How We Made €104 400/month Revenue For Women's Clothing eCommerce Shop
Danil Chernukha
I help eCommerce CEOs & Founders 2x their annual revenue through Meta & TikTok ads
Project Description:?
B2C Ecommerce women's clothing shop
Geolocation:?Europe
Budget:?€19 981
ROAS:?x5.22
Earnings:?€104 400
Client’s background
?Client’s goal:
Get a stable ROAS x3 by the end of the first working month. If all goes well, start scaling up in the second month.
Step One - Analytics
First of all, we started setting up the analytics on the website. Due to some factors and complexities, we decided to set everything up through Google Tag Manager.
There was a lot of work, custom scripts were written to read data from the page, such as value, currency, contents, content_type, content_name, content_category. And of course, the basic events were set: Page View, View Category, View Content, Add To Wishlist, Add To Cart, Initiate Checkout, Purchase.
These parameters are primarily needed for proper data analysis as well as for creating dynamic retargeting through the product catalogue.
The product catalogue was created manually and was updated immediately if the slightest changes were made.
Step Two - Product Selection
Since there was almost no historical data on the campaigns (except for two campaigns with a budget of €150 per day, which client set up without involving experts in the field), we had to run the tests from scratch and rely only on our experience.
Initially, the client with us identified the product categories and the products themselves, which would be primarily used in the advertising campaign. We chose the most popular products of all time with gross margin, a new collection, and promotional offers.
Step Three - Selecting the audiences
The decision was made to start working with the widest possible audience, as the average bill is 200€ and the product itself is not specific. In other words, it could well be a simple emotional purchase without a long warm-up period for the users.
We settled on the following audiences for the Prospection campaign:
- Broad audience without targeting by interests, limited to geo, age 18-50
- Users interested in clothing brands (Prada, Gucci, Louis Vuitton, Burberry, Dolce & Gabbana, Fendi, Chanel, etc.)
- Users interested in fashion and shopping (Shopping and fashion, Clothing, Online shopping, Boutiques, etc.)
-LAL audiences created from users who had already bought the product (the audience was created using a database, we later switched to a LAL audience of users who came from Facebook ads and made a Purchase on the website)
For the retargeting campaign, we decided to use a standard and tried-and-true structure:
Closing warm leads:
- view content 30 days
- add to cart 30 days
- initiate checkout 30 days
For up-sale and cross-sale:
- purchase 30 days
Due to product specifics, it does not make sense to use a window for more than 30 days unless new products and/or any promotional offers are available. The audiences for the 90-day and 180-day windows have been collected but are currently not being used and waiting for their time =D
Step four - Advertising campaign structure
The decision was made for Prospection to put a product category on the ad campaign level, and inside each campaign – ad sets with the same geo and with the audiences described above. Each ad set has 4 ad creatives (static banner, animated banner, carousel, and video)
领英推荐
In this way, we could test both the different product categories and the type of ad creatives and placements that were best to work with. We created one ad campaign for each product category for retargeting, with all the ad sets by funnel stage, as well as a separate ad campaign with dynamic retargeting through the product catalogue. The manual retargeting only used (and currently uses) carousel ad.
Step five - Advertising creatives
Here it's quite simple. We decided to use four types of creatives for the tests:
- Banner
- Animated banner
- Video
- Carousel ad
It was done to reach as wide an audience as possible since Facebook selects types of creatives according to user interests (video creatives, galleries, etc.) when ranking impressions.
On the creatives were depicted:
- Happy woman in clothes
- Unpacking and trying on video
- Product on a plain background
A lot of effort and energy was put into creating the ad creatives. Yes, there is nothing special about them, but you can see the result for yourself: we have a CTR of 1.41%, which is excellent for E-commerce and a fairly high level of engagement.
The result of the test work month
During this period, a suitable advertising campaign structure was developed, product categories, advertising creatives, and user audiences were tested. The average ROAS for the month was 5.22, which is even higher than the set KPI. The data was collected for analysis and we were able to start optimising the advertising campaign.
Follow-up work on the ad campaign
During the second work month, efforts were made to optimise the advertising campaign and to scale it up.
We decided to move the campaigns to CBO. LAL and broad audiences were performing better than interests, but switching off these ad sets would not had been the right decision. They still bring in a good income, even though the other groups bring in more.
The most successful advertising creatives were unpacking and trying on videos and simple carousels, all others were disabled.
The retarget campaigns remained intact, although the manual campaign gave slightly better results than the catalogue.
As a result, after the work on optimisation we have had:
- 3 ad campaigns for Prospection
- 3 ad campaigns for manual retargeting
- 1 ad campaign with dynamic retargeting
We are now monitoring the performance of the campaigns after optimisation and starting to scale up.
Result
??ROAS = 5.22 with an attribution window of 1-day post view and 7-days post click for the first work month
? €104,403.04 test month revenue
? Created a campaign structure that is easy to analyse and scale
? Identified the best placements and types of advertising creatives
? A plan for scaling up and optimising advertising campaigns has been drawn up and partly implemented