You've spent $600 this month on Facebook ads. You've got 34 leads in your CRM. You've called every one of them. Six didn't pick up. Eleven told you they're "just looking, maybe in a year or two." Four asked if you could help them rent, not buy. Two wanted to know if you sold mobile homes. You sell nothing under $700,000.
This isn't bad luck. This is what happens when your targeting is broken and nobody told you.
Every agent running Facebook ads has had this conversation with themselves at some point. The ad account says the campaign is performing. Cost per lead looks reasonable. The dashboard is green across the board. And yet the leads are useless, one after another, like Facebook found every person in your zip code who has zero intention of buying a house and handed you their phone number instead.
The platform isn't broken. Your targeting is. And most agents never find out why because they're looking at the wrong metrics to diagnose it.
The Targeting Setting Everyone Gets Wrong
Meta gives you two fundamentally different ways to tell it who should see your ad: broad targeting, where you let the algorithm figure out who's interested based on your ad content and a few basic parameters, and detailed targeting, where you manually stack interests, behaviors, and demographics on top of each other.
Most agents use detailed targeting. Most agents use it badly.
The instinct is to add more filters, thinking more specificity means more qualified traffic. Age 35 to 55. Household income $150k plus. Interest in "real estate." Interest in "home improvement." Interest in "Zillow." Stack them all together and it feels like you've built a laser guided audience.
What you've actually built is a soup of everyone who checked any one of those boxes, because Meta's detailed targeting options work on an "or" basis within most groupings, not an "and." Someone doesn't need to be 35 to 55 AND high income AND interested in Zillow. Depending on how you've layered your groups, they might just need to hit one loosely related signal, and Meta's optimization will chase whoever is cheapest to reach that still technically qualifies.
Cheapest to reach is rarely the same as most likely to buy a $700,000 house from you specifically.
"Homeowners" Is Not an Audience
Here's the sentence that should get burned into every agent's brain before they open Ads Manager again: a demographic is not a buyer intent.
Targeting "homeowners age 30-45" doesn't mean you're reaching people who want to sell. It means you're reaching people who own a home. Most of them are not selling. Most of them never even think about selling on a random Tuesday when your ad interrupts their scroll between a recipe video and someone's vacation photos.
According to Meta's own advertising guidance, detailed targeting options are built from user behavior and declared interests, not intent signals like "actively looking to sell in the next six months." That signal doesn't exist as a targeting checkbox because Facebook doesn't actually know it. Nobody clicks a box that says "I plan to list my house in March." You're inferring intent from people who liked a home decor page once in 2021.
This is the gap between what feels like precision targeting and what's actually happening under the hood. You think you're speaking to sellers. You're speaking to homeowners. Those are wildly different audiences, and the wrong one is expensive in ways that don't show up until three weeks later when you're on the phone with someone who has no idea why you're calling.
If your ad copy assumes buyer intent but your targeting only guarantees demographic overlap, you've built a mismatch that no amount of budget fixes.
Lookalike Audiences Built on the Wrong Seed
Lookalike audiences sound like magic. Feed Meta a list of your best clients, and it finds more people just like them. In practice, the quality of a lookalike audience is entirely dependent on the quality and size of the seed list you fed it, and most agents feed it garbage without realizing it.
A common mistake: building a lookalike audience off your entire contact list, including every open house sign-in, every cold lead who never responded, every random person who filled out a home valuation form out of curiosity. That seed list isn't "my best clients." It's "everyone who has ever had any contact with me for any reason," and the lookalike Meta builds from it reflects that noise.
A better seed is a list of closed transactions from the last 12 to 24 months, ideally 100 contacts or more so Meta has enough signal to work with. Even better, segment it. A lookalike built from your last 20 luxury listing clients will produce a wildly different audience than one built from your last 20 first-time buyer clients, and mixing them together muddies both.
Most agents never revisit their lookalike source list after the first setup. If the seed was wrong in month one, it's still wrong now, quietly feeding bad matches into every campaign built on top of it.
Your Ad Copy Is Doing the Opposite of Qualifying
Even with decent targeting, your ad copy can undo all of it in one line.
"Thinking about buying or selling? Let's talk!" is an invitation for absolutely everyone. It doesn't repel anyone who isn't serious. It doesn't filter for budget, timeline, or motivation. It's the digital equivalent of standing on a street corner asking strangers if they'd like to chat about real estate. Some will. Almost none of them are ready to do anything.
Ad copy should do the opposite of what most agents write it to do. It should scare off the wrong people, not attract everyone.
Compare that generic line to something like: "Looking to sell your Sacramento home for $600k or more in the next 90 days? Here's what buyers are actually offering right now." That copy pre-qualifies on price, timeline, and location before anyone even clicks. A person who's just curious about the market, not selling, not in that price range, not in that city, reads it and moves on. That's the point. You want them to move on. Every click from someone who was never going to close is money you can't get back.
HubSpot's research on ad copy specificity consistently shows that narrow, specific copy converts at lower volume but meaningfully higher quality than broad, come-one-come-all messaging. Lower volume with better fit beats higher volume with garbage fit every single time, and yet most agents chase the volume number because it feels like progress.
If you've already worked out how to position your services around a specific problem instead of competing on price, your ad copy should reflect that same discipline. Speak to one person with one problem. Let everyone else scroll past.
