Experiments ROI dashboard
This feature is currently in early access.
This tutorial guides you through the Experiments ROI Dashboard: AB Tasty's strategic planning tool for measuring and projecting the financial impact of your experimentation program. You'll learn how to use its advanced features including accuracy settings, quarterly analysis, and revenue projections powered by RevenueIQ™.
What's new
The Experiments ROI Dashboard uses an optimized calculation method that ensures every euro/dollar of revenue is counted exactly once, even when customers are exposed to multiple experiments simultaneously. This gives you a true picture of your experimentation campaigns financial impact.
Accessing the dashboard
In the main AB Tasty navigation bar, click on Strategy
Select Impact Dashboard, then Experiments ROI
The three main indicators of the dashboard are:
1) Global experiments ROI
It’s the sum of the testing phase revenue impact and the uplift that will be generated overtime by your winning variations from Web Experimentation campaigns only.
2) Testing phase revenue impact
It’s the sum of each quarter revenue impact based on real production data.
3) Expected winners uplift overtime
It’s the sum of each quarter winning variations uplift over the selected month based on RevenueIQ most probable scenario.
ROI impact
A green display indicates that running the campaign positively impacted revenue generation.
A yellow display indicates a negative impact on revenue.
Understanding accuracy settings
Testing phase impact calculation
You can select the Testing phase impact calculation method
ROI optimized data
Data is attributed homogeneously between campaigns to deduplicate transactions and perfectly represent the real business impact.
Report standard data
As in A/B test reports, it doesn’t use an attribution strategy to manage parallel running campaigns. This approach is optimal for A/B test analysis but not suited for ROI calculation as it can’t be summed up safely.
Calculation rules
To ensure accuracy, these rules apply:
Campaign types: A/B tests and multipage campaigns
Statistical threshold: Only uplifts with 95% or higher chance to win are counted
Expected uplift calculation
Uplift Estimated (Time Period)
This dropdown controls how far into the future you want to project the revenue uplift. You have three options:
Over 1 month - Projects uplift for the next month
Over 3 months (default) - Projects uplift for the next quarter
Over 6 months - Projects uplift for the next half-year
Winner Selection Method
This dropdown controls which winning experiments are included in your projection calculations:
1) Best winner of each quarter only (Default) Only the single best-performing winner from each quarter is included in projections.
The system identifies the experiment with the highest projected monthly ROI in each quarter. Only that "best winner" is used for the projection calculation and the other winning experiments are visually "dimmed" in the campaign list You can use it When you want to
Focus on your top performers only.
Budget planning where you want the most reliable single number
When explaining to stakeholders which test had the biggest impact
2) All winners
Every winning experiment from each quarter is included in projections. The system sums the projected monthly ROI from ALL winning variations. All winners are displayed in green.
Last updated
Was this helpful?