How I Grew This: Real Stories of Digital Growth

There are genuine practitioner insights on auction mechanics, publisher-side AI as a counter to DSP bid-shading, and the retention-vs-monetization tradeoff. However, roughly a third of the episode is origin-story throat-clearing and the '3x revenue growth' promised in the title is never actually explained or evidenced in the transcript.

“We realized that 1.3 seconds hit some of the best benchmarks”

“we are doing the opposite. We are essentially introducing the same concept and flipping it on its head and delivering only the auctions in which have a high return probability of best price”

The framing of a publisher building its own ad stack and then spinning it out is a mildly fresh narrative, and the 'flip DSP AI on its head for supply-side' angle is interesting, but most of the episode covers well-trodden ad tech territory—waterfall vs. auction, signal loss, UID 2.0—without a genuinely contrarian or first-principles argument.

“We purposely omitted our potential to buy. It absolutely hindered our ability to scale and prioritize revenue. But that's okay because I think long term it is one of the main catalysts of why Nimbus was able to really stand on its own two feet”

“the product kind of spoke for itself. We were able to transform monetization part of it, which was never really done at the time”

David Leviev is a genuine practitioner who built real ad tech out of necessity at a live consumer app and then commercialised it—not a career thought-leader. His domain expertise is evident throughout, though his company remains relatively niche and he occasionally retreats to vague forward-looking language rather than hard operational detail.

“Timehop, this independent app was able to build a fully established ad tech product”

“We piecemealed together what we anticipated, the success of what a monetization stack for Timehop can be”

A handful of concrete numbers appear—1.3-second visibility benchmark, 90-second average session duration, 30-second fixed interstitial rates, named third-party IDs—but the episode lacks hard revenue figures, publisher scale data, or the '3x revenue growth' substantiation its title promises, leaving many claims at the level of assertion.

“I remember Timehop having about an average session duration of about 90 seconds. And so whenever we introduced interstitial ads. We held the user for a minimum of what at that time, 30 second fixed rates”

“We realized that 1.3 seconds hit some of the best benchmarks”

The hosts occasionally land sharp, specific follow-ups—particularly the Reddit/scale question and the DSP integration query—but they also telegraph answers, call questions 'leading witness' questions out loud, and never press on the unsubstantiated headline claim of 3x revenue growth, leaving several important threads unchallenged.

“this is kind of a leading witness question”

“if I'm Joe's app and I don't have the volume or the understanding of the brand, if I say audio enthusiasts to the advertiser, how big do I have to be or how do I actually let them know that, like there's some actual performance data behind this?”