Summary
- The economic unit is a connected-TV or digital video impression bought through an adtech platform. Nexxen's claim is that an integrated DSP, SSP, data platform and CTV inventory package can raise the probability that the impression is real, relevant, measurable and operationally cheap enough to renew.
- The strongest public evidence is not a case study but the audited group record: 2025 revenue of $364.8 million, 627 active customers, 1,304 active publishers, 247.8 billion ad impressions, $353.1 million of Contribution ex-TAC and CTV revenue of $109.4 million in the 2025 Form 20-F (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The current update is more favorable: Q1 2026 CTV revenue rose to $29.4 million and programmatic revenue to $81.9 million (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm).
- The main doubt is that CTV quality is not self-proving. The buyer still has to ask whether the viewer was incremental, whether the supply path avoided waste, whether identity and ACR data were lawful and useful, whether measurement changed budget decisions, and whether a large walled garden could have delivered enough certainty with less operational friction.
- The judgment is constructive but conditional. Nexxen has credible assets in video, data and CTV, but the audited record also shows customer concentration, a 2025 CTV decline, reduced spend from one DSP customer after supply-path optimization, reliance on data and hosting partners, privacy pressure and intense competition.
The paid unit is a real viewer, not an auction event
Start with the buyer, not the platform. A retailer, auto brand, healthcare advertiser or entertainment marketer has a fixed media budget and wants incremental reach on the living-room screen. The buyer can pay Amazon for streaming TV ads across Prime Video, Twitch, Fire TV Channels and third-party supply, with Amazon claiming 200 million-plus monthly ad-supported U.S. reach and first-party shopping and streaming signals (https://advertising.amazon.com/solutions/products/streaming-tv-ads). Or the buyer can keep money in YouTube, Meta video, retail media, a direct broadcaster deal, a publisher's private marketplace, or an agency desk that hides the messy plumbing behind a service fee.
The unit being bought is therefore not simply "one impression". It is an operating outcome: a chance to put a video or CTV message in front of a person who is probably real, probably in the intended audience, probably watching acceptable inventory, probably not duplicated too often, and probably measurable against a business result. That probability is the product. Nexxen earns its place only if it raises that probability enough, or lowers the effort of managing it enough, to beat the substitute.
The burden transferred to Nexxen is large. The buyer wants a platform to value each bid, avoid bad supply, join identity signals across screens, apply frequency controls, handle creative and delivery standards, produce reports, manage publisher access and give the media team evidence that the next campaign should renew. If the advertiser can get easier certainty from Amazon's commerce graph, a broadcaster's direct sales team or a smaller trusted list of publishers, Nexxen's take-rate has to be justified by broader reach, better data, cheaper supply, better yield or less operational work.
The best public proof is narrower than the sales language. Nexxen International Ltd.'s 2025 Form 20-F shows the group generated revenue from transactions where it provides a platform for the purchase and sale of digital advertising inventory; it also describes fees based on percentage of spend, flat fees or fixed CPMs (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The Q1 2026 6-K shows a strong current quarter, with revenue of $86.8 million, programmatic revenue of $81.9 million and CTV revenue of $29.4 million (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm). Those filings prove that advertisers and publishers are using the platform at meaningful scale. They do not prove that every impression is superior to an Amazon, Google, Meta, direct publisher or buyer-side desk alternative.
That distinction matters because adtech often describes itself with the language of control while selling into a market defined by uncertainty. A buyer paying a CPM is buying a statistical claim about attention, identity, placement and outcome. A seller such as Nexxen has to turn that uncertainty into a manageable renewal decision. The thesis to test is whether Nexxen has enough differentiated supply, data and measurement to make a paid impression more likely to be a viewer-quality event than wasted spend.
The UK entity is the directory anchor, but group evidence carries the economics
The assigned entity is NEXXEN GROUP LTD in the United Kingdom. Companies House lists company number 05411297 as active, incorporated on April 1, 2005, with a registered office in High Holborn, London, and business classifications covering video distribution activities, other information technology service activities and advertising agencies (https://find-and-update.company-information.service.gov.uk/company/05411297). The same public record shows previous names including UNRULY GROUP LIMITED from March 2011 to June 2023 and UNRULY MEDIA LIMITED from October 2006 to March 2011. That is important because Unruly is part of the history that ties the UK row to a video advertising business rather than to a generic corporate shell.
