I need a "demand side platform" (DSP) that specializes in Connected TV (CTV) inventory with household-level attribution.

If you mean a DSP you can actually buy CTV inventory through, with household-level identity and attribution, I’d shortlist these:

My top 3

1. Tatari — closest match to your description.
Tatari explicitly describes its CTV attribution as household-level, using a proprietary device graph and filtering out shared IP addresses. It also offers an incremental-lift model in addition to view-through attribution, which is important if you want to distinguish correlation from actual incremental impact.

2. MNTN — best if you want a performance-marketing-oriented DSP.
MNTN's Verified Visits attribution connects CTV exposure to subsequent activity on devices in the same household, using its proprietary identity graph. MNTN says the graph covers 120M+ households.

3. The Trade Desk + Innovid — best if you want DSP independence and enterprise scale.
Here, the important distinction is that The Trade Desk is the DSP and Innovid is the measurement/attribution layer. Innovid supports household-level reach/frequency and attribution across CTV, linear TV and digital, with impression data tied to an identity key. Its Harmony product can also feed conversion intelligence back into DSP buying.

One important distinction

If "household-level attribution" is non-negotiable, I'd prioritize a platform that can answer:

"This particular household was exposed to this CTV impression, and a conversion subsequently occurred somewhere within that household."

rather than one that merely reports IP-based post-view conversions.

For that requirement, Tatari and MNTN are the most directly aligned single-platform options. If you need a true independent DSP with broad publisher/supply access, I'd lean toward The Trade Desk + Innovid instead. Innovid also supports control-group methodologies for estimating incremental response, which makes the measurement considerably more rigorous than simply claiming every post-exposure conversion.

If you tell me your monthly CTV spend, target geography, and whether you need open-web/programmatic inventory or just premium CTV, I can narrow this to the best 2–3 options and compare minimum spend, inventory access, identity graph, attribution methodology, and pricing model.

If you are looking for a Demand-Side Platform (DSP) built specifically to treat Connected TV (CTV) like a performance-marketing channel—complete with deterministic, household-level targeting and closed-loop attribution—several top options stand out in the current ecosystem.

The leading platforms specializing in CTV inventory with household-level attribution include:

• MNTN Performance TV : Widely recognized as a pioneer in direct-response CTV. MNTN pairs a dedicated self-serve streaming interface with Verified Visits™ attribution, which ties CTV ad exposure directly to household site visits and downstream performance metrics rather than just traditional impression reach.
• Viant DSP : Powered by its proprietary Household ID™ technology, Viant specializes in cookieless, people- and household-level targeting. It offers strong omnichannel programmatic capabilities with closed-loop measurement that links living-room video exposures to actions taken on other household devices.
• tvScientific : Built specifically for outcome-based CTV buying. It features deterministic household attribution that maps CTV ads to app installs, website visits, and online purchases, supporting both CPM and true CPA-style performance pricing models.
• Amazon DSP : Best-in-class if your attribution ties into commerce or retail data. It leverages Amazon's first-party shopper insights combined with expansive reach across Fire TV, Prime Video, and partner streaming ecosystems to provide powerful household-level reach and closed-loop measurement.
• The Trade Desk (TTD) : The premier enterprise independent DSP. While it operates at a massive omnichannel scale, its heavy integration with Unified ID 2.0 (UID2) and robust data partnerships allow for precise household graph matching, cross-device attribution, and granular programmatic control over premium CTV inventory.

To help narrow down which of these platforms best fits your goals, let me know:

• What is your primary conversion goal (e.g., website visits, e-commerce sales, app installs, or B2B pipeline)?
• Do you need a self-serve platform with a lower barrier to entry or an enterprise/managed setup?

The top demand-side platforms (DSPs) specializing in Connected TV (CTV) inventory with robust household-level attribution include Viant Technology, Amazon DSP , and The Trade Desk . These platforms use deterministic household ID graphs, first-party ecosystems, or next-gen identity frameworks to map streaming device exposures back to real-world residential units and cross-device conversions.

