InMobi Advertising launched FRAME Attention on July 24, 2026, a certified media product spanning mobile in-app video and connected television that promises a minimum floor of 8,000 Attention Per Mille and carries validation from Lumen Research. The launch pushes attention measurement from a reporting layer into a supply-side guarantee, at a moment when rival systems are embedding the same class of signal into bidding algorithms.

InMobi Advertising, the advertising arm of the Singapore-headquartered technology group, made the announcement from New York through Business Wire. The product is sold via Deals and insertion order to brand advertisers and agencies worldwide, and it is bundled into two packages: one for mobile in-app video, one for CTV.

What separates this from the wave of attention products that preceded it is the direction of the promise. Most attention tooling scores inventory after the fact, or filters it before a bid. FRAME Attention sets a contractual floor on the supply itself.

What Attention Per Mille actually counts

The metric at the centre of the launch is APM, or Attention Per Mille. According to InMobi, APM measures the percentage of impressions actually viewed multiplied by average view time, expressed per 1,000 impressions, producing a count of attentive seconds delivered.

That construction matters. A viewable impression under the Media Rating Council standard requires 50% of pixels in view for at least one continuous second for display, and two continuous seconds for video. The threshold is binary and says nothing about whether a human looked. Attention metrics attempt to close that gap, and the industry has spent roughly three years arguing about how.

APM is a volume measure rather than a rate. Multiplying viewed share by dwell time and normalising to a thousand impressions yields seconds, not percentages, which makes it additive across a campaign in a way that a viewability rate is not. An 8,000 APM floor translates to eight attentive seconds per impression across a thousand impressions, assuming full delivery.

"Brands have spent years optimizing for impressions that were never really seen," said Kunal Nagpal, Chief Business Officer at InMobi, in the announcement. He described the model as the first in-app video attention model built at scale, arguing that buyers should not have to choose between reach and measured attention.

The numbers, and what the release does not disclose

Independent testing by Lumen Research found InMobi's mobile in-app video inventory performed 3.5 times better than standard YouTube equivalent video ads, according to InMobi. On connected television, the company reported 9,593 APM against a streaming benchmark of 7,286 APM, a gap of 31%.

A brand outcomes study measured by Cint across four million impressions recorded a 67.4% lift in ad recall and a 16.5% lift in brand awareness, both described as above Cint's established benchmarks for CTV brand lift studies.

Several figures are absent. The release does not publish an absolute APM value for the mobile in-app inventory, nor the YouTube baseline against which the 3.5x multiple was calculated. It does not define what constitutes a "standard YouTube equivalent" video ad, a definitional choice that carries most of the weight in a comparative claim. Nor does it disclose the composition of the four million impressions in the Cint study, the campaigns involved, or the control methodology. The 7,286 APM streaming benchmark is presented without a stated source or panel size.

Lumen conducted the validation, and Lumen's chief executive supplied a supporting quote. "These are unusually high levels of attention - and due to a combination of simultaneous factors," said Mike Follett, CEO of Lumen Research, who attributed the result to a mix of interactive environments, load speed, creative quality and persistent onscreen presence.

The testing was commissioned by the party whose inventory it evaluates. That does not invalidate the methodology, but it places the results in the same category as most vendor-published attention research: independently measured, commercially motivated.

A supply-side bet against buy-side optimisation

InMobi framed its position as structural rather than algorithmic. Operating at the supply layer, the company said, it can directly curate high-attention environments, promise a minimum APM floor, and apply creative authoring and SDK-level controls for consistent user experiences, with third-party verification on top. The release states this level of certainty cannot be replicated through buy-side optimisation tools alone.

That claim lands in the middle of an argument the market is currently having with itself.

The dominant trend over the past six months has been the migration of attention scores into bid-time systems on the demand side. Adelaide's AU metric became a pre-bid targeting signal inside Amazon DSP on June 10, 2026, complete with a Quality Floor that excludes inventory scoring in the bottom 10% and defaults to filtering made-for-advertising sites. Six weeks later, the same metric arrived as a signal inside DV360 Custom Bidding, giving Google buyers impression-level attention valuation through the mechanism that survived the removal of bid multipliers in May 2025. Index Exchange embedded xpln.ai attention signals into its supply-side platform in February 2026.

InMobi is arguing the opposite direction: that curation at origin beats filtering at auction. Both propositions cannot be equally true, and neither has been settled by evidence available to buyers.

There is precedent for the supply-side approach. Teads extended its Lumen partnership to CTV HomeScreen inventory globally on May 6, 2026, securing exclusive access to Lumen's CTV attention measurement for that placement across the United States, EMEA, APAC and LATAM. Lumen had earlier brought attention measurement to Netflix advertising across five European markets in March 2026. The pattern across those deployments is consistent: a publisher or platform pairs its inventory with a measurement vendor and sells the combination as a quality signal.

The outcome question the guidelines already flagged

InMobi's announcement states that measured attention translates directly into brand outcomes clients are measuring against. That framing runs against explicit guidance issued by the industry's own standards bodies.