Most operating and financial evidence, however, is group-level evidence under Nexxen International Ltd., the listed foreign issuer. The investor site describes Nexxen as an advertising technology platform made up of a DSP, an SSP and the Nexxen Data Platform, headquartered in Israel and traded on Nasdaq under NEXN (https://investors.nexxen.com/). The current marketing site describes a DSP, SSP and data platform working together for advertisers, publishers and data partners (https://nexxen.com/). The 2025 annual filing places the listed company's principal executive office in Tel Aviv and gives consolidated results. The public record therefore supports a group economics article about the Nexxen platform, with the UK directory entity as the GB corporate anchor. It does not support a claim that the UK subsidiary alone produced the group revenue, CTV growth, customer count or publisher footprint.
The one private metric that would settle the thesis more cleanly is not public: net revenue retention and gross margin by connected-TV buyer cohort after fraud filtering, supply-path changes and identity costs. If a CTV buyer that shifted from a walled garden or direct publisher bundle to Nexxen kept renewing at higher spend after measurement, the platform claim would be much stronger. If renewal depended mainly on low CPM reach or one-off political cycles, the case would be weaker. Public filings give enough to judge direction, not enough to audit unit quality.
That is why the article treats ASNs, RIPE context, domains, seller IDs, bid records, data partnerships and platform pages as evidence of operating surface, not as entities in their own right. The subject is a company selling an advertising technology operating unit. The evidence is about whether its machinery makes an impression worth more than the buyer's alternatives.
Nexxen's business model is a take-rate on reducing uncertainty
Nexxen's public product map is straightforward. The DSP is pitched as a path to omnichannel media using proprietary data, differentiated ad experiences, optimization and measurement (https://nexxen.com/nexxen-dsp/). The SSP is pitched to advertisers as scaled supply with native units, proprietary data and creative enhancements, while the company argues that an integrated DSP and SSP means fewer intermediaries between advertiser and publisher (https://nexxen.com/nexxen-ssp-for-advertisers/). The data platform says it brings together first-party data, Nexxen data assets and partner data for onboarding, enrichment, expansion and activation (https://nexxen.com/nexxen-data-platform/).
Economically, this is a take-rate on coordination. The advertiser could assemble the stack itself: one DSP, a clean room, a measurement vendor, multiple SSP paths, several publisher PMPs, a fraud vendor, a brand-safety vendor and a reporting layer. That can work for a sophisticated media team, but it raises people cost, data-contract cost and reconciliation cost. Nexxen's offer is that integration can reduce those costs and let the company capture more of the value chain. The investor-day deck states the same logic more explicitly: onboarding first-party data, enriching audiences, activating media, measuring and optimizing are presented as one connected commercial loop (https://investors.nexxen.com/static-files/3451b0b7-c4e1-4798-af8b-3e00efa10bf8).
The audited numbers show why the distinction between gross revenue and contribution matters. In 2025, revenue was $364.8 million, cost of revenue excluding depreciation and amortization was $55.0 million, research and development was $58.1 million, and selling and marketing was $123.0 million (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). Contribution ex-TAC was $353.1 million, up 2.8% from 2024. Nexxen defines Contribution ex-TAC as gross profit plus certain revenue-cost items minus Performance media costs, which reflects management's view that some media costs are passed through rather than controlled inventory economics.
This definition is useful but not magic. A buyer does not care whether a cost sits above or below a management metric if the campaign fails. The buyer cares whether platform fees, data fees and service effort leave enough working media to move an outcome. The take-rate is justified when the platform prevents waste that would otherwise be bought: duplicate reach, non-viewable delivery, weak inventory, irrelevant audience matches, broken creative, bad frequency or reporting that cannot influence spend. The take-rate is exposed when a large buyer can optimize the supply path itself.
The 2025 filing gives a clear example. Nexxen said revenue was constrained in part by "a significant reduction in spending by one DSP customer" in the open marketplace channel as part of that customer's supply-path optimization efforts (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). That sentence is a small but important window into platform economics. A DSP customer can decide that a supply path is not worth the fee, duplication, opacity or inventory mix. Nexxen's answer has to be either differentiated supply that cannot be cheaply replicated, or measurement and data that make the path demonstrably better.
