Vita4you.
How custom-built software — Symfony, React, and PostgreSQL — generated a full Google Ads account from the Merchant Center feed and more than doubled revenue for one of Greece's largest online pharmacies.
Scaling ad coverage from a handful of brands to the full pharmacy catalog — fast
Vita4you is a Thessaloniki pharmacy that has operated from the city center since 1987 and moved online in 2010. It now stands as one of the largest online pharmacies in Greece, with 25+ employees including in-house content marketing and web development teams. By 2019 the brand had strong fundamentals; what it lacked was ad coverage across its full catalog.
The brief demanded a sharp increase: +40% revenue growth within a specific profit margin, in a compressed timeframe. Analysis of the existing ad accounts revealed the root cause — advertising activity was concentrated on a small number of products, brands, and categories, while the long tail of the catalog was either unadvertised or running campaigns that did not cover their own cost. The majority of SKUs were earning zero ad support.
To drive the revenue lift in weeks rather than quarters, the project needed campaigns for every category and every brand Vita4you sells — fully optimized from day one. Building that account manually would have taken months, during which the brand would keep losing the very revenue the brief was asking it to capture. The answer had to be programmatic.
Custom software that generates the account and mines the search layer
We built a custom software platform on Symfony 2.0 (backend), React.js (frontend), and PostgreSQL (data layer). The platform has two main modules, each addressing one half of the problem.
The first module runs in the cloud. It ingests categories directly from the Google Merchant Center feed and automatically creates the matching campaigns, ad groups, keywords, and ads. Every SKU in the catalog gets ad coverage on launch day — not via manual copy-paste but via deterministic rules applied to the feed. When Vita4you adds a new brand or category, the module regenerates the corresponding account slice without touching the rest.
The second module focuses on search discovery. It mines search queries that hit Google Search using existing account segments as triggers, filters the output intelligently (removing duplicates, obvious low-intent terms, and already-covered queries), and surfaces suggestions for account expansion. What a manual analyst would need days to compile, the module produces in minutes — and the filtering logic keeps the signal-to-noise ratio high enough that a manager can act on every suggestion.
Build the platform, not the campaign
Platform Foundation
Chose Symfony 2.0 + React.js + PostgreSQL as the stack — mature, well-supported, and fit for long-running systems that must stay reliable as the account grows. Designed with clear module boundaries: one for account generation, one for search mining, shared data layer.
Account Generation Module
Feed-to-account module reads the Merchant Center export and emits the full Google Ads structure programmatically — campaigns grouped by category, ad groups per subcategory, keyword pools from product titles, ad copy templated against brand and category attributes. New brands propagate automatically.
Search Mining Module
Query-mining module pulls search terms from Google Search, filters against duplicates and low-intent patterns, and ranks recommendations by expected impact. Account managers approve from a short list each morning instead of hunting through thousands of raw queries.
The Results
Revenue Growth
Revenue grew 135% in the window, more than 3x the +40% brief. The lift came from catalog-wide ad coverage: every SKU that previously had no ad support suddenly had fully optimized campaigns from day one.
Transactions
Transaction volume expanded 125% — the order count that follows when ad coverage flips from a handful of brands to the entire catalog. Average order value stayed stable, suggesting the new audiences were high-quality.
Ad Spend
Ad spend more than doubled to fund the expansion, but it stayed within the target profit margin throughout — the filtering logic in the search-mining module kept low-intent terms out of the account.
Months to Full Catalog
Full catalog ad coverage achieved in two months end-to-end — the kind of timeline that would have taken an entire team a year of manual work. The platform remains in production, regenerating account slices whenever the catalog changes.
More Case Studies
View all
Ready to scale ad coverage as fast as your catalog?
Let's build a feed-to-account platform that turns catalog expansion into a build command, not a quarter of manual work.
Start a conversation