Nano Interactive launched Nano Screen Graph - CTV Intelligence on August 11, 2026, an artificial intelligence layer that cross-matches incomplete content records held by different streaming publishers into single merged listings. The announcement carries no coverage figures, no named publisher or platform partners, no pricing and no performance data.

The product addresses a mechanical problem rather than a strategic one. When a connected television impression becomes available, the metadata describing what is playing behind that impression arrives in fragments. One publisher passes a Content ID and a programme title but omits the episode number. Another passes the episode number and omits the title. Neither record, on its own, tells a buyer enough to distinguish one episode of a series from another, or to establish whether the surrounding content is suitable for a given brand.

According to Nano Interactive, Screen Graph merges those partial records into one complete, consistent listing. The company describes the merged output as filling missing data including recency, cast, show runtime information, popularity, and content sensitivity flags for material involving addiction, abuse, or violence.

What the merge actually produces

The stated outputs fall into two categories, and they serve different buyers.

The first is planning granularity. According to the company, planners can analyse CTV content topic relevancy by geographic region and build custom segments from the merged records. All segments, including what Nano Interactive calls Intent personas and topic rankings, adjust in real time based on the target audience specified.

The second is brand suitability. Content sensitivity flags covering addiction, abuse and violence are the classification layer that suitability decisions run on. In display and video environments, that layer has existed for years through page-level semantic analysis. In connected television, where the unit of content is an episode rather than a page, the equivalent classification depends entirely on whether anyone knows which episode is playing.

Nano Interactive describes Screen Graph as a continuously evolving data layer. The company states the product is available now.

Niall Moody, chief revenue officer at Nano Interactive, framed the problem in terms of a forced trade-off.

"CTV has been booming for years, but the data underneath it has never caught up. Buyers have been forced to choose between broad, low-precision targeting and identifiers that simply don't hold up on the big screen," said Moody. "Advertisers know they need better CTV audience data, but have largely accepted wasted spend as the cost of doing business in the channel, a reality that shows up every time a viewer sees the same ad repeated over and over. Nano's Screen Graph - CTV Intelligence closes that gap. We built this technology to give planners the same intent-driven precision they already trust in display and video, without asking the channel to rely on IDs it was never built to support."

That last clause is the positioning. Screen Graph is offered as an alternative to identity resolution, not a complement to it.

The dependency the announcement does not address

A cross-matching system enriches records. It does not create them.

Screen Graph works by combining what publisher A passed with what publisher B passed. Where neither passed anything usable, there is nothing to merge. The scale of that constraint has been measured. Gracenote research published on May 14, 2026 cited a March 2026 analysis by contextual data company Peer39 finding that only 40% of connected television bid requests contain usable programme-level content signals. Within that 40%, 33% carried some genre data and 7% carried shallow or inaccurate genre classifications.

The same research documented how uneven the passing is at the individual publisher level. A sample data set from Rain The Growth Agency showed three supply-side platforms serving inventory for one publisher passing series information in 72.3%, 70.3% and 53.1% of bid requests, with title data ranging from 65.4% down to 6.1%.

Those figures describe the raw material Screen Graph operates on. The announcement does not state how many publishers contribute records, how many titles the merged layer covers, what share of connected television impressions it can describe, or how the merged records reach a buying platform. Nor does it name a demand-side platform, a supply-side platform or a launch client.

The absence matters because the value of a content data layer is a function of its coverage. A layer that resolves episode-level detail on a small fraction of available impressions changes what a planner can see on that fraction, and nothing elsewhere.

Content-level versus impression-level

The distinction the product rests on is worth stating precisely, because it separates Screen Graph from most of the tooling launched in this category over the past two years.

Impression-level solutions attach information to a specific ad opportunity: was it viewable, was it fraudulent, was a human present, which household received it. Content-level solutions attach information to the programme itself: which series, which episode, which cast, which runtime, which sensitivity classification.