Scale exists, but scale and quality are not the same thing
Nexxen is not a marginal tool. The 2025 Form 20-F reports 627 active customers, 1,304 active publishers and 247.8 billion ad impressions in the year (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The filing also says the platform served ads in about 180 countries and had direct access to 184,090 active sites and apps. A public sellers.json file under the legacy Tremor domain lists 1,802 seller entries, including publishers and intermediaries, and carries a TAG-ID identifier (https://tremorhub.com/sellers.json). That is a real public operating surface.
But scale can hide opposite realities. A large supply file may mean breadth, resilience and negotiated access. It can also mean more paths to monitor, more resellers to classify and more ways for buyers to ask whether they are paying unnecessary tolls. In programmatic advertising, the question is not only "how much supply can I reach?" It is "which supply path gives me the best probability of a useful viewer at an acceptable clearing price after fees and risk?"
Nexxen's own policies show that the company understands this. Its advertiser policy bans misleading formats, malware, fake interactions, low-resolution CTV creative and other behavior that would corrupt ad quality (https://nexxen.com/advertiser-policy/). Its publisher policy covers content standards, fraud prevention, ad refresh, onboarding, monitoring and review; it says CTV apps must be approved by platform operators and that Nexxen may use MRC-accredited vendors in determining invalid activity (https://nexxen.com/publisher-policy/). These rules are valuable because ad quality is a cost-control system, not a compliance ornament. Every fake view, bad refresh, hidden placement or unsuitable creative converts media budget into leakage.
The harder point is that policies do not prove enforcement quality. They define the contract. The buyer still needs evidence that the platform enforces the contract faster and more consistently than substitutes. The public filing acknowledges the same problem, warning that fraudulent, deceptive or malicious activity may attempt to inflate or divert advertising spend, generate fraudulent activity or misuse the platform (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The company says it uses third-party tools and industry collaboration, but no platform can make fraud a solved issue.
That makes "viewer quality" the right frame. A CTV impression bought through Nexxen has value only if the viewer and environment are better than the cheapest auction event. The economics of a scaled SSP/DSP platform depend on selection, not merely volume.
Connected TV is the strategic story, with a 2025 warning label
Nexxen's most persuasive commercial story is connected TV. The company says video represented 66% of total revenue in 2025 and about 71% of revenue excluding non-programmatic performance activity (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). It presents CTV, mobile video and programmatic home-screen units as channels less exposed to search-page disruption and more tied to logged-in, high-attention environments. The investor-day deck says CTV was 36% of programmatic revenue in Q1 2026 and 32% for full-year 2025 (https://investors.nexxen.com/static-files/3451b0b7-c4e1-4798-af8b-3e00efa10bf8).
The market tailwind is real. IAB and PwC reported that U.S. digital advertising revenue reached nearly $300 billion in 2025, up 13.9% year over year (https://www.iab.com/insights/internet-advertising-revenue-report-full-year-2025/). Nielsen-related coverage reported that streaming represented 44.8% of total U.S. TV usage in May 2025, edging past broadcast and cable combined for the first time (https://www.tvtechnology.com/news/viewing-of-ad-supported-services-grew-to-73-6-percent-of-tv-viewing-in-q2). Nexxen's own 20-F cites eMarketer expectations for U.S. CTV ad spend to grow at about a 12% CAGR from 2025 to 2029 (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm).
Yet the audited 2025 result contains the warning. Nexxen's CTV revenue fell 3.8% to $109.4 million from $113.8 million in 2024, even as the wider CTV category was expected to grow. The company attributed the decline to macroeconomic uncertainty, more competitive CTV CPMs, reduced spend by some customers due to U.S. trade dynamics and tariffs, the absence of political CTV spend versus 2024, and a major reduction by one DSP customer in the open marketplace channel (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm).
