Official certified partners of Google & Meta
This was a launch from literally nothing. A new mid-to-premium kids' clothing brand had just opened a Shopify store, with no sales history, no established audience and no ad account with any track record.
Yet the niche is far from empty: budget players from China compete with niche brands that run their own production, so the mid-price segment required clear positioning, not just launching ads.
Our task was to build Meta Ads from scratch, find a creative–audience–website combination that would work effectively and bring the brand to a steady flow of sales — while helping the client remove the technical and communication obstacles, of which there turned out to be quite a few at the start.
The first decision we had to make was the choice of channel. The brand had organic Instagram activity even before the website launched, so at first sales came through two parallel paths: Direct messages and the new website. We tested both and within a few weeks deliberately shut down the Direct funnel — lead quality there was lower, and the CPM was unfavorable for a new brand's small budget. We moved the entire focus to the sales funnel through the website.
Next came the search for a working creative format. We went through product cards, outfit sets, carousels and UGC videos (unboxings, reviews from kids and parents) — and it was the catalog format, a dynamic product catalog via Meta Commerce Manager connected to Shopify, that consistently showed the best return (ROAS 200–300%+). It became the main working combination, while other formats were tested in parallel as a second scaling track.
The technical side of the project turned out to be just as important as the advertising. The site's checkout ran on a third-party solution that initially didn't pass correct data to analytics at all — GA4 only saw sessions from the checkout stage, not from the whole site. We worked out the cross-domain connection between Shopify, the checkout app and GA4 ourselves and made sure analytics showed the real picture across the entire site.
A separate technical puzzle was that the product catalog didn't sync directly to Meta because of how Shopify stores outside the US work. We prepared a workaround for the client: a hidden collection on the site plus a product set in Meta Commerce Manager filtered by a unique product tag. This let us run catalog ads without showing the technical collection to site visitors.
We also carried out an in-depth target audience analysis and market research — we studied competitors and identified about 20 specific fears and barriers that stop people from buying kids' clothing online. The main reasons that held people back from buying were:
- fear of getting the size wrong,
- distrust of a new brand requiring prepayment,
- a psychological price threshold.
We gave the client concrete recommendations on positioning, how to communicate discounts and product page content.
The most critical moment of the project came at the end of July, when ROAS suddenly dropped from 157% to 100%, and checkout-to-purchase conversion fell from 10% to 2%. Instead of guessing, we did a detailed funnel breakdown and matched the dates of the drop with the dates of the latest changes to the website. It turned out that fixing one technical bug had accidentally created a new one (in the checkout form autofill). We told the client the cause directly and recommended taking the site's UI/UX seriously, up to bringing in an external specialist for a paid audit. We resolved differences in how we saw priorities (website vs. ads) with arguments and numbers, which preserved the client's trust in us.
In parallel, we worked systematically on the main bottleneck of the funnel — the “ad click → add to cart” step, which stayed at 3.5–5.5%. We tested optimizing campaigns for the “add to cart” event instead of “purchase” (to give the algorithm more learning signal with a small number of sales) and gave recommendations on the product page: a consistent photo structure, clear fabric specifications and fewer empty clicks on the size chart.
The numbers show the classic trajectory of a project that started from zero: a drop, diagnosis, recovery and growth. The deepest dip (100% ROAS) at the end of July coincided with technical changes on the website; after diagnosing and partially fixing the issues, results began to recover steadily.
In August, ROAS grew from 106% to 332%, weekly purchases increased from 4 to 22, and the cost per purchase fell more than threefold — from 1,517 UAH to 473 UAH. The catalog format remained consistently the best-performing for return throughout the period, and the “click → cart” conversion rate rose to 5.56% by the end of August — the best figure of the entire collaboration.
The main takeaway from this case for our team and for clients entering the market from scratch: a successful new brand launch is systematic work on several fronts at once. Dropping a weaker funnel in favor of a stronger one, consistently searching for the most effective ad creative format, quickly diagnosing problems by comparing data rather than guessing, and reasoned communication with the client even in difficult moments — all of this together gave us the result: ROAS grew from 106% to 332% in a single month, and the cost per purchase dropped more than threefold. The project is ongoing, and our team continues to scale the funnel that has proven its effectiveness.
