On December 4th, Business of Apps ran a webinar in partnership with Verve that featured speakers from Verve, Dataseat, and Kochava. Moderated by Nicole Weiss, the session focused on how Connected TV (CTV) is becoming a viable and measurable channel for user acquisition.
The discussion brought together the perspectives of a DSP, an SSP, and an MMP to give marketers a clearer understanding of what is now possible when inking CTV exposure to mobile installs.
Speakers included:
- Nikunj Sureka, Sr Director Product Management, CTV at Verve
- Alexei Moltchan, VP of Product Management at Dataseat (Now part of Verve)
- Julia Kan, Managing Director, Advertising Solutions at Kochava
Host:
- Nicole Weiss, Founder at Brass Finch
CTV is shifting from a branding medium to a measurable performance channel
The speakers agreed that CTV has undergone meaningful changes in the past few years. Increased ad supported inventory has created more liquidity and publishers have become more familiar with programmatic monetization. This shift allows buyers to transact at an impression level instead of purchasing large upfront packages.
The result is a channel that is more accessible for performance-driven teams. Alexei Moltchan noted that measurement and optimization were previously constrained by the lack of granular signals, but the ecosystem has matured enough for DSPs to model outcomes with a high degree of confidence.
Julia Kan also explained that this shift accelerated after the Covid pandemic when streaming viewership spiked and advertisers demanded more accountability from CTV platforms.
Making CTV work for UA: From exposure to install
Source: Business of Apps via YouTube
Measurement has become more practical due to advances in household-level attribution
A central point in the discussion was the progress made in connecting CTV impressions to mobile installs. CTV inventory is not clickable, so all attribution flows through probabilistic matching using the household IP.
Julia Kan pointed out that IP based methods are more reliable for CTV because televisions remain stationary while mobile devices can move across networks. To address multi-day conversion delays and fragmented identifiers, Kochava has built a household graph known as DeviceLink. It maps devices within the same home to allow attribution to succeed even when the CTV impression and the mobile install occur on different networks. This gives marketers a more accurate view of CTV’s contribution to both direct installs and downstream channel uplift in areas such as search and social.
Contextual and environmental signals are becoming the foundation of effective CTV buying
The panel emphasized the importance of contextual intelligence as a performance lever. Since CTV is typically a one-to-many environment, identifying the individual viewer is challenging.
DSPs must rely on signals such as content type, viewing format, time of day, device characteristics and household patterns to estimate the likelihood of an install. Alexei Moltchan explained how Dataseat’s models use a wide set of environmental attributes to interpret user intent.
Nikunj Sureka, on the other hand, expanded on the types of signals actually available in the bidstream. These include IP address, user agent, device identifiers when permitted, and content level metadata. The presence or absence of these signals affects how accurately a DSP can optimize and how clearly an MMP can attribute.
Final thoughts
Signal variability, data permissions, and implementation gaps still shape what marketers can reliably measure on CTV. Progress depends on how well SSPs, DSPs and MMPs align their data and delivery frameworks, since performance outcomes improve only when the full chain operates in sync.
Greater transparency in how signals are passed, stronger consistency in device and content metadata, and closer collaboration on integration standards can all raise the ceiling on what CTV contributes to user acquisition. As these components mature, teams will gain clearer attribution and a more dependable view of CTV’s role across the entire funnel.