That does not break the thesis, but it makes it conditional. A category can grow while a platform temporarily loses share, price or mix. The Q1 2026 update was stronger: CTV revenue of $29.4 million was up 12% year over year, and management said Q2 strength was driven particularly by CTV, mobile and data products (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm). The decisive question is whether Q1 2026 was the start of a durable return to CTV share gain or a quarter helped by timing, easier comparison, political and event budgets, home-screen novelty and enterprise customer onboarding.
CTV is also no longer a niche product. IAB Tech Lab's CTV Ad Portfolio guidance, updated in June 2026, defines and standardizes formats such as pause ads, screensaver ads, menu ads, in-scene insertion, overlays and squeeze-back ads (https://iabtechlab.com/standards/ctv-ad-portfolio/). That standardization can help companies like Nexxen scale advanced units, but it also lowers the mystery premium. Once formats become standardized, differentiated access, data and measurement matter more than format novelty.
Amazon is the clearest substitute because it prices certainty
Amazon is the cleanest comparator because it turns the CTV buy into a walled-garden certainty story. Its streaming TV ads page says full-screen, non-skippable video ads can appear before, during or after content across Prime Video, Twitch, live sports, Fire TV Channels and third-party TV publishers and broadcasters (https://advertising.amazon.com/solutions/products/streaming-tv-ads). It also advertises no minimum spend for one route, DSP self-service and managed-service options, retail and streaming signals, and the ability to use 20,000-plus first-party audiences.
For a buyer, the substitute is not merely inventory. It is fewer explanations. Amazon can say: here is the audience, here is the shopping signal, here is the media environment, here is the interface, here are the campaign minimums, and here is the measurement system. That is why walled gardens take budget even when buyers complain about opacity. They reduce cross-vendor work and convert uncertainty into a single platform relationship.
Nexxen's response is different. It cannot out-Amazon Amazon on retail identity, and it cannot out-YouTube YouTube on logged-in video scale. Its claim is that the open internet still contains valuable video, CTV, publisher and OEM supply, and that a platform with data, identity resolution, measurement and supply control can make that open supply perform. Its DSP page argues that integrated DSP and SSP access means fewer hops, less waste, lower cost and cleaner data (https://nexxen.com/nexxen-dsp/). Its SSP page claims top-10 omnichannel reach, citing Jounce Media Exchange Profiles from April 2026 (https://nexxen.com/nexxen-ssp-for-advertisers/).
The buyer's decision becomes a portfolio question. Amazon may be the easiest choice for commerce-heavy brands and Prime Video reach. Direct publisher deals may be best for a tentpole show, sports event or sponsorship. Google and Meta may be better for platform-native video and social conversion. A buyer-side desk may be better if the brand wants to keep supplier optionality and own the model. Nexxen has to win where the buyer wants broad video reach, CTV data, home-screen units, supply-path efficiency and cross-screen measurement without surrendering the whole plan to one giant platform.
This makes price discipline essential. If Nexxen's CPM is low because supply is weaker, the bargain is false. If it is high because data, home-screen placement and measurement increase incremental outcomes, the premium can renew. The company therefore competes in the space between "cheap reach" and "closed certainty."
Identity and TV data are the strategic bet
Nexxen's data story centers on TV viewership signals and identity resolution. Its ACR page says the company has an exclusive partnership with Hisense and V, formerly VIDAA, bringing automated content recognition data across the open internet, and says its TV Intelligence product integrates set-top-box, ACR and streaming viewership data from 45 million-plus global households (https://nexxen.com/acr-data/). The Nexxen TV page says its cross-platform approach uses TV, web and social data to reveal what audiences watch, how they behave across screens and where they are reachable (https://nexxen.com/nexxen-tv/). The TV Intelligence page repeats the 45 million-plus household footprint and describes activation and measurement for fragmented TV viewing (https://nexxen.com/tv-intelligence/).
The filings show that the V relationship is not just marketing copy. Nexxen invested $25 million in V on August 18, 2022 and another $20 million on August 7, 2025; as of December 31, 2025 it held 759,453 ordinary shares and had committed to an additional equity investment expected in 2026, subject to conditions (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The Q1 2026 6-K says Nexxen expected to invest another $15 million in V during Q3 2026, bringing total investment to $60 million and about 6% equity ownership (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm).
