When Syedewa Cosmetics first came on board, the business had strong products but no growth engine behind them. The brand was scaling slowly and unpredictably, mostly on organic reach and repeat customers. The few ads that were running had no ROAS target, spend was going out with no reliable way to tell if it was profitable. There was no GA4, no GTM, no BigQuery, so decisions were made on gut feel instead of data. And no dedicated team owned any of it.
The starting point: RM500,000 in annual revenue, no paid ads infrastructure, no tracking, and no team.
The Strategy
1. Meta Ads, built around ROAS, not guesswork
The first priority was getting Meta Ads to a point where every ringgit spent could be measured against a clear ROAS target, instead of "boosting posts" and hoping for sales.
- Set a minimum ROAS benchmark before scaling any campaign, so budget only moved toward what was already proven profitable.
- Pulled ad angles and creative directly from customer feedback, reviews, comments, DMs, repeat-purchase reasons, so the copy mirrored what customers were actually saying about the product.
- Scaled spend by product category rather than treating the whole catalogue the same way. Categories with stronger margins and ROAS got budget priority first, then scaling extended outward once each proved itself.
2. TikTok Ads, hero-product strategy
Instead of spreading budget thin across the full catalogue, the strategy concentrated on one hero product at a time, the item with the best organic pull and highest repeat-purchase signal, and used it as the entry point for new customers. This gave TikTok's algorithm a tighter signal to optimise around, and let the brand build a repeatable creative formula on one winner before rolling the same playbook out to the next.
3. Web optimisation, product-page focus
Traffic from Meta and TikTok is only as good as the page it lands on, so product pages became a dedicated optimisation focus, not the whole site. Product-page structure and content were reworked to reduce friction between "landed on page" and "added to cart," prioritising the pages carrying the most paid traffic first.
4. Data infrastructure, GA4, GTM & BigQuery
Before this, there was no reliable way to see what was actually working. This got fixed first, in parallel with the ads work:
- Installed and configured GA4 for proper event and conversion tracking.
- Set up Google Tag Manager so tracking could be managed and expanded without a developer for every change.
- Connected BigQuery to hold and query the data at scale, so performance across Meta, TikTok, and the website could be reviewed in one place instead of three separate dashboards.
This is what made ROAS-based scaling possible in the first place, you can't scale by category or by ROAS target without trustworthy numbers behind it.
5. Team management
Finally, this needed a team to run it. A dedicated in-house team was built around the workflow above: someone owning ads execution and daily ROAS monitoring, someone owning creative (working directly off the customer-feedback loop), and someone owning the data side (GA4/GTM/BigQuery reporting). Instead of one person juggling everything reactively, each part of the system had clear ownership, which is what made consistent year-over-year scaling possible instead of one good year followed by a plateau.
The Results
| Year | Annual Revenue (RM) |
|---|---|
| Year 1 | 500,000 |
| Year 2 | 1,200,000 |
| Year 3 | 2,400,000 |
| Year 4 | 4,800,000 |
| Year 5 | 9,600,000 |
From RM500,000 to RM9.6 million in five years, roughly doubling revenue every year, driven by a ROAS-disciplined Meta Ads system, a focused TikTok hero-product strategy, product-page optimisation, and a GA4/GTM/BigQuery data foundation that made every scaling decision measurable instead of a guess.
