$3.5M a Year on Three Moonshots: The Stage Gate That Saves It

Every founder has a list of bold ideas. The question is which ones deserve capital and which ones are expensive distractions dressed as strategy.

What It Actually Costs

Unvetted moonshots are the single biggest drag on mid-stage growth. A company with $30M annual recurring revenue (ARR) that commits 15 percent of engineering capacity to three speculative bets with no stage gates is spending roughly $3.5M per year on projects that may never produce revenue. Two of those three will typically be dead inside 18 months. The capital isn't the worst cost. The opportunity cost of the engineering hours is.

Companies that vet moonshots poorly don't fail from one catastrophic bet. They fail from accumulated small ones. Each project survives a board meeting, consumes a quarter of roadmap, gets quietly shelved, and repeats. The aggregate drain is visible only in retrospect, when three years of gross margin compression traces back to underpriced speculative work.

Google's X division is often cited for its moonshot ambition, but the real mechanism is discipline. Projects like Waymo advance only after meeting pre-set milestones for technical viability and commercial scalability. The appetite for bold ideas is matched by a willingness to kill projects that miss the signal. That combination is what you want to copy.

The Approach

Step 1: Define success in advance with numbers

Before committing resources, write down what success looks like. Specific number, specific date, specific customer segment. "A new revenue stream" isn't success criteria. "$2M ARR from mid-market manufacturing customers within 18 months" is. If the team can't articulate this without hedging, the idea isn't ready for capital.

Step 2: Validate market need before building

Tesla's early EV traction came from identifying a specific gap in the luxury sustainable vehicle market, not from general interest in electric cars. For each moonshot, run 15 to 20 structured customer interviews with the target segment before writing code. Test whether the problem is painful enough that someone is already spending to solve it imperfectly.

Step 3: Build an MVP at under 5 percent of full build cost

Airbnb's first version was a site renting air mattresses. The goal wasn't to prove the full model. It was to prove that strangers would pay to stay in strangers' homes. Constrain the MVP to the single most uncertain assumption. If the MVP costs 30 percent of full build, you've over-scoped it and you'll commit to the project emotionally before the evidence justifies it.

Step 4: Pressure-test with a cross-functional review

Amazon's Alexa work pulled input from engineering, marketing, and design to stress-test the concept early. Run a structured review with three questions: is this feasible to build, is this desirable to the target customer, is this profitable at scale. Separate this input phase from the go decision so cross-functional feedback sharpens the plan without diffusing accountability.

Step 5: Secure executive sponsorship with data and narrative

Apple's iPhone advanced because Jobs championed it with both data and narrative. Leadership sponsors are the reason moonshots survive the first difficult quarter. Present the vetted concept with the market evidence, the MVP results, and the specific capital ask tied to a defined stage gate. Vague asks lose funding first when the budget tightens.

Step 6: Run cost-versus-impact analysis with scenario ranges

SpaceX pursued reusable rockets because the cost curve made the economics work at scale. For each moonshot, model three scenarios: base case, upside, downside. The base case shouldn't depend on the upside to justify the investment. If it does, the bet is priced on hope.

Step 7: Install stage gates with kill criteria

Microsoft limits further investment until early milestones prove out. Your stage gates should include explicit kill criteria: what evidence would cause this project to stop. A gate without kill criteria is a ritual. Each gate should force a real decision: double down, pivot, or stop.

Step 8: Capture the lesson, even on failures

Google Glass didn't become a commercial product, but the underlying technology informed Google's AR roadmap for a decade. Moonshots that fail aren't wasted if the organization extracts the learning. Require a one-page retrospective for every shelved project, circulated to the strategy team and the next round of proposers.

Where This Breaks

The common failure is treating moonshot vetting as a one-time approval rather than a running discipline. Ideas get funded at quarterly off-sites with enthusiasm and no follow-through on gate criteria. Six months later the project is still running, the original hypothesis is forgotten, and the team is building toward a drifting target.

The second failure is asymmetric accountability. Executives fund the bet, a team executes it, and when the evidence turns negative no one has the authority or the incentive to recommend killing it. Build the kill authority into the same group that approved the funding, and require a kill-or-continue vote at each gate.

What would change about your next three funding decisions if every moonshot had to pass a real kill gate at 90 days?

Priorities for the Quarter

• Document success criteria for every active moonshot with number, date, and target customer
• Run 15 customer interviews per idea before any engineering commitment
• Audit your current moonshot portfolio against actual stage gate evidence, not intent
• Install explicit kill criteria for each gate and assign the kill-or-continue decision to a named owner
• Require a one-page retrospective for any project you shelve or stop

If your team is wrestling with which moonshots to fund and which to kill, take the free assessment to identify the gaps draining growth capital quietly.