Though both commerce media spend and measurement capabilities are growing rapidly, the gap between investment and the ability to accurately measure performance continues to widen.

"The reality is, spending is outpacing the pace of capability," said Matt Barresi, president of digital commerce and capabilities at Kimberly-Clark, during EMARKETER’s virtual summit on commerce media.

Brands are increasingly moving beyond platform-reported metrics toward more sophisticated approaches like incrementality testing, but fragmentation across media networks makes comparisons nearly impossible.

"Really, every retail media network or platform gives you a different view of success," said Jack Kneuper, senior digital performance marketing manager at Perdue Farms. "It might be the attribution window or the KPIs that they have available, but that makes it difficult to compare channels in an apples-to-apples style."

ROAS becomes irrelevant for CPG brands

Return on ad spend (ROAS), historically the default metric in retail media, has lost relevance for consumer packaged goods (CPG) companies as they shift focus to incrementality and lifetime value.

• "For us, ROAS is sort of irrelevant," Barresi said. "I kind of view ROAS as spelled J-U-N-K for our business."
• While ROAS may remain relevant for durable goods businesses, CPG brands are prioritizing two different measures: Incrementality and lifetime value.

Incrementality has become the primary focus, though lag time in receiving results remains a significant barrier. Kimberly Clark is working to accelerate the time associated with incrementality measurements, though substantial room for improvement remains, said Barresi.

Lifetime value proves particularly important for certain categories.

• Baby care and similar categories with high loyalty rates benefit disproportionately from understanding user acquisition value, retention, and growth into adjacent categories.
• But ROAS provides only a short-term view that doesn't capture the full picture.
• "Incrementality, margins, repeat rate, lifetime value, they help us truly calculate what we're activating," Kneuper said. "Is it making a long term impact as opposed to just the short term?"

Budget thresholds determine incrementality testing

Not every campaign warrants the investment required for rigorous incrementality testing, forcing brands to make strategic decisions about when testing delivers sufficient value.

Perdue Farms evaluates incrementality testing based on two primary factors: Whether the platform represents meaningful budget allocation, and whether results would actually change investment decisions.

"If there's a threshold, will the results help us reallocate the budget, and is the upside of learning that bigger than the cost and time that it would require to run that test," Kneuper said.

Uncertainty also plays a role in testing decisions: Channels with clear roles and stable performance may not require immediate incrementality testing, while platforms showing performance that appears too good or where attribution may overstate value warrant deeper investigation.

• For smaller budget initiatives, brands can rely on directional data to maintain speed.
• Higher-budget campaigns or major strategic decisions require more rigorous analysis despite the additional time investment.

Transparency drives partner collaboration

Leading brands are sharing granular, blinded performance data with retail media partners to drive mutual problem-solving rather than pure selling relationships.

Kimberly Clark provides media partners with transparent data showing their performance relative to alternative investments and the impact of capabilities they've developed.

• “If you approach it with the right approach of true partnership, you get the truth on the table, and you get to joint problem solving," Barresi said.
• This helps align objectives between brands and partners but it requires two-way transparency, with brands willing to hear what partners want to communicate and what's important to them.

Even with stronger collaboration and transparency between brands and partners, broader industry alignment on how performance is measured remains a key missing piece.

"I would just love to see more consistency across the board with attribution windows, definitions of metrics, new to brand logic, incrementality and that framework," Kneuper said.

Barresi pointed to brand safety standards as precedent for industry-wide collaboration.

"It wasn't that long ago when brand safety standards were all over the place across the industry," he said, noting advertisers collectively stepped up to create industry requirements, and a similar collaborative approach could work for measurement standards.

Ultimately, standardization benefits everyone in the ecosystem.

"The more clear we are on how things are working, the money will follow," Barresi said. "We're all in the growth business."

AI accelerates workflow but doesn't replace strategy

AI is making measurable differences in measurement workloads today, primarily by reducing manual work and accelerating insights rather than replacing measurement strategies.

Perdue Farms has reduced reporting time from over an hour to less than five minutes using AI for tasks like summarizing trends, identifying outliers, and creating clearer stories from raw data.

"It really frees us up to be able to spend more time to actually interpret the data," Kneuper said.

Kimberly Clark is advancing rapidly on AI-powered content generation, which accelerates the ability to conduct A/B testing across various content types at scale. The company expects this to become standard operating procedure in the near term.

• Looking ahead, brands are asking partners to deliver more machine learning-based automation.
• "It's not a crazy concept. Frankly, it's here in some avenues, but it's got an ability to be here in a more meaningful breadth and depth," Barresi said.

Watch the full session

This was originally featured in the Commerce Media Weekly newsletter. For more marketing insights, statistics, and trends, subscribe here.