Atrium

Pete Kazanji's journey to founding Atrium began not in sales, but in product marketing at VMware. When he co-founded Talent Bin in 2009—a recruiting software company—he became the company's first sales rep out of necessity. His co-founder told him bluntly: "you're the only one who's job is to sell this." This forced education in sales, combined with his product marketing background, led him to instrument Talent Bin's sales operation obsessively. They tracked win rates, customer-facing meetings, demos, ASP, and deal cycles—essentially turning sales into a measurable factory.

When Monster Worldwide acquired Talent Bin in 2014, Pete scaled from running a 20-person sales team to overseeing new product sales for 1,000 salespeople. The contrast was jarring. Monster was an old-school organization struggling with data-driven sales management. Pete's team discovered something shocking: when they accessed Monster's exchange server to pull meeting data, they found only 60% of the expected selling activity was actually being logged. The sales leadership was defensive—"It's happening, it's happening"—but the data told a different story. It was like having a boat with 10 rowers all moving at 50% speed. Someone had to know.

After leaving Monster, Pete and his team spent the second half of 2016 and 2017 interviewing customers and validating their hypothesis. The pattern was universal: organizations knew which metrics mattered (bookings, win rates, ASP, pipeline, customer-facing meetings), but they couldn't operationalize them. Sales ops would spend weeks building dashboards in Salesforce, Tableau, or Looker. Then nobody would look at them. The visualizations became "a spray of spaghetti across the screen." Managers were too busy firefighting to consume analytics. Pete's insight: build software that does the statistical analysis for them, not tools for data nerds.

Atrium launched with a simple architecture: integrate with Salesforce for opportunity data, Gmail/Google Calendar or Office 365 for activity data, and automatically calculate the key metrics managers actually need. Within 90 seconds, a sales leader would have KPIs that normally took a sales ops person weeks to assemble. But the real innovation was the software's ability to do what humans hate doing—continuous statistical analysis and anomaly detection. Instead of asking managers to stare at dashboards, Atrium surfaced insights: "This rep has half the customer-facing meetings of her peers this month. That's a problem." Or: "This SDR's new opportunity creation is declining compared to his own baseline. What's going on?"

Pete began founder-led selling in 2018, acquiring roughly a dozen customers. This wasn't passive—it was discovering that the true buyer was the sales manager, not the sales ops person. Sales ops wanted flexible, customizable reporting tools. Sales managers wanted answers, not dashboards. The product pivoted from being positioned as a reporting tool to being positioned as a coaching and management tool.

In early 2019, Pete hired his first sales rep. The rep closed deals in March and April. Emboldened, Pete then made a critical mistake: instead of hiring one more rep to prove repeatability, he hired a new sales manager and two reps concurrently in April-May 2019. He jumped two steps at once on the staircase and tripped. The sales motion wasn't sufficiently packaged. It lived in Pete's brain and his first rep's muscle memory. The new manager had no foundation. The new reps didn't ramp smoothly. The existing rep stalled. Within three months, Pete reintervened, got back in the weeds, and by September-October the team was humming again—but he'd lost 3-4 months unnecessarily.

The biggest early mistake was abstracting away from day-to-day sales management too early. Pete's insight became the centerpiece of his later book *Founding Sales*: there are stages in scaling a sales organization, and attempting stage N+2 improvements before mastering stage N leads to disaster. The recipe works like this: first, prove you can repeatedly sell it as a founder. Then, prove someone else who isn't you can repeatedly sell it (the second rep is the "beta test"). Only after that cohort ramps successfully should you hire management.

What worked was Atrium's core hypothesis: non-technical sales managers will pay for software that tells them what's wrong and where to coach. The product's ability to surface anomalies—a top performer with zero new pipeline, an SDR's win rate deteriorating against baseline—resonated deeply. Sales leaders didn't want Tableau. They wanted a minivan, not a drag racer, to borrow Pete's later phrase.

Atrium targets tier two and tier three organizations—those with repeatable sales motions and 10-150+ sales reps. Tier one (pre-sales-ops) organizations don't have enough repeatability. Tier four (massive enterprises with existing Tableau/Looker infrastructure) have already solved the problem with human analysts. Atrium's sweet spot is the sales manager managing 5-20 reps who needs to know: who's struggling, why, and what to do about it.

By positioning himself as a founder-led seller and later author of *Founding Sales* (2021), Pete built credibility and attracted customers organically. He also co-founded Modern Sales Pros, a 15,000-member community for sales operations leaders. Atrium operates on a subscription model, integrating seamlessly with Salesforce and calendar tools to deliver a metrics harness that used to require weeks of manual data work.