Both have been retrofitted onto connected television after the fact. IAS launched Total TV on April 27, 2026, combining show, genre, rating and language measurement with device, viewability and invalid traffic verification across Disney, NBCUniversal, Paramount and Prime Video inventory. FreeWheel debuted a series-level reporting tool for seven publishers in July 2026, drawing on impression-level delivery data its own ad server already recorded. Gracenote introduced contextual CTV ad targeting in July 2024 with a shared taxonomy of programme genres, ratings, content types and advisories, and Content Connect followed on December 4, 2025.

What separates Screen Graph, on the company's account, is the merge step. The other products surface a signal where a publisher chose to pass one. Screen Graph attempts to reconstruct a fuller record by reconciling several partial ones held by different sellers. Whether that reconciliation resolves titles a single publisher never disclosed, or only completes records already partially present, is not specified in the announcement.

The identifier route runs in parallel

Connected television has produced two competing answers to the same question, and both are being built simultaneously.

The identifier answer resolves the impression to a persistent code. Viant's IRIS_ID, inherited from the November 2024 acquisition of IRIS.TV, tags inventory by network, app, show, scene and pod, and its presence in the CTV bidstream more than tripled within a year of that acquisition. Adobe Advertising began ingesting Eyeota audience data into its proprietary CTV identity system on July 30, 2026, resolving impressions to consent-based identifiers and layering demographic, interest and purchase attributes onto them.

The contextual answer avoids the identifier entirely. IPG Mediabrands built Acxiom Contextual CTV on a taxonomy of more than 1,200 content categories. Ogury extended persona-based targeting to connected television on February 24, 2026, combining zero-party survey data with contextual signals rather than chasing individual identifiers. Viant integrated with Wurl in August 2025 to deliver scene-level contextual alignment.

Nano Interactive sits in the second camp by design. The company describes itself as a privacy-first signal intelligence business operating across more than 100 markets, powering intent-driven programmatic campaigns across display, video and CTV through a platform it calls LIIFT, which it states enables supply-side platforms, agencies and brands to plan, activate and optimise campaigns.

The company already has infrastructure exposure on the sell side. Index Exchange listed Nano Interactive among the partners building on Index Cloud when the containerised platform surfaced in April 2026, alongside Bedrock Platform, Chalice AI and inPowered AI, with contextual and content classification among the supported use cases. That platform remained in closed beta with select partners at the time.

The frequency claim

Moody's statement anchors the wasted spend argument to a specific viewer experience: the same advertisement appearing repeatedly. That is a frequency management failure, and it has a structural cause distinct from the metadata problem Screen Graph addresses.

Frequency capping across connected television requires either a persistent household identifier or coordination between sellers. Content metadata does not, by itself, cap frequency. A richer content record allows a planner to diversify placement across titles and reduce the chance of repetition within a narrow content pool, but it does not deduplicate exposure across supply paths. The announcement makes no frequency capping claim beyond the illustrative reference in the executive quotation.

The broader waste figure has been quantified elsewhere. The Gracenote research cited ANA benchmark data putting global ad waste attributable to programmatic inefficiencies at $26.8 billion, with connected television representing 44% of total spend, and separate estimates from Truthset attributing more than $7 billion in waste directly to poor data signals.

Buyer demand is documented; supply is the constraint

Appetite for content-level signals in connected television is not in question.

Among programmatic traders surveyed by Gracenote, 100% described show-level transparency as very or extremely important for brand safety and inventory quality, 95% said the absence of those signals stopped them advocating for larger CTV budgets during planning, and 80% said they would move budget from audience-targeted to contextually targeted CTV if actionable content signals were available. Among media planners, 86% said consistent show-level targeting and reporting would prompt them to shift linear television budget into the channel.

That demand sits against a market where content signals are the exception. According to the Gracenote report, programmatic purchasing accounts for 85% of CTV inventory, up from 75% the previous year.

