From Photoshoot to Prompt: How AI Rebuilt Our Ad Creative Process
Catalyst Agents | AI for Marketing Industry Insights
From Photoshoot to Prompt: How AI Rebuilt Our Ad Creative Process
For years, the ad creative process for real estate clients followed the same script. The client hired a photographer to shoot the neighbourhood, and they also hired a creative ad agency or digital rendering artist to create renderings of a building that doesn't exist yet. They would then hand over these images to us, and our designers took what they were given and built a static ad around it. If the light was flat, the season was wrong, or the angle didn't work, that was the ad. Reshoots meant waiting for a weather window and rebooking a crew.
AI has broken that ceiling in two connected ways: image manipulation and image-to-video animation.
Manipulation: the same base image, endless variations
Clients still provide the foundation: photography of the neighbourhood, or a rendering of the townhome or high-rise. But that foundation is no longer the finish line. We can now shift lighting, change the mood, add or swap characters, all from a single source image. A rendering shot in overcast light becomes a golden-hour hero shot. A hallway with no people in it becomes a hallway with a young family walking through it, or a couple downsizing into their next chapter, depending on which buyer segment the ad is targeting.
That last point matters more than it sounds. We are already using this to stage the same property differently for different buyer personas within the same campaign, one aesthetic for young professionals, another for downsizers, rather than picking a single look and hoping it lands broadly, giving our real estate developer clients many more options than they would usually have.
Animation: turning enhanced stills into a full video suite
Once a still image has been enhanced, AI video tools can animate it: a subtle pan across a lobby, a slow push into a kitchen, a seasonal variant with snow instead of sun. Building a comparable suite of seasonal, character, and angle variations the old way meant multiple shoot days, multiple crews, and a client invoice that could run 10 times higher than what AI-assisted production costs today.
What this actually changes for clients
The obvious win is speed. No waiting for optimal weather. No booking a videography crew because a stakeholder wants the building from a slightly different angle. But the deeper change is in how clients collaborate with us.
Paid media agencies used to be creative receivers. A client handed over assets, and the agency ran ads against whatever they'd been given, with little room to adjust once a campaign was live. Now we can adjust the creative itself based on what the performance data tells us will convert better, mid-flight, without waiting on a new shoot. That shifts an agency's role from executing creative to actively improving it based on results. This has been game-changing for us because we're accountable for results.
Two things this doesn't solve
AI removes the production bottleneck, but it introduces two new ones worth planning for.
Disclosure. Meta and Google both enforce AI-content labeling in 2026, not as a suggestion but as a compliance requirement. Meta requires an "AI-generated" label on any ad where AI generated, substantially modified, or composited visual or audio content, and undisclosed AI use is now one of the leading causes of ad rejection on the platform. Google applies the same logic, with account suspension as the penalty for repeat violations. For agencies, this means disclosure must become a pre-flight checklist item on every AI-assisted creative, not an afterthought.
Quality control. AI editing that's rushed or over-processed still looks synthetic, and a fake-looking ad erodes trust faster than a plain one ever did. AI hasn't replaced the need for a strong base photograph or rendering; it's replaced everything that used to happen after that image was captured. The agencies getting this right are the ones treating AI as a multiplier on good source material, not a substitute for it. In the case of a real estate project ad, where buyers understand that nothing exists yet except for a hole in the ground, there's more leeway in using AI characters to help them picture a lifestyle at the location. They understand a rendering, or an animated rendering, isn't the real thing.
Frequently Asked Questions
Do AI-generated real estate ads need to be disclosed on Meta and Google?
Yes. Both platforms require an AI-generated label on ads where AI created or substantially modified visual or audio content, and enforcement includes ad disapproval or account suspension for non-compliance.
Does AI replace professional real estate photography?
No. AI expands what can be done with a base image or rendering, but a weak or low-quality source image will still produce a weak result. Professional photography remains the foundation; AI is what happens after.
How does AI change the cost of producing seasonal or multi-angle ad creative?
Instead of rebooking a photography or video crew for every new angle, season, or character variation, agencies can generate those variations from existing assets, cutting production costs significantly compared to traditional reshoots.
