In 2015, Oracle hired the engineering department from the shuttered company Nebula, a private cloud company that promised a turnkey OpenStack solution for the most discerning of customers. I was one of the engineers in that department. (Oracle later acquired the company's IP as well, retroactively branding the move an acquisition within the company.)
What I found on the other side of that "acquisition" was a company in transition. Oracle had recently hired a soft-spoken Principal Engineer named Don Johnson away from AWS. It was Don who visited us at the Nebula office in Seattle to announce the offer, quietly laying out his vision to compete head-on with AWS—or at least to go work with a bunch of really smart people to build something really cool on Oracle's dime.
On one hand, I was excited at the vision of what could be. On the other, he had proactively countered any reservations I had about joining what I had then considered the Evil Empire. Yes, they are. But we're not. We're the change agents trying to right the ship. The hubris was palpable. And boy am I a sucker for moonshots.
What struck me about Don's pitch—whether to new employees, other executives within Oracle, or to prospective customers—was that it always started with the team. This was a group of battle-tested veterans who had built the first cloud that everyone knows and loves/hates—and they were doing it again, starting with a blank sheet of paper, carrying forward every lesson they had learned the hard way.
And that part of the pitch was solidly true. Don made some very solid early hires—people who were also suckers for moonshots. Five of these really smart, very early hires (among many others) were Clay Magouyrk, Marc Levy, Chris Newcombe, Pradeep Vincent, and Jag Brar. They convinced more great people to join. And those people pulled on their networks. And very soon our merry band of ~30 engineers grew to 100, 300, 600, 1000, 2000, and far, far beyond.
This was my school for nearly four years. It was filled with crazy ideas, crazy-smart people, crazy-hard execution. My job description changed dramatically during my time there. I built products. I wrote code. I designed systems and platforms. I onboarded and integrated acquired companies and their products. I managed people. I created new job families. I managed programs. I did dive-and-saves on broken initiatives.
But what amazes me about my experience there was that it wasn't unique. If you polled the first ~300 hires, you'd hear the same story: their jobs changed all the time. And that willingness to step into new and unfamiliar roles to get the work done—arguably before any of us were really ready—was ultimately our secret weapon.
That sense of ownership—of "If I don't do this now, we won't succeed" was all around us. It was the common thread that bound us all together in those early days. It was quality that you would immediately recognize the first time you met a new hire. And if you didn't see in it in them, chances are they wouldn't be with us long.
How did Don, Clay, Marc, Chris, Pradeep, and Jag create this culture? I have a theory—one which I continue to test every day.
1. Start with a trusted core.
Big things have small beginnings. The bigger the idea, the more fragile it is in the beginning. The trick to spinning up any moonshot initiative is finding those first few followers who are willing to go on a grand adventure with you.
They don't even have to believe the moonshot. But they do have to believe in you. This is why "a compelling vision" isn't the first item on the list. For the very first set of people, the vision doesn't matter. People aren't going to bet their careers on an unfounded, untested vision. But they'll bet it on a person they trust.
2. Cultivate a moonshot idea with a believable proximate objective.
Even with your trusted core established, you probably still don't have enough people to make the crazy idea real. So your number one job is finding people for the cause.
But which people? What qualities do they need to have? And what will convince them to come? You need a tool that attracts the right kinds of people, and aligns them to a singular, unambiguous cause. These are the moonshot, and the proximate objective.
These two things must go hand-in-hand. A moonshot idea with no believable first milestone—your believable proximate object—won't be taken seriously. It will be too nebulous. Too open to interpretation. The tactical priorities at any given moment will be unclear.
The moonshot acts as both a magnet and a filter. It attracts the dreamers, the people willing to take risks. It repels the people who want to simply go along for the ride.
The proximate objective plays the same role. It attracts the pragmatists—the ones who can translate a dream into next steps, rather than playing forever in the theoretical. It repels people who aren't prepared to get their hands dirty building something. And it ensures that the thing you're iterating on at any given moment has tangible, quantifiable value.
3. Hire only the best people you know, and hire them before you know what to do with them.
Getting the initial conditions right is key. The behavior of your first hires will define the culture of the group for the rest of its existence. If you hire great people, they will both attract similar people, and act as an insulator against mediocrity.
Get them in any way you can, however you can, even if you don't have an immediate use for them. When your initiative is in its infancy, you won't know exactly what you need. What you're actually doing here is building organizational capacity and capabilities—a sort of... store of potential energy that you can apply to different problems, depending on the makeup of your team.
