How a Real Estate Brokerage Cut Cost Per Lead 41% with AI Video Ads

A residential brokerage with 45 agents across three metro markets was spending $35,000 per month on Meta and Google Ads to fill agent pipelines. The leads were coming in — sort of. Their average cost per lead on Meta had climbed to $52, up from $38 just eight months earlier. Google Search was worse at $94 per lead. The math was getting uncomfortable.

The team knew the problem wasn't targeting. Their audience segments were dialed in: homebuyers by zip code, income bracket, and life-stage signals. The problem was creative. Every ad in their account was built from the same playbook: a hero shot of the listing, an overlay with price and beds/baths, and a "Schedule a Tour" CTA. These static image ads had worked fine in 2023. By early 2026, they were invisible.

Listings with video receive 403% more inquiries than those without, according to NAR data. The brokerage knew this. But producing video for 200+ active listings across three markets was a logistics nightmare they hadn't been able to solve — until AI changed the production economics entirely.

The Static Photo Problem

Real estate advertising has a creative problem that most other industries don't face: the product changes every week. A retail brand can run a hero product video for months. A real estate brokerage has new inventory constantly, and each listing has a limited shelf life. By the time a traditional video is produced, edited, and uploaded, the listing may already be under contract.

The brokerage had tried video before. They hired a videographer to shoot walkthrough tours for their premium listings — properties above $750K. The results were strong: video listings generated 2.8x more lead form submissions than static ads for the same properties. But the production cost was $500–$1,500 per listing, the turnaround was 5–7 business days, and it only covered their top 15% of inventory.

The remaining 85% of listings — the bread-and-butter properties between $300K and $700K that generated most of their lead volume — were stuck with static photos.

The data was clear: video worked dramatically better. The economics of traditional video production made it impossible to scale.

The AI Video Shift

In Q4 2025, the brokerage's marketing director tested AI video generation tools that could transform existing listing photos into short-form video ads. The workflow was simple: upload 8–12 listing photos, select a template style (modern walkthrough, lifestyle showcase, or neighborhood highlight), and the AI generated a 15–30 second video with motion, transitions, and text overlays.

Cost per video: under $4. Production time: 3 minutes.

The team started with a controlled test. For 30 days, they ran parallel campaigns on Meta: static photo ads for half their listings, AI-generated video ads for the other half. Same targeting, same budgets, same landing pages. The only variable was the creative format.

The results after 30 days were decisive:

Video ads weren't just performing better — they were performing better at a fraction of the creative cost. The $3.80 per video was negligible compared to the $16 CPL reduction.

After the initial test, the team went all-in. Every new listing got at least two AI video variations: one walkthrough-style and one lifestyle/neighborhood angle. Premium listings got four variations for A/B testing. Within 60 days, they had 180+ active video creatives in rotation — more than their entire creative library from the previous two years combined.

Creative Volume Solved the Fatigue Problem

The unexpected benefit wasn't just that video outperformed static images. It was that high creative volume eliminated the fatigue cycle that had been silently inflating their CPL for months.

With static ads, the brokerage had been running the same 10–15 creatives until they stopped performing, then scrambling to replace them. Meta's algorithm penalizes repetitive creative — frequency scores above 3.0 typically correlate with CPL increases of 20–40%. Their average frequency had been hovering around 4.2.

With AI video, they could retire underperforming creatives and replace them in minutes. Fresh creative kept frequency scores below 2.0 across all campaigns. The algorithm rewarded the variety with lower CPMs and better delivery.

How Soku Fit Into the Workflow

Producing 180+ video ads was the easy part. The hard part was figuring out which of those 180 videos were actually generating qualified leads — not just clicks, not just form fills, but leads that converted to showings and eventually to closed deals.

The brokerage connected Soku AI to their Meta Ads, Google Ads, GA4, and CRM (Follow Up Boss). All read-only integrations, set up in under 20 minutes. No engineering support needed.

Creative-to-Outcome Attribution

Soku mapped every creative asset to downstream outcomes: lead → showing scheduled → showing completed → offer submitted. This revealed patterns invisible in platform-native reporting:

• Walkthrough-style videos generated more leads overall, but lifestyle/neighborhood videos generated leads that were 2.3x more likely to schedule a showing
• 15-second videos outperformed 30-second versions on CPL, but 30-second versions produced higher-intent leads
• Text overlay with price performed best in Tier 2/3 markets; no price overlay performed best in Tier 1 markets where buyers expected higher prices

Automated Creative Scoring

Instead of manually reviewing which creatives were working, Soku scored each video on a composite metric that weighted CPL, lead-to-showing rate, and creative freshness. The marketing team got a weekly digest: top 10 performers to scale, bottom 10 to retire, and specific recommendations for what to produce next.

Cross-Market Intelligence

With listings across three metros, the team could see which creative approaches worked in each market. A cinematic drone-to-interior transition that crushed it in their suburban Dallas market fell flat in their urban Chicago campaigns, where lifestyle-focused neighborhood walkthroughs performed 40% better. Soku surfaced these patterns automatically rather than requiring the team to dig through platform dashboards.

The Results After 90 Days

After three months of running AI video ads with Soku-powered optimization, the numbers told a clear story:

The $35K monthly budget didn't change. What changed was what that budget produced: 3.2x more leads at 41% lower cost per lead on their primary channel. The lead quality improvement — measured by lead-to-showing conversion — meant agents were spending less time chasing unqualified leads and more time with serious buyers.

What Made This Work

The brokerage's success wasn't just about switching from photos to video. Several specific decisions compounded the results:

1. Volume over polish. AI-generated videos aren't cinematic masterpieces. They're functional, attractive, and fast. The team stopped optimizing individual creatives and started optimizing the system: more variations, faster testing, quicker retirement of underperformers.

2. Format matched to funnel stage. Walkthrough videos for cold audiences (awareness), neighborhood lifestyle videos for warm audiences (consideration), and price-specific urgency videos for retargeting (conversion). Each stage got different creative, not the same ad shown three times.

3. Market-specific creative, not one-size-fits-all. What works in suburban Dallas doesn't work in urban Chicago. AI production volume made it feasible to create market-specific creative libraries instead of running the same templates everywhere.

4. Attribution beyond the click. Knowing CPL is useful. Knowing which creatives produce leads that actually show up to viewings is transformative. The CRM integration through Soku closed the loop between ad spend and real business outcomes. For a deeper dive on building this kind of measurement stack, see our guide to measuring AI ad creative ROI.

5. Creative freshness as a strategy. Instead of running winning ads until they die, the team proactively rotated creative on a 2-week cycle. This kept frequency scores low and CPMs favorable, even as competitors' costs rose during spring buying season.

The Bigger Picture

Real estate advertising is entering a phase where the teams that can produce and test creative at volume will have a structural advantage. The industry benchmarks bear this out: Meta CPL for real estate averaged $51.90 in 2025 and is projected to reach $54–$57 in 2026. Teams still running static photo ads from their MLS feed are going to feel the squeeze.

AI video generation removes the production bottleneck. Creative intelligence tools like Soku remove the optimization bottleneck. Together, they turn what used to be a creative team's full-time job into an automated workflow that scales with your listing volume — not your headcount.

For this brokerage, the shift wasn't a marginal improvement. It was a fundamentally different way of running their lead generation engine. And the agents filling their pipelines at $31 per lead instead of $52 aren't looking back.