# Cost per approved loan down 41% with server-side tracking
> Source: https://shakalakaa.my/case-studies/loan-attribution-meta-capi-malaysia · Updated: 2026-11-16 · Market: MY
JCL Credit Leasing Sdn Bhd

# Cost per approved loan down 41% with server-side tracking

A licensed Malaysian lender moved its Meta tracking server-side and cut cost per approved loan from RM410 to RM242.         Last updated November 2026 Quick answer

shakalakaa moved JCL Credit Leasing's Meta tracking server-side: the Conversions API through server-side Google Tag Manager, an offline conversion sync from its loan CRM and a Looker Studio attribution dashboard.

Between May and August 2026, cost per approved loan fell from RM410 to RM242, and the lead-to-approval qualification rate rose from 6.8% to 14.9%.     Client JCL Credit Leasing Sdn Bhd  Industry Financial services  Market Malaysia  Period May to August 2026  Services Meta Conversions API, offline conversion sync, attribution dashboard  Build 3 weeks (4–22 May 2026)  Measured to 21 August 2026

## The client

JCL Credit Leasing Sdn Bhd is a licensed lender in Malaysia.

## The problem

Before the build, Meta could see 61% of the conversions the CRM recorded, and the Event Match Quality score in Meta Events Manager was 4.4.

## What we did

1. **Server-side tracking.** Moved Meta tracking to server-side Google Tag Manager and the Meta Conversions API, so conversions reach Meta from the server, not only from the browser.
2. **Offline conversion sync.** CRM webhooks send each lead's later stages back to Meta as offline conversions, so Meta can see which leads went on to an approved loan.
3. **Attribution dashboard.** GA4 and the CRM feed a Looker Studio dashboard with Slack alerts, which replaced the weekly report built by hand.

Connected systems: server-side GTM, the Meta Conversions API with offline sync, CRM webhooks, the WhatsApp Business API, GA4, Looker Studio and Slack.

- −41% cost per approved loan Ads Manager + disbursement CRM
- +119% lead-to-approval qualification rate Internal loan CRM
- 4.1× return on ad spend, from 2.4× Looker dashboard
- 96% of conversions now visible to Meta Events Manager vs CRM

## Results (May to August 2026)

Metric · Before · After

1. Event Match Quality (Meta) Before: 4.4 8.6 (+4.2)
2. Conversions Meta can see Before: 61% 96% (+35 pts)
3. Lead-to-approval qualification rate Before: 6.8% 14.9% (+119%)
4. Cost per approved loan Before: RM410 RM242 (−41%)
5. Return on ad spend (ROAS) Before: 2.4× 4.1× (+71%)
6. Weekly reporting hours Before: 7.5 h 0 h (automated)

Measured to 21 August 2026: over 2 months for Event Match Quality, conversions Meta can see and reporting hours, and over 3 months for qualification rate, cost per approved loan and ROAS. Sources: Meta Events Manager, Events Manager compared with the CRM, the internal loan CRM, Ads Manager with the disbursement CRM, and the Looker dashboard.

Reporting hours are the client's operations estimate.

## What this means for your lending business

The plain-language version. Once Meta could see almost every conversion, including which leads became approved loans, each approved loan cost RM242 instead of RM410.       Client results

## Work we can show
  [Proven Wealth Advisory Wealth advisory · trust and estate planning · Meta Ads, Landing page14 → 62 booked consultations a month · month 1 to month 6 Problem-aware video, a fast dedicated landing page, a multi-step quiz that qualifies on net worth, and self-serve booking with WhatsApp reminders.](https://shakalakaa.my/case-studies/wealth-advisory-lead-funnel-malaysia)[Sunway Serene 2 Property · Sunway Property · Meta Ads, Landing page100–180 leads a month Meta ads into a dedicated landing page with a qualifying lead form, and the Conversions API sending real enquiries back to Meta.](https://shakalakaa.my/case-studies/property-developer-lead-funnel-malaysia)[MechCool Engineering HVAC and M&E contractor · Malaysia · WhatsApp AI quoting, n8n, SQL Accounting18 min site inspection to quote, from 2.5 days · Aug–Nov 2026 WhatsApp quoting from a nameplate photo and a voice note, branded PDF quotes and an n8n sync to SQL Accounting.](https://shakalakaa.my/case-studies/hvac-quotation-automation-malaysia)  [All case studies](https://shakalakaa.my/case-studies)       In depth

## Meta conversions api case study, in depth

The supporting detail behind each section of this page, by topic. 4 topics

### The client

It finds applicants through Meta ads and tracks every application in its own loan CRM, through to approval and disbursement.

### The problem

In the same period, 6.8% of leads qualified through to approval, each approved loan cost RM410 and return on ad spend was 2.4×. Reporting was assembled by hand, about 7.5 hours every week.

### What we did

**What we did**

Built in 3 weeks, from 4 to 22 May 2026.

### What this means for your lending business

The reporting that took 7.5 hours a week now runs on its own, from the same data.

What this says about shakalakaa: for a regulated lender, the tracking under the ads matters as much as the ads. It is the measurement layer behind our [Meta ads](https://shakalakaa.my/services/meta-ads) and [performance marketing](https://shakalakaa.my/services/performance-marketing) work, and the reason our [financial services marketing](https://shakalakaa.my/industries/financial-services-marketing) reports in approved loans, not form fills.       FAQ

## Frequently asked questions

### What did shakalakaa build for JCL Credit Leasing?

Server-side Meta tracking through server-side Google Tag Manager and the Meta Conversions API, an offline conversion sync from the loan CRM, and a Looker Studio attribution dashboard with Slack alerts. The build took 3 weeks.

### What changed in the results?

Over 3 months, cost per approved loan fell from RM410 to RM242 (−41%), the lead-to-approval qualification rate rose from 6.8% to 14.9% and ROAS from 2.4× to 4.1×. Over 2 months, the share of conversions Meta could see rose from 61% to 96%.

### Why does server-side tracking lower cost per approved loan?

Meta optimises towards the conversions it can see. When the CRM sends back which leads reached an approved loan, Meta has the right outcome to learn from, not only form submissions.        Let's talk

## Let's start the convo.

### Want results like this for your own lending business?
  [Book a free strategy call](https://shakalakaa.my/contact)[Chat on WhatsApp↗ (opens in a new tab)](https://wa.me/60327700355?text=Hi%20shakalakaa%2C%20I%27m%20enquiring%20from%20your%20Meta%20Conversions%20API%20Case%20Study%20for%20a%20Lender%20page%20%28MY%29.%20Ref%3A%20my-site)
