How Colin Zima Lost 5 Deals Then Raised $250M for Omni

Colin Zima thought Omni would have traction in a month. It took nine. After losing all five of his first committed deals, he spent two months killing bugs, went on one podcast, and won every trial that followed. Omni has since raised over $250M.

Key Facts

• Time to first real traction: 9 months from founding to first meaningful user (Colin Zima)
• PMF signal: Won all 4 trials from one podcast appearance after losing all 5 earlier verbal commitments (Colin Zima)
• Looker acquisition price: $2.7 billion by Google (Colin Zima)
• Founding capital: Raised $30M across seed and Series A within 6 months of founding (Colin Zima)
• LinkedIn outreach response rate: ~90% response rate to personal connections; 3% for BDR cold outreach (Colin Zima)

Losing All Five Deals—And Why It Wasn't the End

When Omni lost all five verbally committed customers, Colin Zima didn't pivot—he fixed bugs. The product had real differentiation, but too many rough edges overwhelmed the positives. Two months of polish changed the outcome completely: they won every trial that came next.

After nearly a year of building and 100 demos, Omni finally had five unaffiliated prospects at the verbal commit stage. Colin believed they were in paperwork. Then all five walked away. The feedback: company too young, product not quite there.

Most founders in that situation question the idea itself. Colin questioned the execution. He knew the product was genuinely better because he used it every day—he could see through the bugs to what it was capable of. The problem wasn't the concept; it was the friction.

The team spent two months eliminating every bug they could find. Then Colin went on a podcast with Sam Blond, generated three or four trials from completely unaffiliated listeners, and won all of them. That was the PMF moment: strangers, paying full price, actually liking the product.

"We were using the product every single day and we knew it was better. We knew it was buggy, but we also knew it was better. We could see through hitting a full page error and hitting the refresh button because everything else underneath it was so awesome." — Colin Zima

"When you build product, you need a balance of things that differentiate you and you need to avoid things that differentiate you in a negative way. We had too many of those negatives that were overwhelming slightly the positives." — Colin Zima

Why the Innovator's Dilemma Created Space for Omni

Omni exists because Looker couldn't reinvent itself. Once a product is built around a core thesis—like enterprise BI with rigid data modeling—it's nearly impossible to compromise that foundation to serve adjacent needs. That structural rigidity is exactly where new entrants win.

Colin spent eight years at Looker watching it become exceptional at enterprise BI and inflexible at everything else. When Google acquired the company for $2.7B, the product stopped evolving at the pace the market needed. That stagnation gave Colin conviction to leave and compete directly.

The key technical insight Omni built around was an inversion of the Looker model. Looker required you to build the semantic data model before you could ask questions. Omni lets you ask questions first, then formalize the model afterward. As Colin described it: 'Do you build the foundation of the house before you build the house, or do you just build the house and figure out the foundation after?'

This matters even more in the AI era. AI needs to explore freely when answering questions—you can't constrain it to a rigid model. But once it explores, you need to lock in consistent metrics and permissions. Omni's architecture made that possible in a way Looker's couldn't.

"It's really hard to take your core foundational thesis and compromise it at the margin to do these other things well. And so it leaves space for other people on the other side." — Colin Zima

"Teams get tired. If you have a team that does something for ten years, you kind of do need to reshape it a little bit and inject different energy." — Colin Zima

The LinkedIn Social-Selling Playbook That Built Omni's First Pipeline

Colin systematically messaged his 6,000 LinkedIn connections—including nearly 900 former Looker employees—and achieved up to a 90% response rate. The formula: short messages, direct problem framing, and genuine credibility in the space. No automation, no templates. Just relevant outreach at scale.

Most founders think of LinkedIn as a passive brand channel. Colin treated it as an outbound sales tool with an unusually warm database. He printed out his connection list, went through it manually, and flagged everyone who might have a data problem his product could solve.

