IBG Property - 1,721 qualified leads at $34 over nine months

Context

IBG Property is one of the largest real estate agencies in Thailand, with a wide portfolio - from $100K studios to villas above $5M - and buyers from around the world. When the engagement began the advertising worked, but it would not scale: budget was being spent, yet as volume grew the cost per lead rose disproportionately and lead quality fell.

What the audit found

Mistake one: three languages out of six. Only English, Russian and Ukrainian were covered. Polish, Spanish and Arabic were left alone - and Arabic is the cheapest CPL in this niche.

Mistake two: the offer attracted the wrong people. The Russian-language campaigns ran a mortgages-in-Thailand offer. It pulled in people without money who expected to buy on credit, and qualification dropped to 5-10%.

Mistake three: qualification meant nothing. Anyone who replied to a message was counted as qualified. “Hi” was a qualification. “Not interested” was a qualification. There was no way to tell which campaigns worked, and Meta’s optimization was receiving noise.

Mistake four: launches were rare. One property ran for weeks, burned out, and only then was the next one launched. No flow of fresh campaigns means no data to optimize against, no material for A/B tests, and no way to find working combinations quickly.

What we changed in lead generation

Six languages instead of three. Polish, Spanish and Arabic added.

A real qualification rule. The new definition: a genuine inquiry, a budget from $100K, and readiness to place a deposit within three months. Hard criteria instead of “replied to a message”. Campaigns were then optimized on qualifications rather than form fills, with a weekly export from the CRM, source analysis and scaling of whatever held up.

A new offer for Russian speakers. From mortgages to the investment case: 8-12% ROI, guaranteed rental income, concrete yield figures. Qualification rose immediately.

Launches on a schedule. A new property every two or three days, producing a constant flow of fresh campaigns - material for A/B tests and data for optimization.

Video instead of static. Drone footage, interiors with people in them, short formats built for Reels. Click-through rate roughly tripled.

What we built on Telegram

A five-touch warming sequence over ten days.

• Day 0: lead magnet
• Day 2: client case
• Day 4: objection handling
• Day 7: offer
• Day 10: deadline

Lead magnets per segment: 15 properties yielding above 10%, how to buy property in Thailand in seven steps, developer due-diligence checklists.

Optimizing on the subscription itself. The subscribe event is passed back into Meta Ads, so the algorithm optimizes toward people who actually subscribe rather than people who click. Cost per subscriber settled at $1.10.

Results over nine months

Lead generation:

• $58,514 spent on acquisition
• 1,721 qualified leads
• $34 per qualified lead, 30% below the previous level

Telegram channel:

• $17,500 spent on subscriber acquisition
• 16,370 subscribers
• $1.10 per subscriber
• 20-30% convert to a lead within ten days
• ~20% churn, against roughly 30% market average

What made the difference

Opening new markets immediately. Not “fix the existing ones first, add new ones later”, but three new languages launched in parallel. That opened the Arabic market at the lowest CPL available and pulled the portfolio’s average cost per qualification down fast.

Hard qualification criteria. When qualified means a $100K budget and readiness to place a deposit, Meta receives a clean signal and starts learning from real buyers. That beats any attempt to optimize campaigns against a vague definition.

Telegram as a separate long channel. A subscriber at $1.10 is an investment in future transactions. That person receives market content for months, sees the cases, works through their objections, and by the time they are ready to buy they already know the agency.