Tatari's Philip Inghelbrecht: from founding Shazam in 1999 to building a $100M+ profitable TV ad tech company

Tatari's Philip Inghelbrecht: from Shazam to $100M+ in TV ad tech

Philip Inghelbrecht co-founded Shazam in December 1999, built a profitable TV advertising platform called Tatari, and did most of it without venture backing. That combination — bootstrapped, profitable, and operating in a sector Silicon Valley dismissed — is the throughline of his career.

Shazam: the company that almost didn't exist

Shazam launched its first consumer product in August 2002, years before the infrastructure existed to support it. The original experience required dialing a four-digit short code (2580, the center column of any keypad), holding the phone up to the music, and waiting for a reverse-SMS charge to deliver the song name and artist. It was functional. It was also nearly unusable.

Inghelbrecht describes it as a convergence of four individually improbable bets: building the largest digital music database in existence at the time, inventing the music recognition algorithm from scratch, assembling a computer cluster before AWS or Google Cloud existed, and signing every major mobile operator. He puts the compound probability of Shazam surviving at roughly 0.001%.

The business didn't flatline on the consumer side by accident. The iPhone didn't exist. The iTunes Store didn't exist. Distribution was nonexistent. What kept the company alive was a pivot to B2B licensing — selling music recognition to royalty-tracking organizations like BMI and ASCAP, who needed to move from pen-and-paper sampling of radio play to a census-level system across roughly 2,000 US radio stations. Shazam would tap the radio signal directly through leased equipment in each market, identify every song aired, and feed that data back for royalty attribution. Multi-million dollar licensing contracts from that work funded the consumer side while it waited for the market to catch up.

When the iPhone launched in 2007, Shazam was ready. Apple needed showcase apps; Shazam was one of very few that demonstrated something genuinely new. Apple eventually acquired Shazam to use it as a user-acquisition engine for Apple Music — recognizing a song and converting that moment into a subscription was cheaper than almost any other acquisition channel.

Inghelbrecht left Shazam operationally around 2004, joined Google as an early YouTube employee, then co-founded TrueCar in Los Angeles before a subsequent startup was acquired by Yahoo.

“We are technology for TV advertisers — managing creatives, planning campaigns, executing buys, measuring and optimizing across streaming and linear. We are 300 people strong, doing well over $100 million in net revenue, and we've been profitable from day number one and mostly self funded. We pretty much doubled our revenue with the same amount of people with AI tools.”

Tatari: built from a bad experience

Tatari came directly out of watching how TV advertising actually worked at TrueCar — and finding it inadequate. Inghelbrecht founded the company in 2016 with the thesis that TV measurement was broken, and that fixing it would unlock a better buying process. The company now has 300 employees, generates over $100 million in net revenue (excluding media spend, which would be a much larger figure), has been profitable from day one, and is largely self-funded.

The self-funded structure was partly strategic, partly a product of timing. Sand Hill Road wasn't interested in a TV ad tech company during the streaming transition, which meant Inghelbrecht didn't face the valuation pressure or competition that VC backing typically attracts.

Measurement first, then buying

Traditional TV measurement ran through Nielsen ratings — reach and audience size. Newer brands arriving with digital backgrounds wanted outcome measurement: did the campaign drive app installs, signups, or customer lifetime value? Tatari built its own measurement stack to answer those questions, layering deterministic and probabilistic approaches across multiple data sets. Inghelbrecht describes the system as continuously iterated, similar in structure to how Google updates its search algorithm.

The commercial model scales with the brand. Clients typically start with $50,000–$100,000 in TV spend, and Tatari takes them up the curve. Last year, the company placed four or five brands in the Super Bowl at roughly $15 million per slot.

On the buying side, Inghelbrecht is skeptical of programmatic TV. The supply of ad inventory is highly concentrated — roughly 90% of impressions come from the top 10 publishers (Disney, Peacock, and a handful of others). Applying programmatic logic built for the fragmented digital display market to that structure doesn't work well for either side. Tatari operates through direct integrations instead.

AI doubling output without adding headcount

Tatari migrated its data infrastructure from Redshift to Databricks several years ago, primarily to handle growth. That migration, timed before the large language model wave, left the company positioned to use AI against its full historical data set immediately.

In practice, that means campaign planning across roughly 40,000 linear network rotation entities and 10,000 streaming opportunities can now be done in seconds rather than by human buyers working through a fraction of the options. Inghelbrecht says Tatari approximately doubled revenue with the same headcount after deploying these tools, and is now considering a four-day work week as a result. The next target is applying AI to media execution — using model inference to select the best impressions rather than running hundreds of thousands of programmatic auctions per second.

The broader TV market

US TV advertising sits at roughly $90 billion annually, holding steady where print and radio have declined. The internal shift from linear cable and broadcast into streaming is the significant structural change. Average American TV consumption runs at three and a half hours per day and growing, compared to roughly 30 minutes on Instagram — a gap Inghelbrecht frames as the core reason TV remains the most important reach vehicle for advertisers.

Inghelbrecht sees influencer media and TV converging. Where a TV campaign once launched with a single 30-second creative, brands now test 10 creators simultaneously and find the winner. Influencer content production — creating 100 videos to find the one that works — maps directly onto how TV creative could be managed with cheaper generative tools. He describes that convergence as a near-term growth opportunity for Tatari, and says the company is currently tracking ahead of a plan to more than double revenue within two and a half years.