That matters because ACR and TV identity are ways to escape the weakest version of programmatic video: anonymous auctions for loosely described viewers. If Nexxen can know what households watch, what ads they saw, which apps they use and how to control overlap across screens, it can help buyers allocate spend away from low-quality impressions and toward incremental reach. Its identity resolution page says the future requires multiple strategies, including a proprietary identity graph with deduplication across first-party data, targeting and measurement (https://nexxen.com/identity-resolution/).
The risk is that data advantage is both expensive and politically fragile. The 20-F warns that the platform depends on proprietary and third-party data, and that third-party data providers could increase prices, restrict use or enter exclusive arrangements with competitors (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). It also says Nexxen collects ACR data that includes device characteristics, online browsing behavior, exposure to and interaction with ads, and inferential data about purchase intentions and preferences. That is commercially valuable, but it sits directly in the path of privacy law, consumer choice, platform restrictions and public suspicion about smart-TV tracking.
Nexxen's services privacy policy says some disclosures may be considered sale, sharing or targeted advertising under U.S. state privacy laws, and it identifies participation in the IAB Europe Transparency and Consent Framework with supply-side platform ID 36 and demand-side platform ID 23 (https://nexxen.com/services-privacy-policy/). That disclosure is a sign of institutional seriousness. It is also a reminder that the data layer is not a free natural resource. If consent, opt-out signals, operating-system rules or TV-platform policies reduce signal availability, the viewer-quality edge can narrow.
Fraud, low-quality supply and supply-path pressure are not side issues
The open internet's appeal is reach. Its burden is quality control. The Wall Street Journal's 2024 coverage of made-for-advertising sites reported industry concern that such sites can deliver poor user experience, dubious advertiser results and heavy auction volume; it cited an Association of National Advertisers estimate that made-for-advertising publishers won roughly 15% of automated online advertising spend, or $10 billion in annual ad revenue (https://www.wsj.com/articles/made-for-advertising-websites-are-the-marketing-industrys-latest-messy-situation-560c79de). That article was not about Nexxen specifically, but it describes the market condition every open-web video platform has to overcome.
CTV adds its own version of the same problem. A viewer on a smart TV looks premium, but the supply chain can still contain resellers, poor app quality, measurement gaps, device-level ambiguity and creative rendering risk. IAB Tech Lab's work on CTV formats exists partly because the industry needs common language and signaling for advanced units (https://iabtechlab.com/standards/ctv-ad-portfolio/). Nexxen's ad specs page discusses TV Intelligence, Hisense VIDAA strategic partnership, creative optimization, green media products, attention solutions, first ad slot targeting and partner audience segments (https://nexxen.com/ad-specs/). These are tools for making a buyer believe that a high-CPM screen can be bought with enough control.
The public sellers.json file illustrates both the promise and the governance task. It lists a large network of publisher and intermediary entries (https://tremorhub.com/sellers.json). For a buyer, that transparency is useful only if it is operationalized: duplicates removed, intermediaries ranked, directness rewarded, bad paths filtered and reporting made clear. A long list without curation is not quality. It is a map of places where quality has to be judged.
Nexxen's filings show that buyers are already making those judgments. The 2025 reduction by one DSP customer tied to supply-path optimization is a direct economic signal (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). Supply-path optimization is not a fashion term here. It means a buyer or DSP asks which routes are redundant, which sellers add value and which take-rates can be removed. Nexxen benefits when it is judged a necessary path. It loses when it is judged an avoidable hop.
That pressure should discipline the company. The strongest version of Nexxen is not "we can reach everything." It is "we can reach enough high-quality video and CTV supply, with enough data and measurement, through paths that buyers keep after pruning." The difference between those two claims is the difference between recurring economics and auction leakage.
Measurement is the renewal test
Connected-TV advertising has an old television advantage and a digital advertising burden. It is immersive, sound-on and living-room oriented. It is also harder to click, harder to attribute and easier to overstate if the buyer relies on proxy metrics. Nexxen's measurement page addresses that directly. It describes cross-device attribution linking linear, CTV, online video and digital exposures to results; third-party validation; brand lift; incremental reach; app downloads; foot traffic; sales lift; healthcare; attention; tune-in by episode; cost per tune-in; and real-time optimization (https://nexxen.com/measurement/).
