Developed a large-scale creative pack (25+ assets)

Diverse content production: We prepared over 25 creatives, ranging from aesthetic carousels for those who appreciate fine details to dynamic UGC videos designed to build trust;

The Result: This high volume of content allowed us to rotate ads constantly, preventing audience burnout. This variety was key to maintaining a high CTR over the long term;

Switching attribution to “Purchase (after click)”

Refining the model: We moved away from broad attribution models (such as View-through), where a sale was credited simply after an ad was displayed. Instead, we focused exclusively on post-click events;

This sent a clear signal to Meta: “We need people who actively click through to the site and buy.” This approach makes ROAS data transparent and allows for scaling only those campaigns that generate real revenue rather than just statistical views;

Launching 8 targeted ad campaigns in 14 days

Fast-track strategy: Instead of a gradual rollout, we opted for a high-velocity launch. All 8 campaigns were optimized for final sales, allowing the system to gather a critical mass of training data within the first two weeks;

The Result: This intensive launch enabled us to immediately identify the highest-converting product categories and regions, filtering out ineffective hypotheses right from the start.

Development of scripts and technical tasks for video production

5+ specialized scripts: We didn’t just order videos; we developed detailed technical briefs focusing on the audience’s pain points and desires: Collection overviews, Trending colors, Expert reviews from nail technicians;

The Result: Having a clear production structure allowed us to generate content that was “pre-loaded” for sales, rather than just creating a pretty picture.

Funnel segmentation (Cold, Retargeting, LAL)

This structure provided the flexibility to distribute the budget strategically—investing more in new client acquisition for growth, while precisely targeting retargeting for maximum ROI;

Gradual scaling of winning combinations

Scaling “Golden” pairs: Once we identified the winning combinations of Audience + Creative + Offer, we began gradually increasing their budgets. We avoided sharp spikes to prevent disrupting Meta’s learning algorithms;

The Result: We successfully multiplied sales volume while maintaining a consistently high return on investment. Every subsequent scaling step was mathematically justified and safeguarded the project’s overall profitability.

Fixing tracking errors in Google Analytics

Full audit and cleanup: We conducted a comprehensive audit of event transmission and identified issues with duplicated or premature conversion tracking. The setup was adjusted so that the “Purchase” event triggers exclusively on the successful payment confirmation page (Thank-you page); The Result: We eliminated “junk” conversions that were confusing Google’s algorithms. Now, the system receives only clean data on actual revenue, allowing bidding strategies to operate with much higher precision by focusing on high-intent, paying audiences;

Correct implementation of Server-Side Tracking

Server-to-server data transmission: We implemented direct server-to-server data transmission instead of relying on the user’s browser. This allows for consistent conversion tracking even for iOS users or those utilizing AdBlock software; The Result: This increased the volume of conversions visible to Google by 20–30%. More data leads to faster campaign learning and the ability to scale budgets more aggressively, as we now have a complete and accurate picture of overall performance.

Launching 7 multi-channel campaigns (Search + PMax + Discovery)

Integrated system approach: We didn’t limit ourselves to a single tool; instead, we launched a comprehensive system. Testing 10 different creatives allowed the algorithm to select the best-performing visual combinations, ensuring a steady flow of orders from multiple traffic sources simultaneously.

Implementation of 100+ recommendations and Quality Score growth

We systematically addressed every Google recommendation, from adding extensions (phone numbers, sitelinks) to optimizing headline relevance for landing pages. The Quality Score increased to its maximum, allowing us to win auctions at higher positions for a lower cost than our competitors. We began generating significantly more traffic while maintaining the same budget.

Implementation of Dynamic Search Ads (DSA)

We configured campaigns that scan the website’s content and automatically generate ads for each specific product. When a customer searches for a particular model or color, Google automatically creates a relevant ad. This allowed us to capture “long-tail” queries that competitors often overlook. We ensured the brand’s presence for every single item in the catalog, leading to a significant increase in orders due to the highly precise match between the user’s search and the ad content.