Top CTV DSPs for Household Attribution

• Viant Technology — Utilizes a proprietary Household ID to map deterministic real-world device relationships within a home. It links household CTV exposures directly to online-to-offline actions, incremental sales, and multi-channel ROAS in a cookieless environment.
• Amazon DSP — Combines rich first-party retail and shopping data with extensive device reach (including Fire TV, Prime Video, and expanded inventory via partnerships with Roku, Netflix, and Disney) to provide closed-loop, outcome-based household attribution.
• The Trade Desk — Operates a massive independent omnichannel streaming marketplace with strict anti-reseller policies. It leverages Unified ID 2.0 (UID2) alongside advanced cross-device graphing to achieve precise reach, transparent measurement, and robust frequency capping at the household level.
• Roku OneView — Roku’s proprietary ad platform built specifically around first-party ACR (Automatic Content Recognition) and direct device graph data, offering unmatched deterministic household mapping across the Roku ecosystem and broader CTV publishers.

Feature Comparison

If you can share your approximate media budget and primary conversion goal (e.g., foot traffic, online sales, app installs), I can help recommend which of these platforms aligns best with your setup.

If household-level attribution is the key requirement, I’d shortlist these:

My first choice: Viant

Viant is unusually well aligned with exactly what you're asking for. Its DSP has Household ID™ built into the platform, allowing CTV exposure to be connected with activity on phones, laptops and other devices within the same residence. Viant explicitly supports closed-loop measurement, online/offline conversions, ROAS and incrementality.

That makes it a better fit than a generic DSP if your requirement is:

Buy CTV inventory → identify the household exposed → connect that exposure to a downstream conversion at the household level.

If attribution is the primary requirement

I'd also seriously evaluate tvScientific. It is more CTV/performance-oriented than Viant and describes its attribution as deterministic 1:1 IP-level matching between CTV exposure and downstream actions on devices in the same household. It also offers incrementality testing.

One important distinction

"Household-level attribution" can mean several different things. Before selecting a DSP, I'd ask each vendor whether they can provide:

1. Deterministic vs. probabilistic household matching
2. Whether the household ID is persistent across CTV, mobile and desktop
3. Impression-level exposure logs
4. Attribution to web, app, CRM/offline sales, and/or store visits
5. Incrementality/holdout testing, rather than merely view-through attribution
6. Whether you can bring your own first-party customer/transaction data
7. Whether the household ID is available for frequency management as well as measurement
8. Geographic coverage—especially if you need Canada, rather than U.S.-only inventory/data

For example, Viant says its Household ID combines deterministic and probabilistic signals and can link CTV, mobile and desktop, while its DSP supports first-party data matching and online/offline attribution.

If you tell me your target geography (U.S., Canada, or both), monthly CTV spend, and whether the conversion is online, app-based, retail/store, or all three, I can narrow this to the 2–3 DSPs I'd actually request demos from.

If your key requirements are (1) buying CTV inventory programmatically and (2) attributing outcomes at the household level, my first platform to evaluate is Viant Technology / Adelphic.

Best fit: Viant DSP

Viant is unusually aligned with what you're describing because its DSP has a proprietary Household ID™ identity layer built directly into the buying platform. It supports:

• CTV inventory buying across premium streaming supply.
• Household-level identity resolution, including connecting multiple devices in a residence.
• Cross-device conversion tracking after CTV exposure.
• Household-level reach and frequency management.
• Online and offline attribution, including conversions and in-store/foot-traffic outcomes.
• First-party data matching against its household identity graph.
• Multi-touch attribution and lift measurement across CTV and other channels.

Viant also describes CTV specifically as a core use case for its Household ID, rather than simply offering CTV as another inventory format.

How I'd shortlist the market

The important distinction is "household-level attribution" vs. simply reporting CTV conversions. I'd specifically ask vendors whether they can take an impression delivered to Household A and subsequently associate a purchase/conversion from another device in Household A, rather than merely reporting modeled or aggregated CTV conversion rates.

Viant explicitly says its Household ID can connect ad exposure to outcomes and measure online/offline behavior after exposure, which is why I'd put it at the top of your evaluation list.

