How mParticle Hit a $36M Run Rate on 150 Customers — and Sold to Rokt for a Reported $300M

In November 2018, Nathan asked Michael Katz to explain what happens when someone orders a grande mocha through the Starbucks app. Katz’s answer was the whole company in miniature: from the moment you open the app to payment confirmation, the order throws off a trail of data that Starbucks needs in ten to twenty different systems — the warehouse, analytics, attribution, marketing automation, Facebook, Google. mParticle’s pitch was to capture that data once and pipe it everywhere, instead of maintaining twenty brittle point integrations. At taping, that pitch was earning roughly $3 million a month from just 150 customers.

Katz had done this before. He took ad-tech company Interclick public in 2009, sold it to Yahoo in 2011, stayed “roughly about a year,” then left in early 2013 to start mParticle with his brother as co-founder and CTO — the same role he held at Interclick — plus COO Dave Myers and their ten best engineers from the old team. When Nathan joked that Yahoo must love him for the raid, Katz didn’t flinch: “Yahoo doesn’t exist anymore.”

Priced on people, not data points

The licensing decision Katz described is the quiet engine of everything that follows. mParticle charged per user record, not per event: one person tracked across app, web, and an in-store kiosk still counts once. Enterprise contracts ran six and seven figures annually, growth accounts started at five figures, and the average customer landed “closer to like 20 to 25” thousand dollars a month. Because the meter was people, the bill grew exactly when the customer’s business grew — Katz said the pricing was designed “to make sure that that was aligned with our customers and how they think about their business.”

Where 150% net retention actually came from

Nathan told Katz on tape that across roughly 3,000 B2B CEO interviews, he could count on one hand the companies claiming 150% net revenue retention. Katz’s number held up under the follow-ups, and he broke the machine into three parts:

• Same-store growth — when a brand attracts more users, the user-record meter runs by itself; mParticle grows because the customer does.
• New properties — accounts that started on the mobile app added desktop web, mobile web, then offline data sources, each expanding the record count.
• Applications on the pipeline — segmentation, enrichment, and transformation features built on top of the data flow, sold as upsells against the install base.

The churn side of the ledger was just as unusual. Logo retention ran 95–97% per quarter, concentrated in smaller growth accounts — “we actually don’t see any churn in the enterprise,” Katz said, putting enterprise gross revenue churn under 5% a year. Revenue mix at taping was 60% new logos to 40% upsells, expected to invert toward 70/30 the following year.

We start to serve as like the main artery. To rip us out is really, really difficult — and something that really nobody even wants to do.

Michael Katz, co-founder & CEO, mParticle

That stickiness had a price on the front end: sales cycles of one to three months for “hair on fire” buyers, six to nine for classic enterprise procurement. Fully weighted, Katz was spending $400–500K to land a $20K-a-month account — a payback period around 24 months he wanted down to 18. Against the CAC benchmarks founders on the show disclose, that’s heavy — and rational only because the revenue essentially never leaves.

The funding ladder

By the taping, mParticle had raised $75 million, all equity — a seed led by Bowery Capital with Google Ventures and Greylock participating, a Series A from Social Capital, a B from Bain Capital, and a C led by Harmony Partners. Katz was already “thinking about our next fundraise” on tape; the dataset records one more round after it.

• Oct 2013 $3M — the seed year, ten engineers rehired from Interclick.
• 2014–2015 $1.5M and $4.3M follow-ons as the platform found its enterprise lane.
• 2016 $15M in January, $17.5M in October.
• Sep 2017 $35M — the round that carried the company to the $36M run rate Nathan taped.
• Mar 2020 $45M, taking total funding to $121.3M.

What the dataset shows after the tape

The record-keeping after 2018 tells an honest, unglamorous middle chapter: growth to $54 million by the end of 2021, an estimated push toward $69 million in 2022, an estimated slip back to $62.8 million in 2023 as marketing budgets tightened, then a confirmed $76 million in late 2024. Headcount followed the same curve — 110 at taping, 271 at the end of 2021, and roughly 251 from 2022 onward.

The Rokt ending

In January 2025, the story resolved the way infrastructure stories often do: consolidation. Rokt — the transaction-moment advertising company and itself a Latka interview subject — acquired mParticle in a deal AdExchanger reported at $300 million, folding the customer-data pipeline into Rokt’s ecommerce network. For a company that had raised $121.3 million, it was a workmanlike exit rather than a triumphant one — but it put Katz’s “main artery” under an owner whose entire business runs on real-time customer data. mParticle’s current numbers are on its GetLatka profile.

Katz, asked at 40 what he wished his 20-year-old self knew, gave the answer of a founder on his second exit ramp: “It’s all going to work out. You put in the work, you don’t get too high or too low when things are going well or not so well, and you just put one foot in front of the next.”

SourcesNathan’s November 2018 interview with Michael Katz; GetLatka dataset rows through November 2024 (2022 and 2023 revenue are estimates); Rokt acquisition per AdExchanger, January 2025.