Anonymised campaigns from live accounts, including the parts that did not work first time. Client names are withheld under NDA — happy to walk through the detail with references on a call.
Launch weekend plus twelve weeks of absorption. The winning angle was the payment plan, not the view — which took two weeks of testing to accept.
Moving income and down-payment questions ahead of contact capture removed most unbankable applicants before the bank ever touched them.
Arabic-first advertorial plus an activity-type filter tripled the share of enquiries genuinely ready to file within 30 days.
Splitting one campaign into eleven community-level funnels cut wasted agent hours and made the numbers legible per team.
Compliance-approved creative and an experience-first question set. Fewer leads, dramatically fewer unsuitable applicants reaching the desk.
Salary banding on step one and employer category on step two. High volume, tight caps, quality held steady across six months.
Source-market campaigns in four languages with a residency and financing-route filter, feeding a Dubai developer's international desk.
Deadline-driven demand with a turnover filter and a service-line router, so consultants received enquiries already sorted by practice area.
Deliberately low volume. Long-form pre-lander, five respectful questions, and an offer of a conversation rather than a download.
No case studies in this category yet.
One case in full, so you can see how a campaign actually behaves rather than only where it ended up.
A Dubai developer with 340 units releasing across three phases. The internal target was a cost per sales-qualified lead well below what they were paying two incumbent vendors, and a contact rate their agents would stop complaining about.
The opening creative led on the view and the address — the things the brochure led on. It produced strong click-through and a poor funnel completion rate: people were interested in looking, not buying. Week two replaced the hero angle with the payment plan structure, which cut click-through by nearly a third and roughly doubled the qualified lead rate.
The original funnel asked for budget first and payment route last. Swapping them moved the hard disqualification earlier: people who could not finance the purchase dropped out at step two instead of step five, which freed budget to buy more of the traffic that could.
By week six the mix looked nothing like the plan. Two regional property portals and one native platform were producing the majority of qualified leads; a well-known international publisher that had looked essential on paper was switched off entirely.
Over 68 days: 1,480 qualified leads delivered, cost per sales-qualified lead down 41% against the client's blended incumbent cost, and a contact rate 3.2× their previous vendor average. Phases two and three opened ahead of schedule.
3,100 placements, six angles, controlled spend. Information gathering.
Payment plan replaces the view as the hero. CTR falls, quality doubles.
Payment route moves from step five to step two.
78% of placements cut. Spend consolidates into three sources.
Daily caps doubled at stable cost. Arabic variants added for GCC buyers.
“The week-two conversation where they told us our own brochure angle was wrong was uncomfortable and completely correct.”
“Our underwriters noticed before our marketing team did. The applications simply started making sense.”
“Eleven separate funnels sounded like overkill. It turned out to be the whole reason the numbers worked.”
Bring your current cost per acquisition and close rate. We will model the funnel shape that improves on it — or tell you if we cannot.
