How I Grew This: Real Stories of Digital Growth

The episode contains solid product and business insights about on-device search, supply-demand alignment, and programmatic expansion, but is hampered by significant repetition (the core product explanation is rehashed multiple times) and conversational filler. The guest delivers useful frameworks on partnerships and company culture, but much of the 35 minutes involves clarifying what Discovery is rather than exploring novel strategic concepts.

“For an ad tech company to succeed, the supply and demand, which is where you get your inventory from and where you get ad order from, they need to be very closely managed. The product and engineering should also be very closely integrated with the go to market side.”

“we have eight projects live. We expect to fail on seven of them or six of them. We expect to because...we know that things fail.”

While the on-device bidding approach and horizontal integration thesis are genuinely differentiated, much of the framing leans on familiar adtech playbook (aligning supply/demand, data-driven targeting, experimentation culture). The comparisons to Google and Spotlight, while useful, are not particularly fresh. The guest retreads standard venture-friendly narratives about failing fast and building team culture without pushing into contrarian territory.

“It's not local yield optimization, it's global yield optimization. You've got to keep the population consistently happy with the things that they're seeing and expressing this happiness by clicking on them.”

“super apps part. It's only been successful in one large market ever. Right. It's only been successful in China. I...wouldn't bet on super apps in North America.”

Harish Thimmappa is a credible practitioner with 23+ years in adtech and genuine operating experience building revenue at scale (InMobi, Supersonic, now SVP at Branch overseeing a half-billion device install base). He has hands-on accountability for a complex supply-demand business with real P&L. However, he is not a founder or C-level executive at a household name, which moderately limits his stature compared to tier-1 operator interviews.

“I've been in adtech for over two decades now, almost 23, 24 years. Love it, love the technology aspect of it, love working with small companies, large companies, working scale businesses, build new revenue models.”

“I was a CRO at a, uh, small company called Unboxed...InMobi where I built their entire North American branch business team on the demand side, then supersonic ads, then Kenshu”

The episode includes concrete metrics (half a billion devices, 40-80 swipes per day, 0.7 clicks average, 10% ad load cap, 200-300 selection candidates, 7-8 ads bundled per update, 90%+ app advertising) and named OEM partners (Xiaomi, OnePlus, Oppo, Realme, Honor). However, there is minimal specificity on advertiser ROI, actual campaign performance numbers, or financial metrics. Top advertiser categories are mentioned but not company names (OpenAI and Gemini are exceptions). Revenue figures and growth rates are entirely absent.

“Half a billion devices around the world”

“a user swipes up on their Device anywhere between 40 to 80 times on average a day...every interaction typically gets on average leads to 0.7 clicks”

The hosts ask competent clarifying questions and show genuine curiosity (e.g., on-device bidding mechanics, signal prioritization, Apple prospects), but rarely push back or probe for deeper insights. Follow-ups tend to confirm rather than challenge, and the guest is permitted to repeat himself without redirection. There are moments of productive exploration (horizontal vs. vertical integration, AI impact on apps vs. ads), but overall the conversation lacks adversarial edge or willingness to test assumptions.

“And so how do you take that signal? So let's use your shocker example. Like everybody, every time they close Slack, they end up opening up a soccer score app. How does that turn into an advertisement?”

“Will you ever be able to get on Apple devices? Now, it's obviously up to Apple, but do you think.”