Shopify has expanded the feature Rollouts to allow merchants to include discounts in controlled experiments, exposing different offers to portions of traffic and comparing the effect on conversion before expanding a promotion to the entire store.
The change turns a traditional promotion tool into a test variable within the platform's own infrastructure. The merchant can, for example, enable a discount only for part of the visitors and keep the current experience for another group, tracking the results before deciding which configuration to adopt.
In experiments, Shopify splits visitors between control and treatment. The control group remains subject to the store's normal discount behavior, while the treatment receives the changes defined in the Rollout.
New experiments use, by default, a traffic split of 50% for each group, although this setting can be adjusted.
The tool also allows coordinating the discount with changes to the theme and checkout, making different components of a campaign go live at the same time. Intended uses include gradual releases of offers, seasonal campaigns and tests before a permanent implementation.
Conversion now guides the choice of offer
After a Rollout begins, the dashboard gathers metrics such as orders, average order value, total sales, store conversion rate and checkout conversion, depending on the changes included in the experiment.
This makes it possible to compare the behavior of the group that received the new offer with the rest of the traffic and assess whether the discount actually improves business performance.
There is, however, an important limitation. Rollouts with discounts do not provide specific metrics for revenue or orders directly attributed to the discount. Results are evaluated using the experiment's general metrics, which requires caution when interpreting the isolated impact of a promotion.
Access also varies by plan. General use of Rollouts is available starting with Basic, while experiments with control and treatment groups require the plan Grow or higher.
Tests that reach only a portion of visitors are restricted to the online store and can create differences between the offers seen on the site and those displayed in external channels, such as ads and product feeds.
The expansion brings discount management closer to a logic of continuous experimentation. Instead of launching a promotion to all traffic and evaluating the result afterward, the merchant can now test the offer at a limited scale and use conversion data to decide whether it should be kept or expanded.
