John Dean co-founded WindBorne Systems, which uses their long-duration smart weather balloons and their AI platform to create the most accurate weather forecasts.

WindBorne has fresh news this week too: a new $37 million Series B, covered in TechCrunch. I've worked with John since 2017, when Ubiquity backed the company, and SBIRs have been part of the WindBorne story from the start. Right after the company closed its first funding round, it matched roughly the same amount in SBIR awards. Six years later, WindBorne has won more than ten SBIRs across NOAA and multiple parts of the military, and every single one has converted into a recurring revenue contract with that agency. That early revenue is part of what carried the company to this week's round.

In our latest Ubiquity University session, John explains how. Government grants are suddenly relevant to a much wider set of founders, especially in defense tech, and almost nobody explains how they actually work.

1. Pick grants the way you'd pick customers

John's central argument is that founders think about grants wrong. A venture-backed startup shouldn't chase an incremental $1.5 million as a one-time award. The only grants worth pursuing are the ones that become long-term customers, so skip any SBIR that will never turn into revenue.

The mechanics are simpler than they look. Phase 1 awards run about $100K to $250K over three to six months and are mostly exploratory. Phase 2 ranges from $750K to $1.5 million over six months to two years, with some agencies now going to $2 million. Phase 3 means the agency is buying your product as a service, which is the whole point. Applications take a month or two of part-time work, and notice of funding usually arrives about three months after submission.

John also suggests dropping ARR when you do government work. A $1.5 million Phase 2 over 18 months is $1 million in annual contract value, with lumpy payments along the way.

2. Sell it like enterprise sales

Agencies look like impenetrable walls, but they're still made of humans. Figure out who inside the agency actually cares about the problem you solve, who controls the budget, and why the solicitation exists at all. Every SBIR topic has a reason behind it, and reading the posting literally will often steer you away from what the agency really needs.

Then find your champion. Awards get decided behind closed doors in review committees, and you need someone in that room advocating for you. As John puts it, "If you're blindly submitting and no one in the agency has ever heard of your company, you're probably not going to get the award." Email the people running the program, ask questions, and do the homework before you walk through the front door.

3. Buy expertise, never outsource

There's a cottage industry of SBIR consulting firms that will run the process for a cut of the award. John found them unhelpful, and his comparison stuck with me: a startup CEO handing the SBIR process to a consultant is like outsourcing the sale to your first commercial customer. The consultants won't know your tech or your story, and the proposal suffers for it.

Targeted expertise is a different matter. An accountant who specializes in government work will save you real time on budget forms, and John estimates he lost 40 hours of his life to line items nobody cared about. Angel investors with government experience give sharp, specific advice. And if government will be a durable revenue line, hire an in-house proposal writer early. John calls it one of the best hires WindBorne made and wishes they'd done it sooner.

One insider detail worth knowing. Proposals can come back "selectable but not funded," meaning the agency liked your work but ran out of program money. Sometimes you can take that exact proposal to a different agency with excess budget and win the award there.

4. The milestones are yours to write

The standard objection to grants is that they chain your roadmap to someone else's priorities. John's answer is that this only happens when you approach it wrong. You propose the milestones and the payment schedule, and the agency approves them. Never ask your technical point of contact what the milestones should be.

WindBorne learned this the hard way. On early SBIRs, all the money arrived at the end of the grant, because that's how they wrote the schedule themselves. Six months of cash they could have banked earlier sat waiting at the finish line.

I'll add the investor perspective. When founders mention non-dilutive funding as free money they collected along the way, it rarely changes my view of the company. A founder who treats the same grant as the first step of an enterprise sale is telling me something real about how they'll build revenue.

Watch the full session on Ubiquity University

This is a preview. In the full video, John goes deeper on the tactics, from reading solicitations to structuring payment schedules, drawn from more than ten funded SBIRs.

Watch Here