About
Ebiketuner is a highly rated online shop for eBike riders.
Founded in early 2020, they have quickly gained the trust of thousands of loyal riders worldwide by providing the best eBike tuning kits, GPS Trackers and accessories.
The Challenge
George, founder and operator of Ebiketuner, had little experience changing his prices. With many products in his Shopify store and faced with higher ad costs and unit costs, it was difficult to determine which price level would work best for his customer base and his margin.
However, George did not have past data on price changes and related sales, which meant it was not possible to calculate his optimum price points.
George turned to manually changing his pricing but realized that this trade off was a difficult balancing act – make too hard of a price increase and you churn customers deteriorating both bottom and topline, make too conservative a price increase and you end up leaving profit on the table.
George reached out to Cartsmart to help experiment with his pricing and recommend what the right price levels should be to meet his goals.
Our Approach
Every business has their own challenges and goals. In the case of ebiketuner, they also had distinct goals for each product as well as requirements around sell-through-rate and competitive positioning, especially for certain hero SKUs.
"While it is important for us to maintain and increase our profitability over time, we also have to make sure that we do not compromise on the sales volume by unit. We needed a price plan that could hit two birds with one stone!” - George
Cartsmart’s AI software broke up the problem into four distinct parts –
Experimentation: First, we rapidly simulated small price changes in his key products to generate order and conversion data (to calculate elasticities)
Apply Business Logic: After price elasticities were calculated, Cartsmart applied ebiketuner's unique business goals and constraints. In George’s case there were heavy competitor pressures which we incorporated into the model.
A/B Test: Utilizing price elasticities, this guided to a range of price points which would theoretically work for George. We then A/B tested these price points at random intervals. Using real-time sales data, this enabled us to see which pricing per SKU performed the best on George's overall store
Reinforcement Learning: Now with further data, this enabled further calculations for pricing ranges. We used these in further A/B tests to improve pricing further over time (whilst removing the price points i.e. the A in A/B, that were less efficient in the following tests).
The Results
At the end of the observation period, George hit well over his sales goals with a 22% uplift but also conversion increased +0.4%.
The increase in conversion was from timed price optimisations when his best customers were on his store.
If you’re interested in seeing what pricing strategy works the best on any channel you sell on, book a call with us to chat about how Cartsmart can help.
About the Author
Ben is the co-founder and CTO of CartSmart. Before building A/B pricing tests for e-commerce stores, he was solving similar problems for John Lewis and Just Eat.