This is commercially central because measurement is how a media team defends renewal. A campaign manager can tolerate complexity if the platform answers budget questions faster: Did CTV add households not reached by linear? Did frequency overlap waste money? Did exposed households visit stores, buy, install, tune in or search? Did a home-screen unit create incremental reach, or did it merely add cost to people who would have seen the brand elsewhere?
The investor-day deck leans hard into operating efficiency. It claims a new interface reduced onboarding and training times by about 50%, and Q1 2026 disclosures say buyers using the enhanced interface reported efficiency gains and fewer steps, while the platform assistant improved troubleshooting, quality assurance and reporting efficiency (https://investors.nexxen.com/static-files/3451b0b7-c4e1-4798-af8b-3e00efa10bf8; https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm). Those are company-reported metrics, not independent audits, but the economic logic is sound. If a platform reduces the labor needed to launch, monitor and prove campaigns, it can defend a fee even when the media itself is comparable.
The danger is that every competitor now sells measurement. Amazon connects streaming ads to shopping signals (https://advertising.amazon.com/solutions/products/streaming-tv-ads). Retail media networks sell closed-loop purchase proof. Google and Meta sell platform-native conversion systems. Direct publishers sell brand lift, sponsorship and first-party audience packages. Measurement has become table stakes.
Nexxen's opportunity is cross-screen independence. A buyer that wants to compare linear, CTV, mobile video and open-web video across more than one media owner may not want a walled garden to grade its own homework. Nexxen can win if its measurement is credible enough to help buyers move budget across channels. It will struggle if its reporting is perceived as another vendor dashboard that requires separate reconciliation.
The cost base reveals what the impression has to pay for
Every platform story eventually returns to cost. Nexxen's 2025 revenue of $364.8 million was almost flat year over year, while R&D rose to $58.1 million and selling and marketing rose to $123.0 million (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). Cost of revenue excluding depreciation and amortization fell to $55.0 million, partly because non-programmatic performance costs declined, but Nexxen also cited a $4.3 million increase related to expanding data capabilities and capacity under a strategic partnership.
These costs are not generic overhead. R&D is the cost of keeping the bidder, identity graph, measurement products, interface and CTV support relevant. Selling and marketing is the cost of persuading sophisticated buyers and publishers that Nexxen deserves budget despite larger alternatives. Hosting and data costs are the cost of ingesting bid requests, ACR signals, campaign events and reporting records at ad-market speed. The paid impression has to carry all of that.
The balance sheet gives the company room, not immunity. The Form 20-F reports $133.3 million of cash and cash equivalents at December 31, 2025 (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The Q1 2026 release reports $94.6 million of cash and cash equivalents, no long-term debt and a $50 million undrawn revolving credit facility at March 31, 2026 (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm). That is a healthier capital posture than many smaller adtech businesses, but the cash decline also shows that working capital and strategic investment can absorb resources quickly.
Capital allocation is a signal. Nexxen repurchased shares and invested in V, indicating management believes the equity and the CTV data partnership are underappreciated. But the buyer of an impression does not pay for buybacks. The buyer pays for less waste, better outcomes and lower operating burden. If R&D and data spending create that, they are productive. If they merely keep pace with Amazon, Google, Meta, The Trade Desk, Magnite, publisher-owned platforms and retail media, they are defensive costs.
The unit economics therefore depend on renewal quality. The 20-F reports Contribution ex-TAC per active customer of $563,204 in 2025, up from $526,035 in 2024, while active customers fell from 653 to 627 and active publishers fell from 1,516 to 1,304 (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). Management attributed part of this to discontinuing smaller, less profitable customers and focusing on larger enterprise relationships. That is sensible if larger customers use more of the stack and renew. It is risky if concentration gives large buyers more leverage to demand lower fees.
Customer concentration turns quality into bargaining power
Nexxen's business is not dominated by one named buyer, but concentration is visible. In 2025, two buyers represented 12.1% and 11.3% of revenue; at year-end, two buyers accounted for 22.6% and 10.7% of trade receivables (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). The filing also says master service agreements with most DSPs and other customers generally auto-renew for one-year terms but can be terminated for convenience on 30 days' notice.
