Case Princess Furs

#E-com

The project is to launch and scale advertising for the princess furs brand after the new website goes live. The goal is to stabilize traffic, go through the strategy learning phase, and get leads from search during the fur sales season.

Problems and Tasks

Problems

A new website meant campaigns were effectively launched from scratch, requiring strategies to go through the learning phase again.
Advertising launches were inconsistent: one week active, one week paused, which prevented the accumulation of sufficient data.
Part of the season (November and half of December) was missed.
Advertising stopped due to budget depletion, which reset the learning phase.
Low client confidence due to the lack of quick sales.

Traffic Facts

12 conversions
CPA — ₽2,346
470+ users viewed more than 3 pages in 4 days
Bounce rate:Search — 22.6%, Ad networks — 26.5%
On one of the days: 39 visitors

Tasks

Stabilize the advertising budget.
Focus on high-intent demand.
Test different strategies (automatic vs manual).
Reduce cost per click by narrowing the keyword semantics.

Growth plan

Hypotheses for Achieving Results

Focus only on search traffic (disable the Yandex Advertising Network / RSYA) to capture high-intent demand.
Narrow keyword semantics to phrases like “buy a fur coat”, “buy a mink fur coat”, etc., to reduce CPC.
Test automatic strategies for low-cost clicks, then switch to manual bidding with restricted keywords.
Run campaigns without pauses to prevent algorithm learning resets.

How the Work Plan Was Implemented

Campaigns were switched to manual bidding during the launch of the new website.
Preparation for Strategy Learning
Offer adjustments:
- Reduced the advertised discount from 40% to 20%.
- Updated size availability in ads (EU 38–60).
User behavior analysis via Yandex Metrica and Webvisor.
RSYA disabled → testing search traffic only.
Increased search budget to ₽20,000.
Step-by-Step Testing
- Automatic strategy (focused on low-cost clicks).
- Manual strategy with narrow keyword semantics.
- Reintroduction of RSYA with limited queries.
- Preparation for auto-targeting in case of no leads.