Blog › Campaign Case Studies

126 Property Enquiries for KES 40,101: Why the Leak Was the Inbox, Not the Ads

By Joseph Maina · · 4 min read

Hand holding a bunch of house keys

In October 2026, a residential developer spent KES 40,101 on Meta ads and received 126 enquiries on Instagram and lead forms. On the dashboard that is KES 318 each, a respectable figure. The more useful number appeared after we read every chat: 14 serious buyers, 6 of them with a phone number. The client, the buyers and the exact unit are left out on purpose.

The month in numbers

These are the figures for 1 to 31 October. Cost per step is the month's ad spend divided by the people at that step.

October 2026 at a glance
spent, about KES 1,200 a day
KES 40,101
enquiries at KES 318 each
126
hot buyers at KES 2,864 each
14
gave a phone number
6

Where the 126 enquiries went

The ads found the right people. 106 of the 126 were real buyers or renters asking about payment plans, mortgages and site visits. The loss happened after that.

From enquiry to phone number
  1. Messaged the page: 126
  2. Real buyer or renter: 106
  3. Showed real interest: 37
  4. Hot buyers: 14
  5. Gave a phone number: 6

The leak was in the inbox

60 people tapped a quick question on an ad and never replied, because the answers did not end with a next step. Many first replies arrived the next day, and two chats were never answered. By then, a buyer comparing several developments had often moved on.

The ads were also getting better. Cost per Instagram chat fell from KES 478 in the first week to KES 140 in the last three days, against a month average of KES 309.

What changes in November

Same budget, different handling of what the ads bring in.

  • Replies within an hour, ending in one question: "Which day can you come for a site visit?"
  • Quick questions on the ad that lead straight to a visit booking.
  • One sheet tracking every lead from chat to visit to deposit.
  • A hot, warm or cold score on every conversation, as set out in my qualification guide.

What this does and does not show

This is October only. Site visits and sales were not tracked yet, so it shows cost per hot buyer, not cost per sale. The month's detail also comes from reading chats by hand, which involves judgement about who counts as hot. November results will be added to the full case study when they exist.

Before you change an ad that brings enquiries, read the chats. The problem may be the reply, not the ad.

Related articles