Smartly, the ad technology company that says it manages more than $7 billion in ad spend for over 800 brands, launched a premium tier called Intelligence Pro on November 2. The add-on bundles four planning and measurement tools inside its existing platform.
What’s inside the suite
- Reach Planner lets teams compare cross-channel media scenarios before launch.
- Headroom Analysis points to areas where extra spend may still drive growth and where returns may start fading.
- Benchmarks compares campaign performance by industry, region, platform and objective using aggregated, anonymised data.
- INCRMNTAL Connect puts a modelled incrementality estimate next to platform-reported conversions.
Smartly is pitching this as an intelligence layer, not a replacement for attribution or media mix modelling. As Maxwell Tang, Chief Product Officer, put it, the challenge is “identifying what signals are useful and turning them into a decision they can act on.”
Why the placement matters
Incrementality measurement usually happens outside the buying tool, often after budgets are gone. INCRMNTAL Connect changes the default by showing a modelled view of what media actually contributed alongside what the platform itself reports. That puts a second opinion in front of the person moving budget, not just the analytics team.
The demand is real. A May 2026 Incubeta survey found 70.4% of marketing leaders were confident their budgets were deployed effectively, while 41.6% admitted part of their investment was not delivering full value. When platform numbers, internal models and incrementality disagree, Smartly’s screen could become the place where that argument happens.
The validation gap
The launch leaves several buyer questions open. Smartly has not published pricing, eligibility criteria, accuracy figures or the number of test customers. Benchmarks don’t say whose data sits behind them. Headroom outputs come as ranges — the sample panel shows a cost-per-acquisition range of $4.41 to $5.21 at 95% confidence on $1,223 daily spend — but the underlying model assumptions are not spelled out.
Modelled measurement can diverge sharply from randomised experiments. A Marketing Science study across 663 Facebook experiments found modelled upper-funnel lift came in at a median 83%, against 29% from randomised tests. That study did not test INCRMNTAL’s model, but it shows why validation matters.
What to do before moving budget
Treat Intelligence Pro as a hypothesis, not a verdict. If you can access it, compare INCRMNTAL Connect’s figures with your own lift studies over several periods. Ask Smartly how Headroom estimates were tested against holdouts, and require ranges when judging budget shifts. A default dashboard number should earn trust through convergence, not convenience.
That practical stance fits the broader shift: as AI accelerates campaign execution, the quality of each budget decision becomes the scarce advantage.
Source: PPC Land
