Affle Launches MediasmartTV to Sell CTV Ads on Outcomes, Not Impressions

Affle claims its new self-serve CTV platform reaches more than 1 trillion monthly ad opportunities in the US alone. On Monday, the Singapore-based, India-listed ad tech holding company launched MediasmartTV, housed inside Mediasmart, the omnichannel DSP (demand-side platform, software advertisers use to buy ads programmatically) it acquired in 2020.

"Every dollar an advertiser spends should drive some kind of outcome, including on connected TV," said Sameer Sondhi, Affle's North American CEO and chief strategic investments officer, who joined the company earlier this year after serving as CRO and CEO of ad tech at Verve Group.

How does the pricing model work?

The platform's core pitch is Affle's cost-per-converted-user model, or CPCU. Advertisers pay only when a viewer exposed to an ad on their TV then takes an action, such as installing an app or making a purchase, rather than buying impressions upfront.

The mechanics depend on linking TV exposure to later behavior. "We don't have a cookie and we don't have device IDs from the TV side," Sondhi said. Mediasmart instead ties ads to households using IP addresses, devices on the same Wi-Fi network and, in Sondhi's words, "everything consented that we're allowed to." When a conversion happens on another device within an agreed window, the advertiser's ecommerce or measurement platform sends Mediasmart a postback (a server-to-server confirmation signal), and Affle bills on that. Purchases are tracked through integrations with WooCommerce and Adobe Commerce.

Supply comes from SSP partners including Magnite, PubMatic, Index Exchange and FreeWheel, plus TV manufacturers LG and Samsung. Affle's optimization algorithm was built for mobile, where device identifiers are plentiful. Its "TV-first" version has been retooled to run without them, learning from Mediasmart's mobile campaigns, advertiser first-party data and signals pulled from the TVs themselves. The platform lacks direct publisher integrations for now, so contextual signals are limited, and Affle is evaluating ACR (automatic content recognition, TV-based detection of what a screen is showing) and ID graph partners to close the gap.

Who is buying it?

Affle is starting with the buyers who need least convincing: app marketers, mainly in gaming, that already run user acquisition and reengagement campaigns with the company. "You'd be surprised how much gaming user acquisition we're already doing through CTV," Sondhi said. Interest extends to rideshare, finance and reward apps, some spending millions on TV.

Next come independent agencies, courted through "lunch and learn" sessions with Affle's mobile measurement partners, then the large holding companies. Sondhi expects the holdcos to resist, given what he calls a "very upper funnel" mentality. "I really do hate the upper-funnel, mid-funnel, lower-funnel pyramid," he said. "Let's be honest, it's not like advertisers spend money without expecting to get something back."

Affle is marketing the model with the slogan "TV to cart" and is seeking trademarks on it, plus "TV to travel," "TV to education" and "TV to finance." "We want to keep it simple," Sondhi said. "No more acronyms, no technical jargon."

What does this mean for the US push?

North America accounts for roughly a third of Affle's business, but the company has been underrepresented there, Sondhi said, and "that's where the big dollars are." Founded in 2005, Affle trades on India's stock exchange with a market cap above $2 billion. Mediasmart is one of eight specialized point solutions assembled since 2012, alongside Vizury, Appnext, Jampp and YouAppi. In June, Affle bought AdColony's brand and SDKs from DT, formerly Digital Turbine — the technology only, not the team.

Sondhi spends more than half his time on M&A and said to expect more deals soon. He declined to name targets but said any acquisition will tie back to CPCU, with CTV an obvious interest. Further out, Affle is eyeing "autonomous integrated connected devices," from watches to retail-store screens, because "these industries are evolving so fast, we don't want to just bet on one thing," Sondhi said — a plan that may test the "no more acronyms" pledge.