Developing
Opinion column with a hypothetical example and no cited data.
In 60 seconds
- Hypothetical example: platform-reported ROAS 4x, estimated true incremental return about 1.5x.
- At a 40% contribution margin, 1.5x returns $0.60 per dollar before media cost, a $0.40 loss per dollar spent.
- The author says marketing and finance can align around contribution profit instead of ROAS.
- He cautions that tests reflect one investment under specific conditions and results can shift as budgets scale.
Impact Layer
- What changed
- A practitioner makes the finance case for incrementality testing.
- Why it matters
- Reported ROAS can overstate returns, and finance teams want contribution-level evidence.
- Who is affected
- Performance marketers, analytics leads and finance partners.
- Martech stack area
- Analytics, Advertising
- Stack impact
- Pick one large channel, design a holdout or geo test and compare the incremental return with the platform's reported ROAS.
- Business impact
- Low A contributed opinion piece on MarTech (owned by Semrush), reviewed by editorial staff. The example is illustrative and no test methodology, sample or dataset is given.
- Related technologies
- What to watch
- Your own test results.
Tom Leonard, a consultant who works with brands on marketing mix modelling and incrementality testing, argues in a MarTech column published October 19, 2026 that incrementality evidence gives marketers confidence to change budgets.
The argument
- Marketing and finance can align around contribution profit instead of ROAS.
- The status quo is not risk-free: spend on inefficient tactics has costs that are less visible than the risks of cutting.
The hypothetical example
Platform-reported ROAS of 4x against an estimated true incremental return of about 1.5x. With a 40% contribution margin, each dollar returns $0.60 in contribution before media cost, a $0.40 loss per dollar spent. The article cites no data; the example is illustrative.
The author’s own caveats
“Incrementality testing doesn’t eliminate uncertainty, nor does it make decisions for you.” A test reflects one investment under specific conditions, results can shift as budgets scale or competition changes, returns are rarely linear, and testing does not specify how much to cut or where to reinvest.
Our read
The point is sound and widely made. The value comes from running your own test, not from the example.
Sources
- MarTech: The psychological advantage of incrementality testing (opens in a new tab)
Source for the column's argument and caveats.