Case studies3 min read
Every ninth euro comes through a recommendation: suvenir.bg
By the BasketBooster team
The short answer
Between 1 August and 17 November 2026, 17% of the orders at suvenir.bg contained a product the shopper had first seen in a recommendation. Those products make up 11.5% of the store’s revenue. Nobody in the store arranges the recommendations by hand.
This is one store and its real orders, for the period from 1 August to 17 November 2026. We also explain how each number was counted and which of them should be read with care.
The short version
| Number | What it means |
|---|---|
| 11.5% | of revenue comes from products found through a recommendation |
| 17% | of orders contain at least one recommended product |
| 34% | of revenue sits in those orders |
| €209 vs €84 | average order with a recommended product and without one |
The store
suvenir.bg is a Bulgarian online store that sells to shoppers and to wholesale buyers. The catalog is large, and most visitors land straight on a product page from Google or an ad. They see one product, and if it is not the right one, they leave.
The store had no way to show them what else it sells, apart from “similar products” lists arranged by hand that nobody kept up to date.
What we set up
The installation took one day and needed no work from the store’s team. We switched on three blocks:
- “Frequently bought together” on the product page and in the cart. It shows what other customers really buy with this product.
- “Recommended for you”, based on what the visitor has been looking at.
- “Trending now”, for visitors we know nothing about yet.
The system learns from orders and page views by itself. Nobody in the store arranges anything by hand. What each block does is covered in what a recommendation engine is.
The results
Over the three and a half months:
- Recommendations were shown 1.16 million times and got close to 16,000 clicks.
- 17% of orders contain a product the buyer saw in a recommendation and then added to the cart.
- Those products make up 11.5% of the store’s total revenue.
- Orders with a recommended product average €209. The rest average €84.
“Frequently bought together” works best. That block brings in almost two thirds of the revenue from recommendations. It shows a product that goes with the one in front of the customer, at the moment they have already decided to buy.
How the numbers were counted
We count a product as sold through a recommendation only if the buyer saw it in a recommendation block and then bought it in the same visit. We do not credit the whole order to recommendations, only those specific products.
That is why a test with a control group runs alongside: 10% of visitors see no recommendations at all. So far the visitors who do see them buy more often, 2.72% against 2.47%. The difference is not statistically certain yet, because the control group is small, and the test is still running.
We are publishing it anyway. We would rather show the numbers as they are than wait until they look better.
Want to see the same for your store?
We install BasketBooster Recommender for you in one day. The first 30 days are free and no credit card is required. From the first week your dashboard shows how many views, clicks and orders go through recommendations. If you do not like the numbers, we take it off.
Plans are on the pricing page, and the sum for when a plan pays for itself is in what BasketBooster costs and what it returns.
Questions merchants ask
- How do you decide that a sale came through a recommendation?
- The buyer has to see the product in a recommendation block and buy it in the same visit. Only that product is counted, not the rest of the order. If someone sees a recommendation today and buys the product next week, it is not counted.
- Did recommendations really raise the average order from €84 to €209?
- Not all of it. suvenir.bg also sells wholesale, and wholesale buyers look at more products and place larger orders anyway, so they are more likely to end up in the group with a recommended product. The number we stand behind is the 11.5% of revenue, which is counted product by product.
- What does the control group test show?
- One visitor in ten sees no recommendations. So far 2.72% of the visitors with recommendations place an order, against 2.47% of those without. The control group is still too small for that difference to be certain, so we treat it as a direction and keep the test running.
- Will my store get the same numbers?
- We cannot promise that. The result depends on your catalog, your traffic and how many orders you have. That is the reason for the 30-day free period: you see your own numbers in the dashboard before you pay anything.
TagsAverage order valueCross-sellConversion
See it on your own store
Every plan starts with a 30-day free trial. You don’t need a credit card.