The Landing Page That Undoes Everything Your Targeting Got Right
Say your targeting is dialed in. Say your copy filters correctly. Then the click lands on your homepage, and the whole thing falls apart anyway.
A visitor who clicked an ad about selling a home for top dollar in a specific price range should land on a page that speaks to exactly that. Not your general homepage with a headshot and a tagline about your passion for real estate. A dedicated landing page that repeats the promise from the ad, asks for minimal information, and gives them one clear next step.
If you've read about how homepage copy converts, the same principles apply here with even less patience from the visitor. Someone who clicked a paid ad is more impatient than someone who found you organically. They expect the page to match what they just clicked. When it doesn't, they bounce, and Meta's algorithm eventually notices that your landing page is failing to convert traffic it worked hard to deliver, which can quietly increase your costs over time regardless of how good your targeting was.
Every mismatch between ad promise and landing page reality is a leak. Fix the targeting and the copy and still send people to the wrong page, and you're back to paying for clicks that go nowhere.
Broad Targeting Isn't Always the Villain
Here's the twist most agents don't expect. Sometimes the fix for bad leads isn't narrower targeting. It's broader targeting, paired with tighter creative.
Meta's algorithm has gotten considerably better at finding conversion-likely users on its own when you give it room to work, rather than boxing it into a small, over-filtered audience. Meta's advertising documentation on campaign optimization notes that overly narrow audiences can actually limit the algorithm's ability to find efficient placements, sometimes driving costs up rather than down.
A broad audience with sharp, self-selecting ad copy and a strong dedicated landing page can outperform a narrow, over-stacked audience with generic messaging. The targeting box isn't the only lever. Sometimes it's the least important one, and the creative is doing the heavy lifting of qualifying who actually engages.
This doesn't mean abandon targeting entirely. It means stop assuming that stacking interest after interest is inherently smarter than a broad audience with copy that does the filtering work instead.
The Budget Trap That Attracts Tire Kickers
Underfunded campaigns behave strangely. When your daily budget is too low relative to your audience size, Meta's delivery algorithm doesn't get enough data fast enough to optimize toward your actual goal, whether that's leads, calls, or form fills. Instead it often defaults to whoever is cheapest to serve an impression to, which tends to skew toward less qualified, more passive users.
WordStream's benchmark data on real estate advertising shows real estate has some of the higher cost-per-lead figures across industries, in part because the buying decision is high stakes and slow. Running $5 a day and expecting the algorithm to find serious $700k buyers in that budget is asking a lot of a system that needs enough spend and data volume to learn who converts.
If you're running multiple small campaigns spread thin across audiences instead of one well-funded campaign with a clear goal, you're likely starving all of them of the learning phase they need to actually optimize. Consolidate. Fund one campaign properly rather than five campaigns barely.
Retargeting the Wrong Behavior
Retargeting ads, the ones that follow someone around after they visit your site, are only as good as the behavior you're targeting them for.
Retargeting everyone who visited any page on your site treats a person who spent 40 seconds on your blog post about mortgage rates the same as someone who spent four minutes on a specific listing page and clicked "schedule a showing." Those are not the same intent level, and serving them the same retargeting ad wastes budget on the low-intent group while under-serving the high-intent one.
Set up custom audiences based on specific page visits and time-on-page thresholds, not just "anyone who's ever been to my domain." A visitor who viewed three listing pages in one session is a different animal than someone who bounced off your blog in eight seconds. Treat them differently, or you'll keep retargeting people who were never close to ready.
What Good Targeting Actually Looks Like
Pulling this together, a well-built campaign usually has these traits working at once, not in isolation. A defined audience based on actual behavior signals rather than vague demographics. Ad copy that pre-qualifies by naming a specific price range, timeline, or situation. A dedicated landing page matching the ad's exact promise. A budget large enough to give Meta's algorithm room to learn. And retargeting segmented by actual on-site behavior, not blanket site visits.
Miss any one of those five and the whole system leaks. Most agents nail one or two and wonder why the campaign still produces junk. It's rarely one catastrophic mistake. It's usually three or four small ones compounding.
Auditing Your Current Campaigns This Week
Before you spend another dollar, pull up your last 30 days of ad performance and answer a few questions honestly.
What does your current targeting actually consist of, and would you describe it as buyer intent or just demographic overlap? What's your seed list for any lookalike audiences, and when did you last update it? Does your ad copy name a specific price point, timeline, or situation, or does it invite literally anyone? Where does the click actually land, and does that page repeat the exact promise from the ad? And is your daily budget large enough for the audience size you're targeting, or are you spreading a small amount across too many segments?
Google Search Console and your site analytics can tell you what's happening after the click, which is often where the real problem hides even when the targeting looks reasonable on paper.
If you've already built out a lead magnet that's supposed to be doing some of this qualifying work for you, check whether your Facebook traffic is even landing on it, or whether it's going straight to a generic contact form that asks for nothing more than a name and email with zero context about what the person actually wants.
The uncomfortable truth is that most agents running bad Facebook campaigns aren't bad marketers. They're busy people who set up a campaign once, watched the lead count go up, and never went back to ask whether those leads were worth anything. If your leads aren't converting, the platform isn't always the problem. Sometimes the fastest fix isn't a new ad. It's turning off the campaign that's been quietly funding a pile of names you'll never close, and rebuilding it with intent instead of guesswork.
What would your lead quality look like if you cut your ad spend in half and only reached people who actually match what you sell?
