A property tour is not an advert. A tour assumes the viewer has already decided to consider the house, whereas an ad has two seconds to make them consider it, which means leading with the price, the location and the one trade-off that defines the property rather than with the front door. The second thing most agents get wrong is production economics: you do not need one ad per listing, you need a small set of ads about buyer decisions and about your own track record, plus a handful of listing-specific cuts for the properties that justify them.
What should a property ad lead with?
The decision, stated as text on screen in the first two seconds. Buyers filter on three things before they look at a single photograph: price, location and size. Put all three in the opening frame and you have qualified the audience before the footage begins.
A structure that works at 15 seconds:
- 0 to 2 seconds: the numbers. Price, area, bedrooms, and one distinguishing fact ("chain free", "12 minutes to the station", "south-facing garden").
- 2 to 6 seconds: the single room that sells the property, in motion, not a static photo.
- 6 to 11 seconds: the objection answered. Parking, the second bathroom, the extension potential, the school catchment, the service charge.
- 11 to 15 seconds: one instruction. Book a viewing, or see the full listing.
What to cut: the exterior establishing shot, the hallway, the ultra-wide lens that makes rooms look like a portal photograph, the piano music, and the agent's logo animation. A hallway has never sold a house.
There is also a strategic point most agents skip. Buyer-side ads compete with the portals, which already own buyer intent, so the highest return video an agency runs is usually aimed at homeowners rather than buyers: proof that you sell property like theirs, at the price they hope for, quickly. Those ads are evergreen, they do not expire when a listing does, and they are what fills the pipeline.
How do you vary creative across listings without one ad per listing?
Separate the ads that are about a property from the ads that are about a decision. Only the first group is tied to an address.
| Ad type | Audience | What leads | Per-listing production? |
|---|---|---|---|
| Vendor acquisition ("thinking of selling") | Homeowners in the area | Your result: days on market, price achieved | No, evergreen |
| Sold proof | Homeowners in the area | The specific street and the timeframe | No, one per completed sale, reused |
| "What X buys here" | Buyers in a price band | Three properties compared in one ad | No, refreshed monthly |
| Buyer segment ads | First-time buyers, downsizers, investors | The constraint that segment cares about | No, evergreen |
| Market update | Both | One number and what it means locally | No, monthly |
| Hero listing cut | Buyers | Price, location, the defining feature | Yes, worth it for a few |
| Stuck listing rescue | Buyers who filtered it out | The objection, addressed directly | Yes, for the ones that need it |
| Vendor testimonial | Homeowners | The vendor's own words | No, reused for a year |
Six of those eight need no per-listing work at all. In practice a mid-sized agency runs 8 to 12 evergreen variations, which is roughly what an active campaign needs in rotation, plus two or three cuts for hero listings and rescues. That is a manageable production load, and it is what our batch video ad variation workflow is designed around.
Fatigue behaves differently here too. Our 2026 fatigue benchmark, a synthesis of published platform and agency figures rather than our own experiment, does not break out property. The closest analogues it measures are electronics at about 21 days to a 40 percent CTR decline and B2B SaaS at about 28, both slower than fashion at 12 to 14 or food and beverage at 9, which fits a considered purchase with a long decision cycle. For listing ads the point is usually academic: the property sells or the instruction ends before creative fatigue arrives. For vendor acquisition ads, which run continuously against the same local homeowners, plan a refresh every three to four weeks and watch frequency rather than the calendar.
What footage do you actually need?
Less than a videographer will quote for, but shot with more discipline than a portal walkthrough.
- Stabilised movement, at half the speed you think. A gimbal or a phone with good stabilisation, walking slowly. Fast motion in a small room reads as chaotic.
- Doorway push-throughs. Moving from one room into another is the shot that conveys layout, which photographs cannot.
- Eight to ten seconds per room, filmed twice. Once wide, once on the detail that justifies the price.
- All lamps on, curtains open, no flash. Mixed daylight and tungsten looks worse than either alone, so shoot at the time of day the room is best lit.
- Avoid the widest lens. Distortion is the visual signature of an ad nobody trusts.
- The walk, not just the house. Front door to the station, the high street, the park. Location is half the product and nobody films it.
- Exteriors near golden hour, and drone footage only where you hold the right permissions.
- Written consent from the vendor covering paid advertising, and a note of what you may show after completion.
One compliance point that changes targeting, not footage. In the US and Canada, housing ads sit in a special ad category on Meta, which removes targeting by age, gender and postcode and forces a minimum radius, and fair housing law prohibits copy or imagery implying who a property is or is not for. That pushes the burden onto creative: if you cannot target the segment, the ad has to qualify the viewer itself, which is another argument for leading with price and constraints in the first two seconds.
What about per-listing personalisation at scale?
This is where we should be plain. If what you want is a template that pulls the address, price, bed count, agent photo and five listing images from a feed and renders a video for every property in your book, automatically, every night, that is feed-driven template rendering. It is a real category and a good solution to that problem.
Genyad does not do it. There is no product-feed or CSV-driven template rendering, no product-URL import, and no way to point us at a portal export and get 400 videos back. Plainly does that job properly, built on After Effects templates driven by data, listed at $69 per month in August 2026, and prices move so check before you commit. If per-listing automation is your actual requirement, use it, and our Plainly alternative comparison sets out where each approach fits rather than pretending they overlap.
What Genyad does is the other half of the problem. You upload the footage you already own once, it transcribes and tags every clip, and each variation is a fresh script, shot selection, voiceover, caption set and export drawn from that library rather than a recut of the same timeline. That suits the evergreen set: vendor acquisition, sold proof, buyer segment ads, market updates. One standard variation is one credit, editing and re-exporting cost nothing, and exports cover 9:16, 4:5, 1:1 and 16:9 at 1080p with no watermark on any plan. Our ecommerce video ads page shows the same mechanism applied to products, which is the closest analogue to a listing set.
Other things it will not do: no AI avatars or synthetic presenters, no static banners for portal or display placements, no predicted performance score, and no direct publishing to Meta or TikTok.
Frequently asked questions
Should a real estate video ad be a property tour?
No. A tour serves someone who has already shortlisted the property. An ad has to earn the shortlist, so it leads with price, location, size and one distinguishing fact on screen in the first two seconds, then shows the single room that sells the place.
How many video ads should an estate agency run?
Around 8 to 12 evergreen variations covering vendor acquisition, sold proof, buyer segments and market updates, plus two or three cuts for hero or stuck listings. Most of the set is not tied to an address, which is what keeps production sustainable.
Can Genyad generate a video for every listing automatically?
No. There is no product-feed, CSV or URL import, so per-listing automated rendering is not something Genyad does. A feed-driven template renderer such as Plainly is built for that, and Genyad is better suited to the evergreen agency-brand and buyer-decision set.
How often should property ad creative be refreshed?
Vendor acquisition ads that run continuously against the same local homeowners need refreshing every three to four weeks, judged on frequency rather than the calendar. Listing ads rarely reach fatigue, because the property sells or the instruction ends first.