I build products and help them grow.

Finance is consistently the most expensive app category to acquire in. Every month measured came in under both the Android and APAC averages, and the good months came in an order of magnitude under them. The range is the honest version, because the cheapest month was not the typical month.

Does not prove: The Malaysian figure is a seasonal bidding peak for finance apps, not a market average. It shows how expensive the category gets, not what a typical month costs. The like-for-like comparison is the APAC and Android averages.

Read the units carefully, because they are the point: the Malaysian market rate for a single ad click starts around RM2, and this engine was landing a completed registration for RM2.18, a whole customer for roughly what one click costs at the cheapest end of the market. It fell 28% quarter on quarter while sign-ups grew 36%.

Does not prove: The benchmark prices a click. The result prices a completed sign-up several steps later. No Malaysian cost-per-registration benchmark was published, so this shows scale against an earlier funnel step at market rate. It does not prove a like-for-like sign-up cost was beaten.

The second row that does not flatter. Against a 6.66% cross-industry average the measured range straddles it, with the good months running about half again above and the weak ones below. The channels also differ: this was Google Ads app campaigns, which serve across Search, Display, YouTube and Play, while the benchmark is weighted to search. Treat it as directional, not a like-for-like win.

Does not prove: App campaigns and search-weighted campaigns serve on different inventory, and the benchmark is not Malaysia-specific. This places the range in a healthy band. It does not prove out-performance of comparable campaigns.

Straight, because it does not flatter: the measured range straddles the Play Store average rather than clearing it. The best month ran about a third above it, the weakest month sat slightly below. The two figures also count different things, ad-click-to-install against store-page-view-to-install, and the benchmark is a US average, so treat this as adjacent, not identical.

Does not prove: The figures count different steps in different geographies. One measures ad click to install. The other measures store-page view to install. This is adjacent context, not proof of beating the store average.

Compared against the band fintech and subscription apps actually plan to, not the ecommerce blended average of roughly 2.9, because an annual insurance premium is not an ecommerce basket. The number that matters is net crossing 1.0 in Q4: the quarter the acquisition engine started paying for itself rather than being funded.

Does not prove: The benchmark is a modelled planning target, not a measured market average. This shows the engine reached self-funding against the band fintech apps plan to. It does not prove out-performance of measured ecommerce ROAS, which describes a different business model.

The single ratio that decides whether an acquisition engine is a business or a subsidy. It moved from roughly at the 3:1 rule to comfortably above it inside two quarters, driven by cost per acquisition falling rather than by assuming a longer customer life.

Does not prove: The sample is B2B SaaS, not B2C insurance. The 3:1 rule of thumb travels across business models. The 3.2:1 sample median does not necessarily.

Acquisition cost recovered in under three weeks. This comparison flatters and it is worth saying why: an insurance transaction books its revenue immediately, while a SaaS subscription recovers the same cost a month at a time. The right read is not that this beats SaaS. It is that the model throws off cash fast enough to fund its own growth.

Does not prove: Insurance books its revenue at purchase. SaaS recovers the same cost a month at a time. This does not prove the model beats SaaS. It only shows that the model funds its own growth.

A combined ratio above 100% means the Malaysian motor market was paying out more than it took in, on the line that is the largest in general insurance at 42.8% of premiums. In a market underwriting at a loss, acquisition efficiency and renewal are the only levers left, which is why renewal was built as a product surface rather than handed to marketing.