Why real estate Facebook ads fail, and it is almost never the targeting

Every few weeks a realtor tells us the same thing. They ran Meta ads, spent somewhere between $500 and $3,000, got a handful of junk leads, and concluded that Facebook does not work for real estate.

Then they show us the ads. In almost every case the targeting was fine. Right city, right age band, right income. What they were running was a carousel of listing photos with "DM me for details" underneath.

Meta will happily show a bad ad to the right people

This is the part that confuses agents. The platform does its job. It puts your ad in front of exactly the audience you asked for. Those people then scroll past it in under a second, because it looks like every other real estate ad they have seen this week and it asks for something before it offers anything.

Meta reads that as a weak ad and charges you more to keep showing it. Cost per lead climbs. The agent raises the budget, which makes it worse, then turns everything off.

The four failure modes we see most

The brochure in video form. Slow pans over an empty kitchen, piano music, price on screen. It photographs the house and says nothing to the person watching. Sound does more of the work than most people credit, but it cannot rescue a brochure. Nobody stops scrolling for a brochure.

English-only creative for a non-English buyer. This is specific to our market and it is enormous. The strongest buying communities in Brampton, Surrey and Mississauga talk about money in Punjabi, Gujarati or Hinglish at home. An English ad reaches them at the wrong moment, in the wrong register. We have watched the same offer, same budget, same audience, produce very different costs per lead purely on the language of the script.

No hook in the first second. A Reel and an ad live in the same feed and get judged the same way. If the first line is "Hi, I'm a realtor serving the GTA," the viewer is gone before the second line. Our hook formula applies to paid creative exactly as it does to organic.

No follow-up system. A Meta instant form takes four seconds to fill in, which means it is filled in casually. If nobody replies for six hours, that person has already spoken to two other agents. Cheap leads with no follow-up are worse than expensive leads with follow-up, because the volume convinces you the campaign is working while none of them convert.

What changes when the creative changes

Across our realtor accounts we have generated more than 4,000 leads through Meta. The spread between our best and worst campaigns is not explained by targeting, budget or bidding. It is explained by the video.

The campaigns that work look like content, not advertising. A person on camera, talking about one specific thing a buyer is actually worried about: what $2,000 a month gets you in Pickering, whether the stress test kills your pre-approval, what a status certificate really means. Captions burned in. Shot vertically. Often in Hinglish or Punjabi.

Some of those videos are made for the campaign only and never posted to the grid at all. They exist purely to be an ad, built to the same standard as a Reel we would publish.

The uncomfortable part

Better creative costs more to make than a carousel of listing photos. That is the real reason most agents do not do it. Making one good ad video takes a script, a shoot and an edit, and the temptation is always to skip that and let the targeting carry it.

It will not. Targeting is a commodity, every agent in your city has access to the same audiences, and the auction sorts the winners by who holds attention. That is a creative problem with a creative solution.

Six of our own ten campaigns were flagged too expensive

On 19 August 2026 our internal ads dashboard was running ten campaigns across six clients. It flagged six of them as costing too much per lead and passed four. We are publishing that number because it is the most useful thing we can tell you about why real estate Facebook ads fail: it happens constantly, to people who do this full time, and the fix is almost never the audience settings.

Those ten campaigns spent CA$2,779 and returned 469 leads that month. The cheapest ran at CA$1.53 per lead. The most expensive ran at CA$34.50. Same agency, same month, same country, a 23x difference in what a lead cost.

What separates the CA$1.53 campaign from the CA$34.50 one

Not the budget, and not the radius. The cheap campaigns were the ones where the creative was already working organically before a dollar went behind it. The expensive ones were built the other way round: a competent ad written to a brief, launched cold, with the audience doing all the work.

This is why we treat paid as an amplifier of content that has already proven it can hold attention. A reel that people watch to the end for free is a reel that converts cheaply when you pay to show it to more people. A reel nobody watched organically does not improve because you spent money on it.

The practical version: before you scale a campaign, check whether its creative earned attention without spend. If it did not, fix the creative. Raising the budget on a bad ad buys you the same failure at a higher price, and Meta will deliver it efficiently to exactly the people you asked for.