From zero to #1 shopping app in Singapore in 6 months.
An unknown Chinese retailer. No tracking infrastructure. No TikTok benchmarks. We built a hypothesis-driven testing engine, and made Vipshop the #1 shopping app in Singapore.
No data. No benchmarks. Compliance delays. Unknown brand.
Vipshop's Singapore launch faced four simultaneous problems: an unknown brand to local audiences, no app store ranking history, no tracking infrastructure (delayed by global compliance review), and zero industry benchmarks for what "good" looked like on TikTok in this category.
Conversion data wouldn't be available for weeks. Competitors were already established. Every day of delay meant audience capture was harder, not easier.
The default approach. Wait for tracking, then optimise, would mean missing the launch window entirely.
Hypothesis-driven testing. Sequential funnel objectives. Cross-channel scaling.
Instead of guessing, we systematised. Every spend decision became a test. Every test had a hypothesis. Winners scaled across channels.
Set up a structured A/B testing engine: micro vs macro KOLs. Product-focused vs storytelling content. Each variant ran with controlled budget and clear success criteria. Within weeks we had a creative winners playbook backed by signal, not opinion.
Layered campaign objectives: video views → landing page views → app installs. Each stage built audience signal for the next. Audiences pre-qualified themselves before the install ask.
Winning TikTok creative was repurposed into Meta and Google UAC. Once we knew what worked on TikTok, we ran the same hooks at scale across the funnel.
While conversion tracking lagged, we optimised against leading indicators (view-through rates, profile visits, app store impressions) so the campaigns kept compounding while we waited for tracking to come online.
#1 shopping app. Top-3 free app overall. 27M views.
Generated across TikTok in 6 months. Pure top-of-funnel saturation that built brand recognition Vipshop didn't have when they started.
Acquired through the systematised funnel. Cost-per-install stayed below benchmark while volume scaled month-over-month.
Plus top-3 free app overall (alongside ChatGPT and TikTok). From unknown to category leader in two quarters.