4. Meet every problem with, "What can we do about it right now?"
As your team grows and they start making the moonshot real, they'll immediately hit obstacles. Maybe it's friction within a much larger organization. Maybe it's technical obstacles. Maybe it's that you don't have enough people to meet some critical deadline.
As a leader, your job isn't to have all the answers. It's to make sure each person is asking, "What can I do about it right now?" and then doing that thing. This question forces you to take personal responsibility, then act on it. At Oracle, we eventually enshrined this mindset in two of our core values: "own without ego" and "act now, then iterate."
5. Focus on outcomes.
The ultimate form of this is customer obsession—framing every major investment, every critical action, every hard decision in terms of the impact to the customer. This is the great unifier—the great tie-breaker when infighting occurs.
But it's sometimes hard to draw a line from "customer obsession" back to smaller decisions that are sometimes two, three, four steps removed from the customer. Focusing on intended outcomes—ideally in terms of customer impact—does the job of making sure a decision is being made with a specific result in mind.
Considering the outcomes—short- and long-term—also forces you to play the long game. It forces you to think through the long-term implications of your actions. It prevents "action for action's sake". It prevents you from building features with no clear customer or objective in mind. It forces you to compare proposals against alternatives in terms of opportunity costs.
6. Create room for productive failure.
Any bet caries an element of risk. The bigger the bet, the bigger the risk.
If you mitigate every possible risk, you might invest so much that no matter the payoff, you won't net out positively. If you choose to mitigate zero risks, then one wrong move could derail you forever.
The game then becomes a constant dance of finding bets, understanding the risks, then choosing which risks to mitigate and to what extent, weighing the costs of execution and risk mitigation against the payoff.
Learning to do this well is hard. Learning to do this at all levels of your organization requires you have systems which encourage this kind of thinking, and which self-correct when someone makes a preventable mistake.
Failure is inevitable. Your people need to feel comfortable knowing they might try something and fail. You need guardrails to prevent catastrophic failure, but not so many guardrails that execution is overburdened by process, or that personal accountability is stripped away. Your guardrails need to be learning tools that help people improve their judgement every time they make a mistake—a productive failure.
7. Reward the right behavior.
If you see someone consistently modeling the right behavior, make it clear that behavior is valued. There are many ways to do this, but ultimately your biggest lever is compensation.
Give a token amount when someone is meeting expectations. But when someone exceeds them, reward them disproportionately well. This sends a strong message, and raises the bar over time. You're doing the right thing. Do more of that. Encourage others to do more of that. Hire people who will also do that.
8. Have zero tolerance for bad behavior.
This may seem obvious, but in practice most people don't follow through. You need to quickly let go of people who don't exhibit the behavior you want modeled in the organization.
The obvious reason: that behavior will result in less-than-ideal outcomes from that person.
The less obvious but far more important reason: if that person stays, other people will begin to think they should act more like that person. The negative impact becomes viral, infecting any project, team, or person they've interacted with.
9. Emphasize the journey, and celebrate progress.
The beginning of a moonshot is the easy part. It's the middle that's hard. The middle is long. The middle is full of obstacles. It's painful. It's chaotic.
If the middle feels never-ending, eventually those great people you hired will start to leave. And if your team is doing well, the vision will always be out there—evolving as you evolve, becoming bigger as you become bigger, always pushing you to do more for your customer, always just out of reach.
So while you sell the vision to attract people, to keep people you need to acknowledge and celebrate the journey and the progress you've made.
Consistently stepping back to celebrate progress creates a sense of community—of shared struggles leading to shared accomplishment. It makes people think, If we could do that, what else could we do?
Repeat steps 1-10 at every level of your growing org. The same mechanics you used to get your moonshot off the ground can be used for each phase of the moonshot, each initiative in each phase, each team building systems for each initiative. It's turtles all the way down.
It can also be used for entirely new moonshots. Don't think just because you built a fantastic organization for Moonshot A that they'll be the right people, have the right temperament, or have the same interests needed to deliver Moonshot B. You need to start at square one.
Big, previously-successful companies get this wrong all the time. They think you can "innovate by mandate"—just stick fifty random people in the corner and let them come up with something great. It doesn't work like that. You must start with a much smaller seed—1, 2, or 3 people, slowly nurturing a thing, adding weight, momentum, and success until it becomes self-sustaining.
You can start living these steps yourself, even if you don't have a team, or a budget, or support in your company or org. Hold yourself accountable to these steps first. Model the behavior you want from your trusted core, bootstrapping this entire thing.
You don't need anyone's permission to follow these steps. You merely need the courage to start.