Proving that a campaign actually worked — not just that people who saw it are different from people who didn't — is one of the hardest problems in marketing measurement. The difference between correlation and causation is where most reporting falls short.
Lift reports on MX8 are designed to close that gap.
Exposed vs. Control, Done Right
Every lift report compares two groups: respondents who were exposed to a stimulus — an ad, a message, a creative — and a matched control group who were not. The platform builds that control group automatically, matching on demographics so the comparison is fair.
This isn't a single comparison. MX8 rebuilds the matched control group across multiple iterations — typically five to ten — and averages the results. This reduces the noise that comes from any single random split and produces more stable, defensible numbers.
What You Can Measure
Lift reports work with both categorical outcomes (brand awareness, message recall, purchase intent) and numeric scales (consideration scores, favorability ratings). For numeric outcomes, the platform calculates lift at thresholds — for example, the percentage scoring 7 or above — so you can see where exposure made a difference and where it didn't.
You can also configure the report to surface negative lift: cases where exposure actually hurt a metric. That's uncomfortable but useful, especially when comparing creatives.
Statistical Rigor Without the Complexity
The P value threshold is configurable. Set it tight at 0.05 for client-facing results where you need strong evidence. Open it up to 0.50 for early exploration when you're looking for directional signals. Only results that pass your chosen threshold are shown, so the report self-filters to what's meaningful.
A Practical Example
You're measuring whether a spring campaign lifted brand awareness and consideration. You tag exposed respondents using an indicator in the survey, select gender as a control variable, and set brand as the stimulus topic. The platform does the rest: splitting respondents, matching control groups, calculating lift per brand, and filtering by significance.
The result might show that Brand A awareness rose 8 points among the exposed group versus control, and consideration (at the 7+ threshold) rose 5 points. That's a defensible, matched-sample result — not just a difference in who happened to respond.
When to Use Lift
Lift reports are most valuable when you need to prove effectiveness, compare creatives or messages, or defend a media investment with evidence. They're available now for any MX8 survey with exposure tracking configured.
For the full configuration guide — including how to set up indicator tags, choose matching iterations, and interpret results — visit our documentation on reporting lift against control groups.