That is the commercial reality of adtech. The same large customers that validate scale can move spend quickly. A global holding company, major DSP, retail brand or performance buyer can test a platform, push spend, then prune the path if CPMs, outcomes or reporting disappoint. The short termination window means Nexxen's revenue quality depends less on contract length and more on daily usefulness.
This is why the company emphasizes enterprise customers and deeper stack use. Investor-day materials say enterprise customers need integrated platforms rather than point-solution DSPs, and that larger budgets, longer-term relationships and higher retention can follow deeper adoption (https://investors.nexxen.com/static-files/3451b0b7-c4e1-4798-af8b-3e00efa10bf8). The 20-F reports a 92% Contribution ex-TAC retention rate for 2025, down from 102% in 2024, partly due to the deliberate exit from smaller customers and partly due to macro, tariff, political and supply-path factors (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm).
The buyer-quality frame explains both the opportunity and risk. If a customer uses only one piece of Nexxen, the platform is easy to replace. If a customer uses data onboarding, identity matching, CTV supply, home-screen units, optimization, measurement and reporting, the relationship becomes harder to unwind. But that deeper lock-in is defensible only if it improves outcomes. Otherwise, it becomes the complexity buyers are trying to escape.
The public record is encouraging on adoption but not decisive on causality. The Q1 2026 release says Nexxen had already onboarded more new enterprise customers in 2026 than in all of 2025 and cited adoption by names including The Trade Desk, StackAdapt, Basis, H/L, TCL FFALCON and TiVo Ads for home-screen or related CTV access (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm). That shows market interest. The missing proof is how much of that interest becomes recurring net spend after initial tests.
There is also a timing issue in the evidence. Home-screen adoption, interface changes and new enterprise onboarding are recent enough that the strongest public numbers lag the strategic claim. Q1 2026 shows renewed momentum, but a buyer deciding on a 2026 plan would still want to see second-half renewal, not only announcements and early adoption. That does not make the claim weak; it defines the stage of proof. Nexxen has moved from "can it build the CTV and data surface?" to "can the surface survive repeated budget reviews by large buyers?" The second question is harder and more valuable.
Competition is not only another DSP
Nexxen competes with independent adtech platforms, but the more important substitutes are budget systems. Amazon offers streaming TV, Fire TV home-screen placement, retail audiences and commerce measurement in one platform (https://advertising.amazon.com/solutions/products/streaming-tv-ads). Google and Meta absorb video budgets through their own consumer surfaces and measurement. Direct publisher sales teams sell sponsorship, context and guaranteed adjacency. Retail-media networks sell shopper identity and closed-loop attribution. Large buyer-side desks sell procurement discipline and cross-platform planning.
The Trade Desk is a useful market signal because it shows where independent demand-side buying is headed. Business Insider reported in March 2026 that LinkedIn selected The Trade Desk as its first DSP partner for connected-TV ads, allowing advertisers to use LinkedIn's professional data for CTV targeting; the same report framed the move as part of the open internet's attempt to compete with walled gardens (https://www.businessinsider.com/linkedin-partners-with-the-trade-desk-for-ctv-ads-2026-3). That is not about Nexxen directly, but it shows that independent platforms are trying to add proprietary data and premium supply to avoid becoming commodity pipes.
Nexxen's differentiators are narrower but real. The V/Hisense ACR relationship, TV Intelligence household data, CTV home-screen access, integrated DSP/SSP design and video heritage from Unruly and Tremor give it a clear identity. Companies House ties the UK entity to Unruly's corporate lineage (https://find-and-update.company-information.service.gov.uk/company/05411297), and the group filings show video remains the central revenue format (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm).
The vulnerability is that every differentiator can be attacked. Walled gardens have better first-party identity. Large DSPs have broader buyer adoption. SSPs can build direct publisher paths. OEMs can sell home-screen inventory themselves or through multiple partners. Broadcasters can package CTV with linear. Retail media can pull budgets away from open-web video because the sales result is clearer. If advertisers cut experimental spend, independent adtech often feels the cut before the largest platforms do.
