A Lift Measured Only on People Who Saw the Ad Is Not Campaign Lift
A lift for ‘exposed users’ compares people who chose to engage with people who did not. Learn which population a lift was computed on, and what to ask.
A lift for ‘exposed users’ compares people who chose to engage with people who did not. Learn which population a lift was computed on, and what to ask.
Attributed conversions can include people who only saw an ad and never clicked. Here is what view-through counts, when it is fair, and what to ask for.
Each ad platform credits itself for a sale it touched, so summed dashboards can exceed real orders. Here is how to reconcile a claimed total with the actual count.
A return on ad spend measured only inside a campaign’s live dates can miss the purchases that follow. Learn to judge what a window cuts off at the back.
A cause marketing total raised can bundle pledges, unpaid corporate matches, and in-kind value in with cash actually collected. Here is how to separate them.
A content count belongs in the methodology, not the results box. Here is how to tell an effort metric from an outcome before you credit either one.
A pilot city that beat its target still needs to resemble the rest of the base before the result travels. Here’s what to check before you believe it.
An agency’s average client result can rise year over year because the client roster changed, not because the work improved. Here’s how to tell them apart.
A platform’s own brand lift survey sold, ran, and reported the test on its own ad. Here’s how to weigh that number before trusting it in a case study.
A ‘no significant difference’ result can mean the tactic failed, or that the test was too small to tell. Here is how to check which one a case study reports.