Case study
Running the whole growth system on a store of our own
A Swedish fashion store we built, funded and scaled ourselves — SEK 1.47M in total sales and 3,409 orders across seven months, with about 70 percent of sales coming from Facebook and Instagram.
The challenge
No audience, no data, no reason for anyone to trust a new name
Svea Essentials launched into Swedish fashion with nothing behind it — no customer list, no pixel history, no organic reach and no reputation. Everything the store needed had to be bought, and bought cheaply enough that the learning paid for itself.
-
Cold start
No audience data meant every early impression was paid for at an untrained price.
-
Trust
A new Swedish store had to answer shipping, returns and payment questions before it could ask for an order.
-
Margin
Buying an audience at launch prices leaves very little room per order.
-
Proof
Nothing could be claimed about the system until the store had traded a full season.
The approach
We used a cheap product to buy the audience, then sold that audience a catalogue worth four times as much
-
Website
Sold from the product page, not the homepage
Meta sent cold traffic straight to the product page, so each one had to carry the whole sale — Klarna, PostNord tracking and free shipping stated in the buy box.
-
Ads
Subsidised the first product to train the account
One high-repeatability product ran at a 59 percent discount until Meta had learned the buyer. The margin we gave up there was the price of the audience data.
-
Checkout
Grew the basket instead of cutting the price
Quantity bundles carried orders to 2.52 items each, so the offer worked on volume rather than on margin.
Subsidised → trained → swapped the catalogue
Once the account knew the buyer, a 617 kr winter catalogue replaced the 143 kr launch product and discounting fell to 8.6 percent. November became the best month of the year.
The work
Meta did the finding. The catalogue did the earning
Facebook and Instagram delivered 82,500 of the store's 117,516 sessions, and the landing pages confirm the shape of it — three product pages took the overwhelming majority of the traffic. Once the account was trained, the winter catalogue replaced the launch product and the discounting was pulled back rather than repeated.
- Facebook sessions
- 56.9K
- Instagram sessions
- 25.6K
The results
The best month was the one with the least discounting
November took SEK 382,694 across 666 orders with the discount rate down to 8.6 percent, against 50 to 60 percent through the launch summer. Average order value rose from SEK 378 in July to SEK 575 in November — the record month was won on what was sold, not on how much of it.
Measured in Shopify analytics across 1 June – 31 December 2024, the store's full active trading span. Total sales are stated after discounts and returns; gross sales before them were SEK 2,829,276. Meta share is Facebook and Instagram attributed sales as a proportion of total sales.
See what this would look like for your organisation
Tell us what you are trying to change. We will be direct about scope, evidence and whether we are the right fit.