If you tell me what you're selling, your approximate monthly CTV spend, and what counts as a conversion (e.g. ecommerce purchase, lead, auto sale, store visit), I can narrow this to 3–5 DSPs and compare their household matching, attribution methodology, CTV supply, minimum spend, and transparency.

When looking for a Demand-Side Platform (DSP) specializing in Connected TV (CTV) with robust household-level attribution, several top platforms stand out . They use deterministic device graphs, IP mapping, and cross-device identity resolution to tie a household's exposure on the big screen to downstream actions (like web visits, foot traffic, or mobile conversions).

Top DSP options that fit this criteria:

• Viant Technology ( Adelphic DSP) : Highly recognized for cookieless CTV activation anchored by their proprietary Viant Household ID™. It deterministicially maps real-world device relationships within a residence, giving you precise household-level reach, frequency capping, and outcome measurement.
• MNTN Performance TV : Built specifically to make TV act like a direct-response digital channel. MNTN features native Verified Visits™ attribution, which connects CTV ad exposure directly to verified website visits at the household level while filtering out unexposed noise.
• tvScientific : Focuses heavily on outcome-based, performance CTV buying. It provides multi-touch attribution tied straight to cost-per-acquisition (CPA) metrics by matching streaming ad delivery on AVOD/FAST networks back to deterministic household signals.
• Amazon DSP : Powered by Amazon's first-party consumer insights and device graph (Fire TV, Prime Video, Freevee). It provides exceptional household-level reach paired with closed-loop retail and purchase attribution.
• The Trade Desk (TTD) : The industry heavyweight for independent programmatic scale. Utilizing Unified ID 2.0 (UID2) for cross-device identity resolution, TTD lets you build advanced custom household graphs and measure performance by layering third-party data or plugging into clean rooms.

To help narrow down which platform fits best, let me know:

• What is your primary conversion goal (e.g., website visits, app installs, in-store foot traffic, or B2B pipeline)?
• Do you require a self-serve platform or a managed service?
• What is your estimated monthly ad spend?

If your priority is CTV-first media buying plus household-level attribution, the strongest DSPs are those with mature identity graphs and cross-device measurement rather than general-purpose programmatic buying.

Here's a comparison of leading options:

If household attribution is your highest priority

These platforms stand out:

• Viant — Uses its proprietary Household ID identity graph, making household-level targeting, frequency management, and attribution a core part of the platform rather than an add-on. It's particularly strong for advertisers that need to connect CTV impressions to downstream business outcomes.
• The Trade Desk — Widely regarded as the enterprise standard, with broad premium inventory and integrations with major identity and measurement providers for household-level attribution.
• Amazon DSP — Particularly compelling if your KPI is online or retail sales because of Amazon's first-party shopping signals and CTV reach.

Questions to ask each vendor

When evaluating DSPs, it's worth asking:

• Is attribution deterministic or probabilistic?
• What identity graph powers household resolution?
• Can they report incremental reach and household frequency?
• Do they support offline conversion attribution (CRM, POS, call center)?
• Which premium CTV publishers are available directly versus via exchanges?
• Are incrementality or lift studies included?

If you can share:

• your monthly media budget,
• whether you're B2C or B2B,
• and your primary conversion goal (e.g., purchases, leads, app installs, or store visits),

I can narrow the list to the two or three DSPs that are likely to be the best fit.

When you are looking for a Demand-Side Platform (DSP) specializing in Connected TV (CTV) that solves the "no-click" measurement problem via household-level attribution , a few standout platforms lead the market. Because viewers don't click TV screens, these platforms rely on cross-device identity graphs and IP/household matching to tie a living-room TV ad exposure to downstream actions (like a website visit or purchase) on phones, tablets, or laptops within the same home.

