Smartly, the advertising technology company that says it manages more than $7 billion in ad spend for over 800 brands, on November 2, 2026 launched Intelligence Pro, a premium tier that adds budget planning, benchmarks and incrementality figures from INCRMNTAL to the software its customers use to build and run campaigns.

The announcement, datelined New York, describes the package as connecting "causal inference models with planning tools to help marketers prove business impact and make faster decisions from increasingly complex data." It is available now to what Smartly calls "eligible" customers. The company did not define eligibility, publish a price or say how many advertisers took part in the testing that preceded launch.

In Short

Smartly, a company whose software big brands use to make and run ads on Meta, TikTok, Amazon and other platforms, on November 2 started selling an add-on called Intelligence Pro that helps plan budgets and estimate whether ads actually caused sales. It matters to advertisers because ad platforms grade their own results, and Intelligence Pro shows a second estimate from INCRMNTAL, a measurement firm Smartly moved to buy this year, right next to those platform numbers. Only some Smartly customers can get it, and the company has not said what it costs or how its estimates were checked.

Four tools in one premium tier

Intelligence Pro combines four capabilities within Smartly's platform, according to the release. Each answers a different budget question, and Smartly's product page for the suite frames them that way: where to invest, where to scale, how performance compares, and what the media actually contributed.

Reach Planner lets teams compare media scenarios and forecast cross-channel reach and frequency before a campaign goes live. The release does not list the channels the forecasts cover, the data source behind the reach estimates or how audiences are deduplicated across platforms - the central technical problem in any cross-channel reach figure.

Headroom Analysis is meant to show where additional investment may still drive growth and where returns may begin to diminish. It is the only one of the four tools a customer discusses in the announcement.

Benchmarks places a campaign's performance against comparable activity by industry, region, platform and objective, "using aggregated and anonymized data," according to the release. Whose data is aggregated, how many advertisers sit in each comparison set and what minimum thresholds apply before a benchmark is shown are not stated.

INCRMNTAL Connect adds what Smartly calls "an incrementality-based view of media contribution alongside platform-reported performance." The release says it "can inform applicable Smartly workflows," without naming which ones. Smartly's product page describes the feature as always on and embedded in the platform, with no additional data setup required, and illustrates it with a chart setting INCRMNTAL-modelled conversions beside platform-reported conversions, period by period.

Maxwell Tang, Chief Product Officer at Smartly, put the rationale in terms of signal overload. "Marketing teams are inundated with data, but the challenge is identifying what signals are useful and turning them into a decision they can act on," he said. "Intelligence Pro brings forecasting, market context and state of the art incrementality signals together to help marketers assess where to invest and what their media is contributing. As AI accelerates campaign execution, the quality of each decision matters more."

Tang joined Smartly as CPO when the company named three executives to its leadership team on October 1, 2025, after more than ten years at Etsy.

What Hearst says Headroom changed

The single customer voice in the release comes from Hearst Newspapers. Sierra Smith, Sr. Marketing Manager - Paid Social at the publisher, described the previous process as largely manual.

"We used to spend a lot of time pulling data from different platforms and manually analyzing where to invest," Smith said. "With Headroom, we can evaluate budget opportunities using historical performance and make more informed decisions without all the manual work. It's been a game changer for how we approach planning."

Two details in that quote matter more than the endorsement. First, Headroom works from historical performance, which places it in the family of response-curve methods that estimate the shape of returns from past spend and outcomes. Second, the benefit Smith describes is time saved on analysis, not a measured change in sales or efficiency. The release attaches no figures to Hearst's use of the tool.

Smartly's product page gives a sense of the output format. A sample Headroom panel shows an average daily spend of $1,223 and a cost per acquisition of $4.89, then lists CPA ranges at three confidence levels: $4.41 to $5.21 at 95%, $4.48 to $5.18 at 90% and $4.79 to $5.04 at 50%. The conversions column reads 235 to 277 at all three levels, which would not happen with genuine intervals of different widths, so the figures appear to be placeholders. Even so, the panel indicates that Headroom presents its estimates as ranges rather than single numbers. Dividing the sample spend by the sample CPA gives roughly 250 conversions a day, a PPC Land calculation that sits inside the displayed range.

The idea of measurable headroom is not new. In December 2025, TikTok data drawn from NIQ modelling put two Romanian beauty brands at roughly 50% of optimal TikTok spending levels, with response curves projecting that a 20% budget increase would not reach saturation. Those were model projections, and the same caution applies to any headroom estimate: it describes where returns are expected to flatten, not where they will.

Smartly already offers a budget tool. In a pilot run through agency Wpromote, one client recorded a 29% increase in lower-funnel return on ad spend while cutting media spend by 26%, results attributed to Smartly's Predictive Budget Allocation. The Intelligence Pro release does not explain how Headroom Analysis relates to Predictive Budget Allocation, whether one feeds the other, or whether Headroom can move budgets automatically rather than inform a person who then moves them.

