Our Take
The 8% stat exposes a fundamental measurement crisis: brands demand proof of incremental impact but won't invest in proper testing infrastructure.
Why it matters
Marketing budgets face increasing scrutiny while attribution becomes more fragmented across channels. Teams need reliable incrementality data to defend spending and optimize allocation.
Do this week
Marketing directors: Audit your current attribution stack this week to identify if all partners report only positive results, which signals measurement failure.
Most brands lack formal incrementality programs
Only 8% of brands operate formal incrementality programs (per Digitas research), despite widespread pressure to prove marketing impact. An ADWEEK panel with executives from Hershey, Nestlé, CAVA, Haleon, and measurement platform Fetch revealed the gap between demand for clean metrics and actual measurement capabilities.
The core challenge: distinguishing between campaigns that change behavior versus those that capture existing demand. Sandra Abla from CAVA described the organizational tension: "They want a clean metric to show incrementality. But, we know human behavior is super messy. And, unless you do have control, it's really hard to understand what is driving that incrementality."
Attribution breakdown occurs across teams claiming the same conversions. Hershey's Vinny Rinaldi noted: "At the end of the day, there's only one unit tied to moving through the register or not. It's not one team that did it. It's everyone did it."
Fragmented consumer paths complicate measurement
Consumer behavior has become harder to track as purchase journeys compress and fragment. Haleon's Marissa Solan explained: "There are just so many pathways, how are you able to track all those? You can go from awareness to purchase in the same moment today."
Most brands operate across disconnected systems with misaligned timelines. C-suites want immediate answers while meaningful incrementality measurement requires six to eight months of observation (per CAVA's analysis). This mismatch forces teams toward faster but less reliable attribution methods.
A key diagnostic signal: if measurement partners report only positive campaign results, the system isn't working. CAVA's Abla warned: "If all your results are positive, then you know something's not adding up."
Randomized control trials remain the standard
Fetch's Geoff Matthews identified randomized control trials as the most reliable measurement method, but emphasized execution discipline: "People throw out control, but it's got to be randomized. People will model it, or they'll stack it after the fact, and that's really challenging."
The measurement requires honest partner feedback. Matthews stressed that platforms must flag likely failures upfront: "If we know our platform, or we know the plan, and we don't think it's going to work, we've got to tell you now."
Fetch's Robin Wheeler highlighted the broader issue behind unclear measurement: "I'm sick of 'we grew sales.' What does that even mean?" The focus should remain on concrete unit movement rather than attribution theater that obscures actual performance.