---
title: "Pakketo"
url: "https://www.omnicliq.com/en/case-studies/pakketo-ai-product-decisioning/"
description: "How an AI system that decides which products deserve ads brought Pakketo more revenue from fewer visits, over a full year."
category: "ecommerce"
client: "Pakketo"
datePublished: "2026-10-28"
author: "Dimitris Bachtsevanis"
locale: "en"
---

# Pakketo

> How an AI system that decides which products deserve ads brought Pakketo more revenue from fewer visits, over a full year.

## Project

**Client:** Pakketo
**Industry:** Furniture & home
**Duration:** 06/2025 – 05/2026
**Services:** ML & AI, Google Ads, Analytics
## Key results

- **+46%** Revenue
- **+24%** Google ROAS
- **+64%** Ad revenue
- **+42%** Orders
## A large range, different demand per product and one budget for all

Pakketo sells furniture and home décor online. Someone looking for a specific piece wants something different from someone browsing ideas for their home, or someone comparing prices and dimensions.
The goal for a full year was more value from Google Ads and more e-shop revenue.
If campaigns give equal weight to the whole range, budget is lost on products that bring visits but not sales. In furniture this happens often, because customers view many items before buying one.
With thousands of products, manual selection cannot keep up with shifts in demand. We needed a decision per product, every day.

## An AI system that decides every day which products get priority

We combined visitor behaviour per product (GA4), ad performance and feed data (category, price, availability).
A forecasting model (MindsDB) gives every product a commercial momentum score, based on what it shows now and not only on its history.
The score becomes a feed label, and Performance Max, Search and Demand Gen put weight on the products most likely to sell, even before that shows in sales.
Products with weak signals no longer absorb the budget of strong ones. The score refreshes daily, so campaigns follow demand all year round.

## Results

- **Revenue** — More online revenue in a year, from fewer visits.
- **Google Ads ROAS** — More revenue for every euro in Google Ads.
- **Google Ads revenue** — More revenue from the Performance Max, Search and Demand Gen campaigns in Google Ads.
- **Orders** — More purchases over the year.