Fragmentation has been the recurring theme across the category. IAB Europe's April 2026 programmatic CTV guidedocumented 18 CTV-validated consent management platforms and 466 TCF-registered vendors supporting CTV environments, and noted that HbbTV environments across European television sets often limit access to persistent identifiers, making contextual and household-level targeting more effective in practice. At IAB Europe's 2026 programmatic day, fragmentation was the word the panels returned to most often.

The commercial consequences are visible in pricing. US programmatic CPMs rose 34% year over year while CTV stalled, a pattern PPC Land attributed to new entrants expanding streaming supply faster than advertiser demand can absorb it, compounded by audience fragmentation that makes reach guarantees harder to fulfil at premium pricing.

A date the record does not settle

The source material carries an inconsistency on when the launch occurred.

The press release distributed to press stated an embargo lifting on Tuesday, August 11, 2026 at 9:00 am Eastern Time. The release text itself uses the word today. A follow-up message dated Thursday, August 13, 2026 stated that the launch was live. Trade coverage published subsequently has carried differing launch dates for the same announcement. Nano Interactive has not published a correction or clarification addressing the variance.

Why this matters for the marketing community

For planners buying connected television programmatically, the practical question is narrow: does a merged content record change what can be targeted and excluded, and on how much inventory.

The targeting case is clear enough on its face. Episode-level resolution allows exclusion of specific programming rather than whole genres, which is the difference between avoiding a category and avoiding a title. Sensitivity flags for addiction, abuse and violence give suitability teams a classification to act on in an environment where, historically, the available signal has often been an app name and a supply path.

The scale case is unresolved. Every product in this category over the past eighteen months has faced the same limitation: the seller decides what to pass, and no standardised requirement compels programme metadata to accompany an impression. Cross-matching improves the yield from whatever is passed. It does not compel anyone to pass more. That is why supply-side commitments and vendor product launches carry different weight, and why the recurrence of near-identical products indicates how much of the problem remains unsolved.

There is a second consideration for buyers evaluating the claim. The intent personas and topic rankings Screen Graph produces are derived segments, not observed ones. They rest on a merged record whose completeness varies by publisher and by supply path. A planner selecting a segment built on that layer is selecting a modelled output, and the announcement provides no accuracy measurement, no independent validation and no beta results against which to calibrate expectations.

Contextual approaches to connected television have accumulated real infrastructure since 2024. Whether Screen Graph adds coverage the existing taxonomies do not already provide is a question the announcement leaves open, because the coverage figure that would answer it was not disclosed.

Timeline

Summary

Who: Nano Interactive, a signal intelligence company that describes itself as privacy-first and operating across more than 100 markets, with chief revenue officer Niall Moody providing the statement accompanying the launch.

What: Nano Screen Graph - CTV Intelligence, an artificial intelligence capability that cross-matches partial content records held by different streaming publishers into single merged listings. The merged records fill gaps including episode numbers, recency, cast, show runtime, popularity, and content sensitivity flags for material involving addiction, abuse or violence. Planners can analyse content topic relevancy by geographic region and build custom segments, with all segments including Intent personas and topic rankings adjusting in real time based on the target audience. No coverage figures, publisher counts, platform partners, pricing or performance data were disclosed.

When: The release carried an embargo lifting on Tuesday, August 11, 2026 at 9:00 am Eastern Time, and the company confirmed the launch was live in a message dated August 13, 2026. Trade coverage published afterwards has carried differing launch dates for the same announcement. The product is stated as available now.

Where: Across connected television inventory, with the company operating in more than 100 markets through its LIIFT platform, which it states serves supply-side platforms, agencies and brands across display, video and CTV.

Why: Programme-level content signals reach connected television bid requests inconsistently, with Gracenote research citing Peer39 analysis showing only 40% of CTV bid requests carry usable signals. Buyers have said the gap constrains budgets: 100% of programmatic traders surveyed by Gracenote called show-level transparency very or extremely important, and 86% of media planners said consistent show-level targeting and reporting would prompt them to move linear television budget into the channel. A cross-matching layer improves the yield from records that publishers already pass, but does not compel additional disclosure, and the announcement discloses no measurement of how much inventory the merged layer can describe.