Case study
Scaling an eight-project residential portfolio on Meta and Google
50,877 leads on PKR 52.7M of managed spend at a blended PKR 1,036 cost per lead — across seven local projects and an overseas programme in five countries.
An established developer needed a steady, sales-ready flow of enquiries for several high-ticket residential projects running at the same time. Units sell for PKR 9M to PKR 40M, the decision takes months, and the same household can be a buyer for four projects at once. The programme also had to reach overseas Pakistanis buying from five countries.
The programme was led by Outecsion’s founder, who owns the developer’s paid media end to end — strategy, budgets, creative direction and reporting.
The results
- 50,877 leads across 66 ad sets on PKR 52.7M of managed spend, at a blended cost per lead of PKR 1,036.
- Two markets run as two businesses: 48,370 local leads at PKR 975, and 2,507 overseas leads at PKR 2,215 on media that costs four times more per impression.
- Year on year (H1 2026 vs H1 2025): media spend down about 45%, cost per lead down about 48%, and lead volume still up about 5%.
- Lead quality: qualified-lead ratio above 10% for three months in a row, measured on sales-team feedback, not form fills.
- Budget trust: monthly budget grew from a PKR 7M plan to a PKR 10M plan, each step justified by delivered numbers.
Portfolio at a glance
| Project | Ad sets | Leads | Cost per lead | CTR |
|---|---|---|---|---|
| Entry-luxury apartment towers | 21 | 23,011 | PKR 787 | 0.63% |
| Premium apartments, city corridor | 9 | 11,311 | PKR 1,438 | 0.54% |
| Luxury apartments and sky villas | 3 | 4,707 | PKR 911 | 0.80% |
| Mid-market apartments | 3 | 3,230 | PKR 763 | 0.50% |
| Large-format family apartments | 10 | 3,229 | PKR 856 | 0.73% |
| Residential plots | 2 | 2,497 | PKR 1,170 | 0.30% |
| Family apartments, broad range | 4 | 385 | PKR 1,850 | 0.44% |
| Local total | 52 | 48,370 | PKR 975 | 0.57% |
| Overseas — premium apartments (Gulf) | 3 | 1,047 | PKR 2,881 | 0.62% |
| Overseas — apartments (KSA, UAE, Australia) | 11 | 1,460 | PKR 1,738 | 0.50% |
| Portfolio total | 66 | 50,877 | PKR 1,036 | 0.57% |
The constraints
- Housing ad rules. Meta limits age, gender and postcode targeting for property ads, so precision had to come from creative and offer.
- Seven projects, one city. Left alone, the projects would bid against each other for the same buyers.
- Long decision cycles. Sales happen offline, weeks later, so campaigns could not optimise to revenue.
- No Conversions API. Leads moved by CSV into a CRM, so the quality loop had to be closed by hand.
- Event-driven demand. Launches and price changes create short spikes that a flat budget wastes.
The operating system we built
| Layer | The rule |
|---|---|
| Audience architecture | Separate by product type and price band first, intent second. Never by demographic alone. |
| Account structure | One campaign per project per objective. Consolidate until each ad set can exit learning. |
| Budget model | A fixed floor per project plus a flexible tranche that follows the launch calendar. |
| Creative engine | Win the first three seconds. Let creative do the segmenting that targeting cannot. |
| AI in the workflow | AI for volume, variation and analysis. People for judgement and the final call. |
| Retargeting and lookalikes | Every prospecting rupee must build an audience. Seed lookalikes on qualified leads only. |
| Measurement | Optimise to the best proxy for intent and close the loop manually where platforms cannot. |
Five findings that shaped the budget
| Finding | Evidence |
|---|---|
| Timing beats targeting | Launch-window leads at PKR 243–602 against PKR 1,845 on the always-on equivalent. |
| Warm beats cold | Video-viewer retargeting at PKR 1,015 against PKR 1,804 on broad prospecting in the same project. |
| The seed list makes the lookalike | A qualified-seed lookalike at PKR 792 on 1,303 leads against a thin-seed build at PKR 1,779 on 31 leads. |
| Consolidation beats fragmentation | A 3-ad-set project held a PKR 887–922 band; a 4-ad-set micro-audience build ranged PKR 1,553–2,062. |
| Engagement and viability are different questions | The highest CTR in the portfolio (1.78%) sat on the worst cost per lead (PKR 5,408) in an expensive market. |
The PKR 10M monthly budget model
| Tranche | Share | Rule |
|---|---|---|
| Local project floors | 70% | Seven projects at PKR 1M each, so no project goes dark and loses its learning. |
| Overseas engine | 20% | Two projects at PKR 1M each, split into per-market budgets and judged on quality, not local CPL. |
| Events and enhancement | 10% | Follows the launch calendar — the cheapest lead-buying windows of the year. |
Because we could not optimise to revenue or target with precision, the two levers that worked were audience architecture and creative. Almost every result in this portfolio came from one of them.
What comes next
- Server-side conversion tracking so the algorithm can optimise toward qualified leads, not form fills.
- Automated CRM write-back to keep lookalike seed lists current every day.
- WhatsApp-first lead capture with instant first response.
- Rebuilt landing pages with payment plans, layouts and construction progress above the fold.
More case studies
Keep reading.
Start a project
Want results like these?
Tell us about your business. You will get a clear plan and a transparent quote within one business day.