INCRMNTAL Connect and the deal behind it

The fourth capability carries the name of a company Smartly set out to buy earlier this year. Smartly signed a letter of intent on March 16, 2026 to acquire INCRMNTAL, a Tel Aviv incrementality platform founded by Maor Sadra and Moti Tal with between 11 and 50 employees. Financial terms were not disclosed. INCRMNTAL's approach, as described at the time, uses causal inference to estimate the incremental value of campaigns without user-level data or tracking, analysing natural fluctuations in campaign activity instead of withholding ads from a control group or pausing campaigns.

When Smartly appointed Kelly Uphoff as chief technology officer on October 8, 2026, the company's press page listed the INCRMNTAL purchase as completed on May 7, 2026, and Smartly counted INCRMNTAL among the acquisitions integrated under her predecessor, Lauri Tarkkala, alongside Ad-Lib.io and Viralspace.ai. The November 2 release does not mention the acquisition at all. INCRMNTAL keeps its own name inside Smartly's product.

That distinction is worth noting because INCRMNTAL also operates beyond Smartly's platform. When OpenAI expanded the measurement partners for ChatGPT Ads in October 2026, INCRMNTAL appeared in a group described as full-funnel and advanced measurement firms, next to Fospha and Measured.

The method matters for how INCRMNTAL Connect's numbers should be read. Incrementality is the share of outcomes that would not have happened without the advertising. The most direct way to measure it is a holdout study, in which a randomly selected group is kept from seeing ads and serves as the baseline. INCRMNTAL's model infers that baseline from variation in spend and results over time. That avoids switching off campaigns, which is why "always-on" measurement appeals to advertisers, but it also means the estimate depends on modelling assumptions rather than a randomised comparison.

How large can the gap between modelled and randomised estimates be? Research discussed on PPC Land in September gives one answer. Across 663 experiments run on Facebook, machine-learning estimates of upper-funnel lift came in at a median 83%, against 29% from the randomised tests, according to a 2023 Marketing Science paper by Gordon, Moakler and Zettelmeyer. That study tested specific observational methods, not INCRMNTAL's, and its results cannot be transferred to Smartly's product. It does show why validation against experiments matters - and the release offers none for INCRMNTAL Connect.

Smartly's product page carries one incrementality result: a case study in which Pandora increased incremental ROAS by 22% using Brand Pulse, another Smartly tool. No baseline, time period, channel or method is given on the page.

Platform numbers and the measurement argument

INCRMNTAL Connect is pitched as a second lens on platform-reported performance, and the gap it addresses is well documented, sometimes by the platforms themselves.

Meta's own Suite of Truth white paper, drawing on 307 studies run by 54 advertisers between March 2022 and November 2024, found that advertisers at the median undervalue Meta by 31% when using rules-based attribution compared with incrementality measurement. The same coverage cited Kantar research from May 2025 showing the average advertiser runs 3.8 measurement solutions and that 55% have seen contradicting results across them. The error can also run the other way. In 2025, practitioners argued that under Meta's default 7-day click and 1-day view attribution, a marketer reporting a ROAS of 10 may be doing worse than one reporting 2, because view-through credit can flow to purchases existing customers would have made anyway.

Modelled measurement has its own failure modes. An August 2026 preprint linked to a Zalando researcher found that, on simulated data, a standard marketing mix model reported a 10.61x ROAS for paid search against a true value of 4.20x. Only when geo-experiments were built into the estimation did the result land near the true figure.

Platforms, meanwhile, have made their own experiments cheaper. Google lowered the minimum budget for incrementality tests in Google Ads to $5,000 at Google Marketing Live in May 2025, using Bayesian methods that rely on prior assumptions.

Smartly does not position Intelligence Pro as a replacement for any of this. Its product page says the suite complements existing measurement, naming attribution, marketing mix modelling, lift studies and first-party business data as approaches it works alongside. That leaves a familiar question for the 3.8-tool advertiser Kantar described: when INCRMNTAL's view, the platform's view and an in-house model disagree, which one moves the budget?

Benchmarks and planners are crowded ground

Neither benchmarking nor reach planning is new territory for ad buyers. Amazon Ads made its benchmarks generally available in 18 markets in May 2026, covering Sponsored Ads and Amazon DSP. Pinterest added a Media Planner to Ads Manager in January 2026 that estimates impressions, reach, frequency and CPM for awareness and consideration campaigns, without publishing accuracy figures.

What separates Smartly's versions, at least in principle, is position. A platform's planner sees one platform. A tool sitting across Meta, TikTok, Amazon, Pinterest, Reddit, Snap, Spotify and Google, the media partners Smartly names in its boilerplate, can in theory compare them on common terms. The release does not say which of those platforms feed Reach Planner or Benchmarks at launch.