The BDR team running cold outreach got about 3% response rates on LinkedIn. Colin got 90% from people who knew him—because trust compresses every part of the sales cycle. When someone has worked with you or bought from you before, 'would you look at this?' is a simple ask.

The messaging principles that worked: shorter is better, more direct is better, no flowery language. State the problem you think they have, say you solved it, ask if they want to see. The worst outcome is no reply.

"To real LinkedIn connections, I probably had a ninety percent response rate and I had five thousand, six thousand LinkedIn connections. If we find one thousand good people in there, there's nine hundred responses." — Colin Zima

"Good selling is just connecting someone with a problem to someone that can solve their problem. If I'm saying I bet that I can either reduce your cost or make you guys smarter and give you a better product, that's a pretty easy message." — Colin Zima

• Print your connection list and manually identify good-fit prospects—it takes ~2 hours for 200 people.
• Message former colleagues and customers first—trust dramatically increases response rates.
• Frame outreach around their problem, not your product features.
• Avoid asking for 'feedback'—it attracts tire-kickers. Ask for a demo instead.
• SalesNav is the only paid tool you need early; skip LinkedIn ads until later.

Hiring From Your Industry: The Rolodex Advantage

Omni's first sellers were all former Looker reps. When you hire someone who already sold into your target market, they arrive with relationships, credibility, and a book of accounts that will eventually need to switch tools. That's not just a hiring shortcut—it's a pipeline strategy.

Colin's team of twenty pre-revenue employees had an average of ten years of shared history. Fourteen came from Looker, five from Stitch. This eliminated most management overhead and let the team move fast without process. When everyone already knows how each other works, coordination costs collapse.

The same logic applies to go-to-market. Omni's third hire now runs their European business out of Dublin. He was a strong seller at Looker—Colin trusted him, brought him over, and let him build the region. A great enterprise seller doesn't just bring skill; they bring a trusted relationship with every account they've ever worked.

The caution: big-company employees don't always translate. Colin specifically looked for people who could operate without frameworks, without performance management, without mentorship. The culture fit question at Omni was simple: 'You will not get mentored. You'll come here and do stuff, and we want to ship code every day. If you like that, you've self-selected into a good setup.'

"A great enterprise seller is almost like a corporate advisor to their accounts. You want them to come with a bag of business that you can go address." — Colin Zima

"If you worked at a big company with a thousand people and you hire the ten best people, you're going to hire the next ten best people also." — Colin Zima

No Silver Bullet: How Omni Actually Scaled Past PMF

Omni's growth came from doing every obvious channel simultaneously and executing each one better than average. Outbound, inbound, partnerships, events—none was the magic lever. The insight: founders hoping for a single breakout channel are avoiding the unglamorous work of doing everything well.

Colin's honest assessment of go-to-market: 'The least satisfying but most obvious conclusion you could possibly imagine is you just need to do more of literally everything.' Outbound via LinkedIn and BDR. Inbound through content and podcast appearances. Partnerships with both hyperscalers like Snowflake and small one-person data consultancies. Events turned on later.

The partnerships insight is worth unpacking. Large tech partners—Snowflake, Databricks, Google—won't move business to a ten-person startup regardless of your history. But small data practitioners who serve as fractional data teams for growing companies will. They get more at-bats, they're less locked into incumbents, and if your product works, they'll sell it for you. Omni paid $5k implementation fees to partners who then built familiarity with the product and recommended it to future clients.

The compounding effect: every channel reinforces the others. A podcast generates trials. Trials become case studies. Case studies attract partners. Partners bring accounts. The flywheel isn't one thing—it's all the things, done consistently over time.

"I've heard only like one or two things here and there sporadically where I'm like, wow, that's really creative. Almost none of them figured something out. They just have a solid product, the product provides value, and then they do the obvious stuff." — Colin Zima

"Even though our team was maybe twenty people at the time, our engineering team was moving faster than the incumbents and that's an enormous advantage. People are buying you for the slope too." — Colin Zima

Looker vs. Omni: Core Architectural Difference