The Advora toolkit
Plan a controlled creative test with a hypothesis, control, variants, fixed conditions, budget allocation and a decision rule. Export to Excel.
Choose one variable, define the control and variants, and pre-commit to a decision rule. The allocation is a planning split, not a guarantee that an ad platform will spend equally.
Works in your browser · No account requiredLoading helpful votes…
02 / THE EXPLAINER
The method behind the result.
Write a falsifiable hypothesis and select one variable, such as hook, visual or offer. Describe the control and each distinct variant. If several elements change together, the result cannot isolate which change caused the difference.
Record the primary metric, test duration, sample requirement, minimum meaningful improvement and guardrails. Hold audience, placement, landing page and bidding conditions as constant as practical. Use randomized, non-overlapping groups when the platform supports them.
The worksheet divides the entered budget evenly across the control and variants and across the entered days. This is a planning allocation; normal platform delivery may be uneven. Do not declare a winner from early fluctuations. Testing several variants needs a multiple-comparison method beyond a simple two-arm significance check.
03 / A LITTLE MORE CLARITY
What to know before you put your result to work.
One control plus up to nine distinct variants.
No. You must supply the sample requirement and decision rule for your metric and expected effect.
Not necessarily. Equal allocation here is a planning assumption, not a delivery promise.
The planner is designed for one main variable. Multiple changes make causal interpretation harder.
An Excel worksheet with each arm, budget, duration, metric, hypothesis, fixed conditions and decision rule.