Nexxen's best defense is to be the platform that helps buyers compare and improve across these alternatives, not merely another exchange. If it reduces path waste, adds CTV and ACR signal, and produces measurement that helps budget move intelligently, it can earn a role. If it becomes a seller of undifferentiated impressions in a market with falling CPMs, the substitute wins.
Regulation, data locality and operating risk cap the upside
Adtech's legal risk is structural because the product depends on data moving among advertisers, publishers, platforms, measurement providers, verification providers and identity vendors. Nexxen's services privacy policy lists these ecosystem participants and data partners, and says it uses personal information for personalized advertising, contextual advertising, measurement, ID synchronization and identity graph building (https://nexxen.com/services-privacy-policy/). The 20-F warns that restrictions on cookies, mobile device IDs, CTV tracking and other technologies could reduce platform effectiveness (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm).
This is not only a legal department issue. Privacy rules change the cost of an impression. If a platform loses signal, it may need more impressions to hit the same outcome. If consent management adds friction, match rates can decline. If data partners charge more or restrict usage, the platform's margin or effectiveness can suffer. If a TV operating system changes permission rules, ACR-based targeting can become less scalable. Data sovereignty and locality matter because cross-border data handling, consent validity and local privacy law can shape which signals are usable in each market.
The company also has geopolitical exposure. The 20-F describes risks related to employees and location in Israel, and its Q1 2026 forward-looking language references conflicts involving Israel, Iran, Hamas, Hezbollah and the Houthis as potential business and market risks (https://www.sec.gov/Archives/edgar/data/1849396/000117891326002619/zk2635298.htm). That does not mean operations are impaired today. It means continuity, hiring, investor perception and customer comfort can be affected by events outside the ad market.
Cloud and hosting dependency is another practical constraint. The annual filing warns that disruptions to third-party data center hosting facilities and cloud computing and hosting providers could impair service delivery (https://www.sec.gov/Archives/edgar/data/1849396/000117891326000712/zk2634437.htm). In adtech, latency and availability are not back-office concerns. A platform that cannot bid, deliver, pace, measure or report reliably loses spend quickly. The buyer may forgive a weak experiment; it rarely forgives delivery failure during a live campaign.
These risks do not make the company unattractive. They describe the price of being an open, data-rich adtech platform. The same openness that lets Nexxen reach across publishers, devices and buyers also exposes it to privacy, fraud, infrastructure and counterparty risk. The investment case depends on whether the data and supply advantages more than offset those costs.
What would change the judgment
The base judgment is that Nexxen has a credible but not fully proven answer to the viewer-quality problem. The company is not merely selling a commodity impression. It has video scale, CTV focus, ACR data, a linked DSP and SSP, privacy disclosures, public quality policies, meaningful cash, current Q1 2026 growth and a plausible enterprise strategy. The company also has a 2025 CTV decline, lower retention, large-buyer concentration, reliance on data partners, exposure to supply-path pruning and intense competition from platforms that can price certainty more simply.
Three facts would strengthen the thesis materially. First, public cohort evidence showing enterprise CTV buyers increasing spend after independent incrementality measurement, not merely after onboarding. Second, audited or third-party evidence that Nexxen's integrated DSP-plus-SSP path reduces waste or improves outcome cost versus other DSP and SSP combinations across comparable campaigns. Third, clearer disclosure of how much revenue comes from unique CTV home-screen inventory, V/Hisense ACR data and direct publisher access rather than broadly available open-market supply.
Three facts would weaken it. First, another year in which category CTV spend grows but Nexxen's CTV revenue underperforms. Second, further reductions by large DSP or enterprise customers after supply-path reviews. Third, privacy or platform changes that reduce ACR or cross-screen identity availability without an offsetting contextual or first-party data product.
For now, the practical answer for a buyer is neither blind trust nor dismissal. Nexxen should be evaluated as a specialist in making CTV and video impressions less uncertain. The buyer should compare it against Amazon or another walled garden for certainty, against direct publisher deals for guaranteed context, and against buyer-side desks for procurement control. Nexxen earns the impression when it can show that its data, supply access, measurement and operating tools make a real viewer more likely than wasted spend. That is a demanding standard, but it is the right one: in modern CTV and video advertising, the impression is cheap only after the viewer-quality problem has been solved.