The MRC and IAB finalised attention measurement guidelines in November 2025, establishing minimum requirements for quality, transparency and comparability across vendors. Those guidelines covered data signals, visual tracking, physiological and neurological observation, and panel or survey-based methodologies. They also stated that attention should not be treated as a measure of outcomes for evaluating campaign performance, positioning it instead as an exposure and engagement signal.

The Cint brand lift data is InMobi's answer to that objection. Rather than asserting that high APM equals effectiveness, the company pairs the attention figures with separately measured recall and awareness lifts. Other vendors have taken similar routes. IAS and Mastercard built a system validating attention scores against actual purchase data in March 2026, precisely to move past attention as a proxy. Stellantis ran IAS Quality Attention across 13 billion impressions in 19 countries, reporting a 33% ad recall improvement and a 165% engagement increase, with results published on June 4, 2026.

Comparability remains the unresolved problem. APM, AU, Quality Attention and xpln.ai's predictive KPI are computed differently, and a score in one system has no defined relationship to a score in another. An 8,000 APM floor is meaningful only against InMobi's own benchmark set. A buyer running InMobi bundles alongside an Adelaide-optimised DV360 campaign is not comparing like with like, whatever the reporting dashboards suggest.

Two bundles, one measurement layer

The High Attention Mobile Bundle delivers certified placements across premium apps, targeting audiences in active sessions, with Lumen attention reporting applied to every campaign. The High Attention CTV Bundle curates premium streaming inventory for engaged viewing, with outcomes reported independently by Lumen Research or Adelaide.

That second detail is worth noting: the CTV bundle accepts two measurement vendors rather than one. Adelaide's AU is the metric now live inside both Amazon DSP and DV360, which means a buyer can in principle reconcile InMobi CTV delivery against the same currency used in programmatic activation elsewhere.

FRAME Attention sits inside InMobi's broader FRAME Ad Experience suite, which the company positions as a move from reach-and-frequency planning toward outcome-driven investment. Availability is global across InMobi's mobile, desktop and CTV supply, including Glance properties. Glance is owned and operated by Glance InMobi Pte. Ltd., a non-consolidated subsidiary of InMobi Pte. Ltd.

Why this matters for media buyers

Attention has moved from conference-panel topic to line item faster than most measurement categories. The reason is the accumulating evidence that impression counts and viewability rates carry limited information about whether advertising worked.

Connected television has been the sharpest case. Research covered by PPC Land found that more than one third of CTV ad impressions are delivered in TV-off environments, an estimated one billion dollars in annual wasted spend. VAB and TVision analysis published in February 2026 documented that a YouTube CTV impression and a premium streaming impression occur under materially different viewing conditions, making them non-equivalent units despite identical pricing structures.

Mobile in-app carries its own version of the problem. Small formats, short sessions and rapid scroll behaviour compress the window in which an ad can register. Mobkoi and xpln.ai reported average attention above 5.5 seconds across more than one billion measured mobile impressions in June 2026, a figure they positioned as substantially above the display market benchmark.

Against that backdrop, a guaranteed floor is a commercially distinct proposition. Filtering removes bad inventory. A floor commits the seller to a performance level. Whether the commitment carries a remedy when delivery falls short, the release does not say, and the mechanics of enforcement are the detail buyers will interrogate first.

InMobi's own position in the market has shifted considerably. The company cut 5% of its workforce in February 2024amid ad tech uncertainty, raised 100 million dollars in September 2024 for AI-driven expansion, and partnered with Scope3 in October 2024 on green media segments for mobile and in-app advertising. It joined Kochava's Certified Partners Program in March 2026 alongside LG Ad Solutions, Digital Turbine and three others.

The competitive stakes are visible in the comparison InMobi chose to lead with. Naming YouTube in a headline claim is a direct challenge to the largest video advertising property in the market, and it invites a methodological argument InMobi has not yet published enough detail to win outright.

Timeline

Summary

Who: InMobi Advertising, part of the Singapore-headquartered InMobi group, with validation supplied by Lumen Research, the London-based eye-tracking company, and brand outcome measurement from Cint. Quotes were provided by Kunal Nagpal, Chief Business Officer at InMobi, and Mike Follett, CEO of Lumen Research.

What: FRAME Attention, a certified high-attention media product promising a minimum floor of 8,000 Attention Per Mille, sold in two bundles covering mobile in-app video and connected television. Lumen testing reported mobile in-app performance 3.5 times better than standard YouTube equivalent video ads and CTV inventory at 9,593 APM against a 7,286 APM streaming benchmark, a 31% gap. A Cint study across four million impressions recorded a 67.4% ad recall lift and a 16.5% brand awareness lift.

When: Announced on July 24, 2026.

Where: Announced from New York and available globally across InMobi's mobile, desktop and CTV supply, including Glance properties, purchasable through Deals and insertion order.

Why: Viewability confirms only that an ad met a pixel and duration threshold, not that a person looked. Attention metrics attempt to quantify the difference, and InMobi is positioning curation at the supply layer as an alternative to attention filtering applied by demand-side platforms. The claim sits alongside November 2025 MRC and IAB guidelines cautioning that attention should not be read as a direct measure of campaign outcomes.