system runs on machine learning. You give it a goal (leads, purchases, calls), a budget, and creative, and its algorithm hunts through billions of users on Facebook and Instagram to find people likely to take that action. Every targeting option you set is an instruction that either helps that hunt or handcuffs it.

The short version: your job shifted from picking the perfect audience to feeding the system good signals. Clean conversion tracking, strong creative, and sensible audience boundaries beat clever interest combinations almost every time now.

That doesn’t mean targeting is dead. It means you need to know which levers still matter. There are four of them.

The 4 Types of Facebook and Instagram Ad Targeting

Every audience you can build in Meta Ads Manager falls into one of four buckets:

• Core (detailed) targeting reaches cold audiences by demographics, interests, and behaviors.
• Custom Audiences retarget people who already know you: your customer list, website visitors, and engagers.
• Lookalike Audiences find new people who resemble your best existing customers.
• Advantage+ audience hands the targeting decision to Meta’s AI, guided by your suggestions.

Each one earns its keep at a different stage of the funnel. Here’s how to use them.

1. Core Targeting: Demographics, Interests, and Behaviors

Core targeting is the classic option and the default starting point when you have no customer data yet. Inside the detailed targeting section of Ads Manager you can narrow by:

• Demographics: age, gender, location, language, education, job titles, relationship status, and life events like recently moved or newly engaged
• Interests: pages people follow, content categories they engage with, hobbies, and shopping preferences
• Behaviors: purchase activity, device usage, travel patterns, and digital habits

Two things to know about this data. First, some of it comes straight from user profiles and is fairly reliable. Second, a lot of it is inferred by Meta from clicks and engagement, so treat interest categories as educated guesses, not census data.

Our take after years of running these campaigns: use core targeting to set boundaries (location, age range) rather than to build elaborate interest stacks. A search for “hiking” in the interest browser feels productive. It usually performs worse than a broad audience with creative that speaks directly to hikers.

2. Custom Audiences: Your Warmest Prospects

Custom Audiences are built from people who already interacted with your business, which makes them the highest-intent targeting available on the platform. You can create them from:

• Customer lists. Upload emails and phone numbers from your CRM, and Meta matches them to user accounts. Use multiple identifiers per contact, because a portion of any list won’t match.
• Website visitors. The Meta Pixel tracks visitors for up to 180 days and lets you segment by page, like everyone who hit your pricing page but never filled out a form.
• Engagement. People who watched your videos, opened a lead form, or interacted with your Facebook page or Instagram profile, with retention windows that stretch up to a year.
• App activity. For businesses with a mobile app, audiences built from installs and in-app events.

One 2026 caution: Meta now blocks custom audiences that imply sensitive traits like health conditions or financial status. Build segments around actions people took, not around who they are.

3. Lookalike Audiences: Cloning Your Best Customers

A lookalike audience starts with a source (say, your past purchasers), analyzes what those people have in common, and finds new users who match the pattern. You pick a similarity percentage from 1% to 10% of the target country’s population. A 1% lookalike is the tightest match and usually converts best. Broader percentages trade precision for reach.

The source is everything here. A lookalike built from 1,000 real customers will outperform one built from 5,000 random page fans, because the pattern Meta learns is only as good as the people you feed it. Meta accepts sources as small as 100 people, but we’d push for 1,000 or more before trusting the results.

Are lookalikes still worth it? Yes, with an asterisk. Advantage+ has absorbed a lot of their job, and on some accounts the AI now beats a standard lookalike. Test both before you commit a budget.

4. Advantage+ Audience: Meta’s AI Takes the Wheel

Advantage+ is Meta’s automated targeting layer. Instead of hard rules, you give it audience suggestions, and the system starts there, then expands wherever it predicts better results. It learns from your pixel data, conversion history, and how users have responded to ads like yours.

For veteran advertisers, this is broad targeting with a brain. And it works better than most owners expect. Meta’s machine learning has more signal about buying behavior than any manual audience you could build.

When we launch a new campaign for a partner, we usually test an Advantage+ ad set against one manually targeted ad set, then follow the data. More often than not, the AI version wins on cost per lead within a few weeks. But not always, which is exactly why we test instead of guessing.

Broad vs. Narrow: The 2026 Targeting Debate

If you read advertiser forums, you’ll see the same argument on repeat: does detailed targeting still matter, or should everyone just go broad?

Our answer is an opinion, so take it as one: broad targeting with strong creative wins for most small businesses. Splitting a modest budget across six narrow ad sets starves each one of the roughly 50 weekly conversion events Meta wants before it optimizes well. One or two broad ad sets concentrate that signal and exit the learning phase faster.

Manual targeting still earns its place in three situations:

1. Local service businesses. If you only serve a 30-mile radius, geography is non-negotiable.
2. Regulated industries. Housing, credit, and employment ads have restricted targeting by law.
3. Genuinely niche B2B. If your buyer is one of a few thousand people nationally, broad delivery wastes money.

Notice that the first category covers most of the businesses we work with. Local is the one filter you should always control by hand.

