Skip to content
serpico.

Food manufacturing

SEO, Google Ads and a Rozetka Launch for a Food Producer: Google Position 12.6 → 5.5 and a 426-Product Feed

We run SEO, Local SEO, Google Ads and Shopping and brought the product range to Rozetka: a 426-product feed, automated attributes and a pricing formula for products sold by weight. We found an error that showed 97 of 422 products as "in stock" with no stock. The average Google position improved from 12.6 to 5.5, and clicks in June 2026 grew by 28%.

Project type
SEO + Google Ads + marketplace launch for a food producer
Market
Ukraine
Timeframe
February 2024 - present
average Google position, June 2025 vs June 2026
12.6 → 5.5
clicks in June 2026 vs June 2025
+28%
Google clicks in September 2026
30K+
products in the Rozetka feed
426
SEO, Google Ads and a Rozetka Launch for a Food Producer: Google Position 12.6 → 5.5 and a 426-Product Feed

Challenge

The client: a food producer with its own branded stores. We have been working together since February 2024, and over this time the scope has grown from website promotion to marketplace expansion.

What made the task difficult

  • A catalog website and physical stores at the same time. Some customers buy online, while others look for the nearest store. We needed to drive both online sales and store visits.
  • Products sold by weight. The products are sold by weight, but the marketplace expects a fixed price per unit. The wrong pricing logic means either a loss or an inflated price nobody will pay.
  • Marketplace data requirements. Rozetka requires completed attributes: package weight, brand country, country of manufacture and the correct category. Doing this manually for hundreds of products takes weeks.
  • Stock levels. Marketplace shoppers expect an item marked "in stock" to actually be available. Mistakes here lead to canceled orders and a lower seller rating.

The goal: grow the client's own sales channels and safely bring the product range to Rozetka, alongside MauDau.

What we did

SEO and Local SEO

We run SEO for the catalog website and manage Google profiles for the stores, bringing shoppers both to the online store and to the nearest point of sale.

We launched and manage Google Ads and Google Shopping campaigns, where products appear with a photo and price right in the search results.

Launching on Rozetka

  • Prepared a feed of 426 products.
  • Mapped 17 of 46 product range categories to Rozetka categories.
  • Automatically filled in package weight for 204 products, brand country for 62 and country of manufacture for 53.
  • Prices for products sold by weight are calculated with a formula: price per kg × minimum weight, 0.4 kg by default. This way the shopper sees the real unit price, and the producer does not sell at a loss.

A critical stock error

While auditing the feed, we found that 97 of 422 products were shown as "in stock" even though they were not available. The cause: the system treated an empty stock value as available. We fixed the logic, so stock levels now come only from the correct warehouse, and an empty value no longer turns into "in stock".

CRM

For this client we also built a CRM for marketplaces, orders and pricing, as well as other internal tools. They are covered in separate case studies.

Results

Since we started working together in February 2024:

  • The catalog website and stores have ongoing support across SEO, Google Maps, Google Ads and Shopping.
  • A 426-product Rozetka feed with the categories and attributes the marketplace requires.
  • Automation instead of manual work: package weight for 204 products, brand country for 62 and country of manufacture for 53 filled in automatically.
  • Correct prices for products sold by weight using the formula price per kg × minimum weight.
  • A fixed stock error: 97 of 422 products are no longer shown as "in stock" when they are unavailable.

The stock error could have been very costly: almost one in four products in the feed looked available when it was not. On a marketplace, that means canceled orders, unhappy shoppers and a risk to the seller rating. We caught it before it started causing damage.

Search Console data

Over the year the site's average Google position improved from 12.6 (June 2025) to 5.5 (June 2026). Clicks in June 2026 were up 28% year over year (28,674 vs 22,332), and in September 2026 the site received over 30K clicks from Google.

What makes this project special

This is a project where marketing grew into a partnership. We started with SEO and advertising, and today we are also responsible for marketplace expansion and have built custom internal systems for the client. Food manufacturing with products sold by weight has technical pitfalls that regular ecommerce does not, and knowing these details is exactly what makes it possible to enter marketplaces without losses.

Average Google position (lower is better)
03.57.110.614.1Jun 25Jul 25Aug 25Sep 25Oct 25Nov 25Dec 25Jan 26Feb 26Mar 26Apr 26May 26Jun 26Jul 26Aug 26Sep 266.4

We have worked together since February 2024. Search Console keeps 16 months of data. Source: Google Search Console, June 2025 - September 2026

Published: · Updated:

Questions about this project

How do you sell products priced by weight on a marketplace?

The marketplace expects a fixed price, so we calculate it as price per kg × minimum weight, 0.4 kg by default. This way the shopper sees the real unit price, and the producer does not sell at a loss.

What do you need to launch on Rozetka?

A product feed, mapped categories and completed required attributes. For this client we prepared a 426-product feed and mapped 17 of 46 product range categories.

Can you avoid filling in product attributes manually?

Some of them can be filled in automatically based on existing data. Here we automatically filled in package weight for 204 products, brand country for 62 and country of manufacture for 53.

Why are incorrect stock levels dangerous on a marketplace?

A shopper orders an item that is not available, the order gets canceled, and that hurts the seller rating. In this project we found that 97 of 422 products were shown as "in stock" because of an empty stock value, and we fixed the logic.

Do you still need SEO if you sell through marketplaces?

Yes, because your own website and stores remain channels without marketplace commissions. That is why we run SEO for the catalog website, Google profiles for the stores and Google Ads with Shopping in parallel.

Do you build tools for working with marketplaces?

Yes, for this client we built a CRM for marketplaces, orders and pricing, as well as other internal tools. You can read more about them in our development case studies.

More case studies

SEO, GEO and Local SEO for a Multi-Location Medical Clinic: +69% Google Clicks in the First Month
SEOGEO / AI SEOLocal SEO

SEO, GEO and Local SEO for a Multi-Location Medical Clinic: +69% Google Clicks in the First Month

We built a local page structure for cities and symptoms, set up a separate map profile for each location, optimized content for AI answers and restructured Google Ads. In the first month, Google clicks grew 69%, impressions nearly doubled, and indexed pages went from 56 to 816.

Google clicks, March vs. February 2026
+69%
Google Search impressions in a month
+99%
Read case study →
Technical Audit and GEO for an AI SaaS: Google AI Search Impressions Nearly Quadrupled in a Month
SEOGEO / AI SEOSEO audit

Technical Audit and GEO for an AI SaaS: Google AI Search Impressions Nearly Quadrupled in a Month

A technical audit found 26 tutorials that a CMS setting had blocked from indexing. We opened them to search and reworked key pages for the AI answer format. In the first month, impressions in AI Overviews and AI Mode grew from 818 to 3,245, and organic sessions in GA4 rose 235%.

organic sessions, 28 days before launch vs. the last 28 days
+235%
impressions in AI Overviews and AI Mode in the first month
818 → 3,245
Read case study →
GEO and Structured Data Fixes for Ecommerce: +546.7% Impressions in Google AI Features
GEO / AI SEOSEO audit

GEO and Structured Data Fixes for Ecommerce: +546.7% Impressions in Google AI Features

We audited the structured data, fixed markup errors on hundreds of pages, cleaned up the index and reworked content for the AI answer format. In two months, impressions in Google AI features grew from 168,250 to 1,088,226, and the number of pages with valid structured data increased almost tenfold.

impressions in Google AI features
+546.7%
AI features' share of all impressions
3.4% → 26.1%
Read case study →

Want results like these?

Tell us about your goals and we'll estimate your project's growth potential.

Book a call →

Free consultation

Tell us about your project and we'll prepare a growth plan