Q&A: Singular on Making CTV a Performance Channel

Connected TV has proven itself as a performance channel, but many app marketers still question how attribution really works. We sat down with Victor Savath, VP of Global Solutions Consulting at Singular, to unpack what makes CTV attribution reliable, how view-through measurement, attribution windows, and prioritization rules impact performance, and why re-engagement, assists, and multi-touch attribution are critical to understanding true incremental value. Watch the video to learn how to evaluate CTV with confidence and integrate it effectively into your performance marketing mix.

Watch the full interview here:

Victor, before we get into attribution, can you introduce yourself and what you do at Singular?

Absolutely. My name is Victor Savath, and I’m the VP of Global Solutions Consulting at Singular. For anyone unfamiliar, Singular is a next-generation mobile measurement provider focused on data movement and AI-driven insights. Attribution has always been close to my heart, so I’m excited to dig into CTV with you today.

Many mobile-first marketers still see CTV as a brand-awareness channel. How should they begin thinking of CTV as a measurable, performance-driven channel?

Coming from a traditional measurement background — particularly media mix modeling and even linear TV — the biggest shift CTV brought is accessibility and iterative control.
CTV isn’t a static upper-funnel channel. It allows rapid testing, granular optimization, real-time learnings, and performance-style scaling similar to what mobile marketers are used to.
While the mechanics differ, CTV can and should be treated as a performance channel rather than just “TV for awareness.”

What makes CTV attribution reliable today? What signals matter most?

It really comes down to the signal infrastructure built around CTV from day one. Unlike traditional TV, early CTV adopters demanded performance-level transparency — real-time delivery data, creative-level reporting, placement details, and sub-second timestamps.

As a result, CTV attribution uses:

• Sub-second impression timestamps
  
• Postbacks for every attributed conversion to feed optimization
• Visibility into the attribution decision
  
• Granular breakdowns by creative, placement, geo, etc.

Importantly, CTV grew up separate from the privacy disruptions that hit mobile (especially iOS). That allowed the ecosystem to maintain device-level attribution methods while still respecting privacy.

Are marketers becoming more comfortable working with less user-level information, especially after iOS privacy changes?

Yes — 100%. iOS privacy changes forced marketers to embrace probabilistic thinking and to rely on the best available signals rather than perfect deterministic ones.
This shift opened the door for more experimentation, including CTV. Marketers now recognize that:

• Multiple measurement methodologies can (and should) coexist
• SKAdNetwork introduced the idea of parallel models
• Incrementality testing and MTA provide complementary insights

How does Singular connect a CTV impression to a later mobile app install? What strengthens that connection?

It starts with capturing the touchpoint properly. Unlike mobile-to-mobile attribution, the conversion won’t happen on the same device. So Singular uses platform-agnostic tracking links that can attribute impressions to any device.

Captured signals may include:

• IP addresses and header-level proxies
• User agent or OS (for CTV delivered on tablets or similar devices)
• Sub-second timestamps
• Placement and creative details

Once an install or re-engagement occurs, Singular checks that event against all eligible touchpoints and applies the configured attribution windows and rules.

What is the ideal attribution window for CTV?

 Singular’s default is 24 hours.

A window that’s too long risks:

• Over-attribution
• Organic installs being incorrectly credited
• Excessive overlap with other channels

A window that’s too short risks missing genuine influence. One day is generally the most balanced and consistent with other view-through channels.

How do prioritization rules — especially click vs. view-through — impact CTV attribution?

This is arguably more important than the attribution window.

CTV is entirely view-through based, and in traditional attribution hierarchies, views are lower priority than clicks. Without adjustments, clicks from other channels will always override CTV impressions — even if the CTV touchpoints drove the outcome.

Prioritization rules allow marketers to:

• Treat CTV impressions as higher-value signals
• Ensure credit is fairly distributed
• Avoid suppressing CTV performance by default hierarchy biases

This is crucial for proper evaluation and optimization.

What about re-engagement? How does Singular attribute re-engagement events triggered by CTV?

We fully support CTV-driven re-engagement attribution.

CTV is naturally conducive to re-engagement because:

• Users often have multiple devices active while watching
• Exposure can prompt immediate in-app actions (e.g., ordering dinner after seeing a QSR ad)

We evaluate CTV impressions the same way we would for acquisition, but classify the resulting conversion appropriately as a re-engagement.

Some marketers worry CTV may “take credit” from other channels. How do you answer that?

This is exactly why Singular introduced MTA (multi-touch attribution) and assists.

Rather than debating “which channel gets the last click,” we now offer visibility into:

• Single-attributed conversions
  
• Co-attributed conversions
  
• Assist-based contributions
  

This paints a clearer picture of channel overlap and how CTV supports — rather than cannibalizes — other campaigns.

What are the most common setup mistakes marketers make when starting with CTV?

1. Incorrect or overly long attribution windows
   
2. Not configuring prioritization rules, leading to CTV being undervalued
3. Evaluating CTV as a single monolithic channel instead of analyzing:
   
   • Creatives

1. • Markets

1. • Audiences
2. • Placements
3. Treating CTV in isolation rather than understanding its role in the broader user journey
4. Failing to use household-level thinking, unique to CTV environments

CTV should be optimized with the same rigor as performance mobile — but with attention to its unique cross-device dynamics.

What KPIs should marketers use to evaluate CTV performance?

The core metrics still apply:

• CPI / Cost per re-engagement
  
• IPM (installs per mille) — more important in CTV
• ROAS and LTV
  

But with CTV, it’s critical to look at contribution metrics:

• Single-attributed conversions (net new users)
• Co-attributed conversions
  
• Assist rates (how CTV supports other channels)

If assist and single-attributed numbers are strong, you know CTV is driving real value, not just siphoning credit.

What developments will make CTV an even stronger performance channel in the future?

Two major areas:

1. Multi-touch attribution and assist modeling

CTV benefits enormously from the added transparency. Marketers can see exactly how CTV fits into the broader funnel and receive more accurate credit allocation.

2. Audience-based incrementality testing

This will allow marketers to:

• Split CTV-exposed vs. control audiences
• Measure incremental re-engagement or incremental conversions
• Combine learnings from MTA, incrementality, and MMP reporting for a triangulated view

Together, these will make CTV not just measurable but highly optimizable as a performance channel.

Any final thoughts for marketers still unsure about CTV?

Understanding attribution overlap and having transparency into signals reduces the fear of cannibalization. With clearer visibility — especially through MTA and assist data — marketers can confidently assess how CTV influences their entire ecosystem.