From creative automation to an intelligence layer

Smartly built its business on creative automation, turning product feeds and brand assets into large volumes of ad variants. Intelligence Pro extends a run of 2026 announcements that pushed the company towards planning, measurement and agent-driven buying.

In March, Smartly connected its platform to Amazon DSP for streaming TV campaigns and opened a product, engineering and commercial hub in Mexico City. In June it plugged into the Roku Ads API, became the first SaaS activation partner for Horizon Media's Blu audience API and introduced Smartly Synapse, an orchestration and memory layer behind planning, execution and optimisation agents. In September it added LinkedIn Ads.

The product page for Intelligence Pro draws the line from those pieces to this one. It describes intelligence as the step that tells marketers what should happen next, and says those signals can increasingly inform AI-orchestrated decisions across Smartly's Creative and Media workflows over time. The launch release is more restrained, describing Intelligence Pro as part of "Smartly's broader vision to bring creative, media and intelligence together." Nothing in it says Intelligence Pro outputs already drive automated budget or bid changes.

Smartly is not alone in reframing creative software around measurement. Bannerflow, another creative management vendor, launched a Creative Intelligence product in October 2026 that scores ads for attention before launch, also without publishing validation data for its models at launch.

What the release leaves out

Several points a buyer would need are absent from the announcement:

  • Price and eligibility. Intelligence Pro is described as a premium offering for eligible customers. Neither the price nor the criteria are stated, and the release does not describe what Smartly customers get without the Pro tier.
  • Validation. No accuracy figures are published for Reach Planner forecasts, Headroom estimates or INCRMNTAL Connect, and no comparison against randomised experiments is given.
  • Data sources. The release says Benchmarks uses aggregated and anonymised data but not whose, nor how many advertisers stand behind each benchmark.
  • Testing scale. Smartly said it tested the offering with some customers before launch. Hearst Newspapers is the only one named.

Smartly's own figures also vary across its materials. The release boilerplate says the company supports more than 800 brands, while the Intelligence Pro product page says more than 700 brands use Smartly - the count that appeared in Smartly's September 2025 Spotify integration announcement and its October 2025 leadership appointments, before it moved to 800. The boilerplate repeats PwC-validated results of a 5.5x return on ad spend and 42 minutes saved every hour, figures unchanged across Smartly releases since at least September 2025, and cites leadership of The Forrester Wave: Creative Advertising Technologies without naming an edition. PPC Land's June coverage of the Roku integration tied the claim to Forrester's Q4 2024 Wave.

Why this matters for the marketing community

For advertisers, the practical significance lies in where the incrementality number appears. Lift studies and mix models usually run outside the buying tool and report after the money is spent. Smartly is placing a modelled incrementality figure on the same screen where campaigns are built and budgets are set, beside the conversion counts each platform reports about itself. That changes who sees the second opinion and when.

The demand is there. In an Incubeta survey of CMOs and CEOs in the UK and US published in May 2026, 70.4% of marketing leaders said they were confident their budgets were deployed effectively, while 41.6% acknowledged that part of their investment was not delivering its full value, and only 34.4% used a unified measurement approach. The survey's sample size was not disclosed.

The risk is the mirror image of the opportunity. A model embedded in a buying tool, owned by the company selling that tool, can shape budget decisions with the authority of a dashboard default. Without published validation, advertisers using Intelligence Pro will be weighing one modelled estimate against another, with the platform's own figures alongside. For agencies and brands that already run lift studies, the useful test of INCRMNTAL Connect will be whether its figures converge with theirs over time. For those that do not, it may become the first incrementality number they see at all.

Timeline

Summary

Who: Smartly, the advertising technology company that says it supports more than 800 brands and manages over $7 billion in ad spend, with Chief Product Officer Maxwell Tang quoted. INCRMNTAL, the Tel Aviv incrementality firm Smartly agreed to acquire in March 2026, supplies the incrementality component. Hearst Newspapers, through Sr. Marketing Manager - Paid Social Sierra Smith, is the one customer named.

What: Intelligence Pro, a premium offering inside Smartly's platform combining Reach Planner for cross-channel reach and frequency forecasts, Headroom Analysis for estimating where extra investment may grow results or hit diminishing returns, Benchmarks built on aggregated and anonymised data, and INCRMNTAL Connect, which shows incrementality-based contribution beside platform-reported performance.

When: Announced November 2, 2026, and available now to eligible Smartly customers.

Where: Announced from New York, inside Smartly's platform, which connects to media partners including Amazon, Google, Meta, Pinterest, Reddit, Snap, Spotify and TikTok. The release names no regional limits.

Why: Smartly says marketing teams have more data than they can turn into decisions and that, as AI speeds up campaign execution, the quality of each decision matters more. The launch puts a modelled incrementality estimate next to platform self-reporting, in the tool where budgets are set. Pricing, eligibility criteria and validation data for the models have not been published.