---
title: "Pakketo"
url: "https://www.omnicliq.com/en/case-studies/pakketo-black-friday/"
description: "How a demand forecast showed Pakketo how to set up Black Friday before it began, without announcing the discount early."
category: "ecommerce"
client: "Pakketo"
datePublished: "2026-04-18"
author: "Dimitris Bachtsevanis"
locale: "en"
---

# Pakketo

> How a demand forecast showed Pakketo how to set up Black Friday before it began, without announcing the discount early.

## Project

**Client:** Pakketo
**Industry:** Furniture & home
**Duration:** Black Friday 2019
**Services:** MMM, ML & AI, Google Ads
## Key results

- **+83%** Revenue
- **+63%** Google ROAS
- **+109%** Orders
- **2** Years' data
## More Black Friday revenue, without freezing the weeks before

Pakketo sells furniture and homeware online across Greece. For Black Friday 2019 the goal was much more revenue, with better advertising efficiency.
The easy route, announcing the offers early, has a cost in furniture: anyone who knows discounts are coming stops buying, and the weeks before the event are lost.
Furniture is a high-value, considered purchase. Demand is not spread evenly across days and categories, and competition in ads jumps sharply in the same week.
If budgets and bids were adjusted on the fly, they would always be a step behind. We had to know before it started how demand would move, day by day and category by category.

## Before the week began, we knew which days and categories would sell

We built a demand forecasting model (R and Facebook Prophet) from two years of sales per category, device and audience.
The model accounts for holidays, seasonality and discount depth, and produces scenarios per day and category for revenue, traffic and conversions.
Every decision came from the forecast: how budget is split across the week, how bids are set before competition rises, and which categories get stock and creatives.
Communication before Black Friday stayed on products, not prices, so sales did not freeze. During the week, actual numbers were compared with the forecast and budget was corrected immediately.

## Results

- **Revenue** — More revenue in Black Friday week, without sales dropping in the weeks before.
- **Google Ads ROAS** — More revenue for every euro in Google Ads, because bids were set before the competition rose.
- **Orders** — Twice the purchases: new buyers, not demand that simply came earlier.
- **Years of sales history** — Two years of sales history fed the forecast that Black Friday was built on.