Location Targeting for Local Businesses

For a Gainesville wellness practice or a Jacksonville roofer, location targeting is the difference between leads and noise. A few settings matter more than people realize:

• Radius targeting lets you draw a circle around your address, down to a tight local radius, or drop pins on specific areas.
• “People living in this location” is the setting most local businesses want. The default includes travelers and recent visitors, which means tourists who will never book with you.
• Exclusions let you carve out areas you don’t serve, so a broad metro audience doesn’t bleed into the next county.

Layer location with an age range and let delivery stay broad inside those lines. That combination respects the algorithm while keeping your spend inside your actual service area.

Retargeting: The Highest-ROI Campaign You Can Run

Someone visits your website, reads your pricing, and leaves. Without retargeting, that’s the end of the story. With it, they see your ad two days later and come back to book.

A retargeting setup that pays for itself looks like this:

Install the Meta Pixel and the Conversions API together. Browser-only tracking lost a lot of accuracy after Apple’s iOS privacy changes. The Conversions API sends events from your server as well, which patches the gaps and gives Meta cleaner data to optimize on.

Build audiences by intent level. All visitors are fine. Pricing-page visitors, form abandoners, and 50%+ video viewers are better, because the message can match how warm they are.

Exclude converters. Take recent customers out of prospecting campaigns and fresh leads out of retargeting. Paying to re-convert people you already won is the quietest budget leak in most accounts we audit.

Keep retargeting on a leash. These audiences convert well but they’re small. Somewhere around 60 to 70 percent of budget on prospecting and 15 to 25 percent on retargeting keeps the pipeline full without fatiguing your warm audience.

An early gut check: if you’re not sure your pixel is firing correctly, fix that before touching anything else on this list. Every audience downstream depends on it. If you’d rather have a second set of eyes on the whole setup, our Facebook and Instagram ads service starts every partnership with a tracking audit.

Your Creative Is Now Part of Your Targeting

Here’s the shift that surprises owners most. In 2026, Meta’s algorithm watches who engages with each ad concept and steers delivery toward similar people. Your creative acts as a filter: an ad that opens with “Gainesville homeowners” finds homeowners, and an ad showing a kitchen remodel finds people dreaming about kitchens.

That has practical consequences:

• Run several creative angles at once. Different hooks attract different segments of the same broad audience.
• Say who the ad is for in the first line. Calling out the customer does the narrowing that interest stacks used to do.
• Refresh before fatigue sets in. A static audience sees the same ad again and again, and performance decays. New creative resets the clock.

When one of our partner campaigns came out hot and then cooled off, the fix wasn’t new targeting. We dug into the data, rewrote the ad copy around the themes that were resonating, and swapped the lead form. Same audience, better inputs, and the cost per lead came back down. If producing that volume of creative sounds like a lot, a batch content shoot can bank months of ad material in a day.

How Much Budget Does Good Targeting Need?

Targeting choices and budget size are tied together. Every ad set needs enough conversions to learn, so the smaller your budget, the fewer audiences you should run. An account on a modest monthly budget does best with one campaign, one broad ad set, and a handful of creatives. An account spending several times that can support separate prospecting and retargeting campaigns with room to test.

Costs vary by industry and market, so we won’t pretend there’s a universal number. We’re upfront with every partner about this: paid social needs a realistic proof-of-concept budget to generate enough data to judge, usually a few months at a spend level that can produce steady weekly conversions. Anyone promising exact lead counts before seeing data is selling you something.

If you want the fuller picture of how paid channels fit together, our Google Ads and Facebook ads pages break down how we run search and social side by side. Meta finds demand; search captures it. For lead-driven local businesses, pairing this with Google Local Services Ads covers both ends.

5 Targeting Mistakes That Burn Budgets

We audit a lot of ad accounts. These five mistakes show up constantly:

1. Over-segmentation. Six ad sets splitting one small daily budget, none of them ever leaving the learning phase.
2. Interest stacking. Twenty stacked interests muddy the signal. Pick one or two clear themes if you use interests at all.
3. No exclusions. Existing customers seeing acquisition offers, fresh converts seeing retargeting ads.
4. Stale audience windows. A 180-day website audience for a business with a one-week sales cycle retargets people who moved on months ago.
5. Browser-only tracking. Skipping the Conversions API starves the algorithm and weakens every audience built on pixel events.

Fix these five and most accounts improve before a single new audience gets built.

Get Targeting That Pulls Its Weight

You could spend the next month testing audiences yourself, and if you have the time, this playbook will get you moving in the right direction. Most owners we talk to don’t have the time. That’s the job they hand to us.

For most growth-stage businesses we build a custom package that combines SEO and paid advertising to drive consistent, qualified traffic, with Meta targeting built and tested by the team that does this daily. You’ll see exactly what we’re doing and why, down to the audience level, because transparent attribution is the whole point.

Happy to jump on a call if you want to talk through your account. Reach out here and we’ll take a look at what’s working, what’s not, and what we’d test first.