Split tests
Use case
"We cannot split traffic to evaluate A/B test results... but we can deploy Holocene, which takes the A/B testing to the whole new level..."
Conducting A/B testing on marketplaces presents a formidable challenge. The nature of these platforms makes it unfeasible to split traffic and analyze outcomes under varying conditions. Moreover, the highly dynamic environment of marketplaces renders consecutive tests ineffectual, as the rapid changes can skew comparability of results. Holocene offers the precise solution needed for validating A/B hypotheses across various elements, including prices, images, titles, or any other variable you aim to test. Utilizing a proprietary algorithm, Holocene delivers real-time validation of outcomes, enabling you to select the variable of interest while the algorithm manages the intricate analysis.
“
"Holocene Services has developed a special tool called the Dynamic Synthetic Control (DSC). Its main job is to show us what could happen in real scenarios if we didn't make any changes to our current settings.
To put it simply, this tool helps us understand the impact of changing a specific detail, like the price of a product, by comparing it to a scenario where everything stays the same. Since it's tough to predict sales outcomes if prices remain unchanged, the DSC steps in by analyzing similar products in the market. It then creates a virtual model that mirrors the sales of the original product using key factors. When we adjust the price of the actual product, the virtual (synthetic) model doesn't change, allowing us to see the difference in sales results clearly. If any of the virtual model's reference points (like price or product description) changes, the DSC algorithm updates the model to keep the comparison accurate."
Application in economics →
To put it simply, this tool helps us understand the impact of changing a specific detail, like the price of a product, by comparing it to a scenario where everything stays the same. Since it's tough to predict sales outcomes if prices remain unchanged, the DSC steps in by analyzing similar products in the market. It then creates a virtual model that mirrors the sales of the original product using key factors. When we adjust the price of the actual product, the virtual (synthetic) model doesn't change, allowing us to see the difference in sales results clearly. If any of the virtual model's reference points (like price or product description) changes, the DSC algorithm updates the model to keep the comparison accurate."
Application in economics →
Use case: Dynamic Synthetic Control validates incremental gains of dynamic pricing
Case study: fly fishing rod & reel combo
This DSC process was demonstrated in a case study where, after observing a rise in demand, Holocene adjusted the product's price on November 23 to increase revenue. This change caused a spike in sales (shown by the blue line) compared to the steady sales of the unchanged virtual model (the red line). Eventually, as the market balanced out, the DSC adjusted the price again to keep maximizing revenue for the client.
This example shows how Holocene's AI tool can help businesses achieve their revenue and profit goals by making smart, data-driven decisions.
This example shows how Holocene's AI tool can help businesses achieve their revenue and profit goals by making smart, data-driven decisions.
Looking to automate dynamic pricing?
Connect Holocene
By clicking the button you agree to our Privacy Policy