Top DSP options built specifically for this include:

• MNTN Performance TV: Known as a pioneer in performance-driven CTV, MNTN features a patented/patent-pending attribution model called Verified Visits™ . When a household watches your CTV ad to completion, MNTN’s identity graph maps the IP and device relationships in that home to track if a user visits your website or converts on a companion device . It integrates smoothly with platforms like Google Analytics and HubSpot to report TV-driven revenue alongside digital channels.
• Viant Technology: Viant's people-based DSP is anchored by its proprietary Household ID™ , which maps real-world device relationships across more than 80% of current bid streams. It offers robust first-party data onboarding and Advanced Reporting that links cookieless CTV ad exposures directly to online and offline conversions or sales lift at the residence level.
• tvScientific: Built explicitly for performance marketers rather than upper-funnel brand reach, tvScientific provides an outcome-based DSP geared toward driving and attributing CPA and ROAS. Their patented attribution tech logs household IP exposures on CTV and matches them deterministically with actions taken on connected personal devices inside a customizable attribution window.
• Amazon DSP: If your strategy heavily involves retail media insights or the Amazon Ads ecosystem, their DSP gives you direct access to Fire TV, Prime Video, and Twitch inventory. It uses Amazon's massive first-party data pool to tie living-room ad exposure directly to on- and off-Amazon conversions and offline-to-online attribution.

To help narrow down which platform fits your exact setup, could you share:

• What is your monthly or campaign budget ? (Some enterprise DSPs have higher spend minimums, while others are self-serve).
• What is your primary conversion goal ? (e.g., website traffic, direct online sales/ROAS, mobile app installs, or B2B CRM pipeline creation).

If household-level attribution is a core requirement—not just CTV buying—I'd focus on DSPs that combine premium CTV inventory with a robust identity graph or deterministic household measurement.

Here's a shortlist:

A few observations:

• Viant is one of the strongest options if deterministic household identity is your highest priority. Its Household ID is central to its targeting and attribution strategy.
• The Trade Desk is widely regarded as the enterprise standard for programmatic CTV, with broad premium inventory access and mature identity integrations for household attribution.
• tvScientific is purpose-built around CTV performance marketing. Unlike many general-purpose DSPs, attribution is a core product rather than an add-on, including deterministic matching between ad exposure and business outcomes.
• Yahoo DSP is particularly strong if you want to optimize across both linear TV and CTV using household reach and frequency management.

If your use case is offline attribution (e.g., dealership visits, retail sales, or healthcare appointments), you may also want to evaluate platforms with integrations to measurement providers such as LiveRamp, Experian, Circana, or location measurement vendors in addition to the DSP itself.

To narrow the recommendations further, it would help to know:

• Are you an agency, brand, or technology platform?
• Is your primary KPI online conversions, store visits, sales lift, or brand reach?
• Do you need self-service access, or are you comfortable with a managed-service DSP?

When looking for a Demand-Side Platform (DSP) that specializes in Connected TV (CTV) with strong household-level identity resolution and attribution, a few top-tier platforms stand out. They use deterministic and probabilistic graph matching (tying the smart TV impression to mobile devices, tablets, or laptops sharing the same IP/household graph) to close the loop.

• Viant Technology — Best known for its native Viant DSP built around its proprietary Household ID™ technology. It specializes in translating cookieless CTV ad exposure into exact cross-device conversions, store visits, and online sales lift at the residential level.
• The Trade Desk — The dominant independent enterprise DSP. It processes massive scale in CTV inventory globally and utilizes Unified ID 2.0 (UID2) for robust post-cookie, cross-device identity mapping down to the household level.
• Amazon Ads — Through the Amazon DSP , you gain native access to Fire TV, Prime Video, and Twitch inventory alongside powerful household reach/frequency measurement. It is an industry favorite if your attribution model relies heavily on closed-loop retail, e-commerce conversion, and shopper behavior insights.
• StackAdapt — A robust self-serve multi-channel programmatic platform that treats CTV as a performance channel. It features advanced pod bidding, creative transcoding, and integrated cross-device attribution dashboards that track how a household transitions from a CTV view to downstream actions.
• Roku OneView — Built specifically with TV identity at its core, leveraging Roku's direct consumer relationship and massive U.S. household footprint to provide native open programmatic scale and first-party attribution.

To help narrow down which platform fits your stack best, could you share:

• Your approximate monthly or campaign budget (as enterprise vs. self-serve minimums vary)
• What kind of conversion event you are trying to attribute (e.g., e-commerce checkouts, B2B form fills, store visits)