Stas Tushinskiy on AI-Driven Contextual Advertising: Transforming Brand Engagement

What led to Instreamatic’s focus on AI-driven contextual video and audio ads, and how does this approach differ from more traditional advertising methods?

Our focus on AI-driven contextual advertising (which we began offering before the industry’s broader AI boom of the past couple of years) came from recognizing 1) a fundamental shift in consumer expectations for more contextualized campaigns, and 2) a disconnect in how brands and agencies were able to deliver on those expectations.

Back when we first started Instreamatic as a programmatic audio ad platform in 2015, we saw that brands were struggling with the traditional ‘one-size-fits-all’ approach to campaigns. But marketers that tried to create more individualized ad variations—manually—were spending enormous resources doing so. And even then, they were never going to achieve true personalization at scale. It would take weeks of production time and significant budget just to create a few variations of an ad. 

Our AI-driven approach fundamentally transforms this process by enabling brands to generate hundreds of personalized video, audio, or CTV ad variations from a single creative within minutes. What really sets our approach apart is that we’re not just creating random variations—we’re using AI to analyze context and data to ensure each variation resonates with its intended audience. The numbers support the strategy: our work this year with one of the largest technology companies in the world showed that even simple contextual relevance in ads drove an 18 percentage point increase in purchase intent compared to standard ads. We’re essentially enabling brands to have meaningful conversations with their audiences at scale, rather than broadcasting the same message to everyone. This shift from mass messaging to personalized communication represents the future of advertising, and we’re proud to be at the forefront of this transformation.

How can marketers use AI to enhance storytelling in video and audio ads, making them more impactful and relevant?

AI is revolutionizing storytelling by enabling dynamic narrative adaptation in real-time. Rather than creating one linear story, we can now craft flexible narratives with multiple elements that can be personalized while maintaining the core brand message. Our platform uses AI to analyze what storytelling elements resonate most with different audience segments—whether it’s adjusting the pacing, music, voice-over style, or even visual sequences in video ads. For instance, we might find that morning commuters respond better to energetic, quick-paced narratives, while evening audiences engage more with relaxed, thoughtful approaches.

The key is that we’re not just making superficial changes or changes for the sake of changes; rather, AI helps us identify deeper patterns in how different audiences connect with storytelling elements. This allows brands to maintain their authentic voice while adapting how they tell their story based on context. What’s particularly exciting is how AI can help brands extend the lifespan of their creative assets by continuously finding new ways to make existing content relevant to different audiences.

What challenges do brands typically face in ad personalization, and how does Instreamatic address these?

The biggest challenge brands face with personalization is scale. Most brands understand the value of personalization. But traditional methods make it prohibitively expensive and time-consuming to create enough variations to truly personalize at scale. A typical video ad campaign might need hundreds of variations to account for different locations, times of day, audience segments, and other contextual factors. Creating these manually could take months and cost hundreds of thousands of dollars. Our platform addresses this by automating the personalization process, allowing brands to generate these variations in minutes rather than months.

We’re also solving the quality control challenge—our solution ensures that each variation maintains brand consistency while still optimizing for relevance. It’s not just about making more versions; it’s about making the right versions for each context.

What key metrics are used to measure the success of personalized ad campaigns for the industry?

We look beyond traditional metrics to understand the full impact of personalization. As we detailed in our recent report “2025: AI in Creative Production,” the industry needs to expand its measurement framework to capture the nuanced effects of contextual relevance. Our platform analyzes the performance delta between personalized and non-personalized versions, measuring not just if an ad worked, but why it worked. For example, in this campaign case study, we saw a 22 percentage point increase in brand favorability with personalized ads. But more importantly, we could attribute which personalization elements drove that improvement. Even seemingly minimal personalization can deliver impressive results, prompting customers to lean into a more tailored experience. As additional contextual elements are then layered into ads—like time of day, location, or audience behavior—the outcomes only become more impactful. 

We also measure adaptation efficiency—how quickly our AI can optimize ad variations based on real-time performance data. This helps brands understand not just the end result, but the ongoing optimization process that got them there. The findings in that new report linked to above reinforce what we’ve seen across campaigns: successful personalization requires looking at both immediate performance metrics and longer-term brand impact indicators.

Can you explain how your platform optimizes the delivery and performance of contextual ads at scale?

Our platform’s optimization process works on three levels, simultaneously. First, we use AI to analyze the initial creative assets and identify elements that can be modified without compromising brand integrity. This could include everything from background music to visual transitions in video ads. Second, our system creates variations based on contextual parameters—time, location, user behavior, and platform-specific requirements.

But what makes our approach especially unique is the third level: real-time performance optimization. As these variations are deployed, our AI continuously monitors performance data to refine the personalization rules. For example, if we notice certain elements performing better in specific contexts, the solution automatically adjusts the distribution to favor those combinations. This creates a virtuous cycle where each ad served helps improve the performance of future ads. We’re essentially building a self-improving system that gets smarter with every interaction, ensuring brands can maintain peak performance at scale. Our platform also integrates with existing DSPs and SSPs, so marketers can incorporate our enhanced personalization into current workflows without additional costs or infrastructure changes.

">