Startups and venture capital
DraftPrivate investment in space grew substantially through the 2010s and early 2020s. The results have been mixed in an instructive way: the failures share patterns, and the patterns are structural rather than accidental.
Why space is difficult for venture capital
Venture funds have a specific shape. They raise a fund with a life of about ten years, expect most investments to fail, and rely on a small number of very large returns. That model fits software, where capital requirements are low, iteration is fast and marginal cost is near zero.
Space inverts nearly all of it.
| Venture assumption | Space reality |
|---|---|
| Low capital to first product | Building and launching hardware costs tens to hundreds of millions |
| Fast iteration | A design cycle is years, and the test article is often destroyed |
| Fail cheaply and learn | A launch failure or an on-orbit failure costs the whole article |
| Near-zero marginal cost | Every satellite costs real money to build and launch |
| Large addressable market | Many space markets are genuinely small, and some are one customer |
| Exit within fund life | Time from founding to meaningful revenue frequently exceeds ten years |
The last row is the structural mismatch. A fund raised in year zero needs liquidity by roughly year ten. A launch company founded in year zero may fly in year seven and reach sustainable revenue in year twelve. The timelines do not align, which forces companies toward public markets or acquisition earlier than their maturity warrants.
The SPAC episode
Between 2020 and 2022 a large number of space companies went public through special purpose acquisition companies rather than conventional initial public offerings.
The mechanism mattered. A SPAC merger permitted the company to publish forward-looking revenue projections, which a traditional IPO prospectus effectively does not. Pre-revenue companies therefore went public on the strength of projections extending five or more years out.
The outcome was consistent and predictable in hindsight:
- Projections were missed by wide margins
- Revenue forecasts made at merger were commonly off by factors of several, not percentages.
- Valuations fell sharply
- Many companies traded far below their merger valuation within two years.
- Some ran out of money
- Companies that went public specifically to raise capital found the public markets closed to follow-on raises once they missed guidance.
- A few were fine
- Companies with real revenue and disciplined projections at merger have performed adequately. The mechanism was not the whole problem.
The transferable lesson is not that SPACs were uniquely bad but that a financing mechanism which rewards optimistic projection selects for optimistic projection. Companies that were honest about a ten-year path were valued below companies that were not, which is an adverse selection problem rather than a fraud problem.
Where capital went
| Segment | Capital intensity | Time to revenue | Assessment |
|---|---|---|---|
| Launch | Very high | 5 to 10 years | Heavily funded; most entrants have not reached sustainable operations |
| Earth observation | High | 3 to 7 years | Constellations built; revenue consistently below projection |
| Communications constellations | Extreme | 5 to 10 years | Only viable with enormous capital; several bankruptcies |
| Components and subsystems | Moderate | 2 to 4 years | Less glamorous, more consistently profitable |
| Ground segment and software | Low | 1 to 3 years | Best fit to the venture model |
| In-space servicing | High | Long and uncertain | Technology demonstrated, market not established |
The pattern is that the segments best suited to venture funding are the least capital intensive ones, which are also the ones that attract the least attention. Components, ground software and data analytics businesses have produced more consistent returns than launch or constellations, at correspondingly smaller scale.
Failure modes
The recurring patterns are worth stating explicitly, because they repeat across cohorts.
A market that does not exist yet. The clearest example is asteroid mining companies founded in the 2010s. The physical case was sound, the demand was not, and no revenue path existed within a fund’s life. See Asteroids and comets.
Underestimating qualification. Hardware that works in a laboratory is years and tens of millions of dollars from flight qualification. Founders from software backgrounds have repeatedly underestimated this by an order of magnitude.
A single customer. Many space businesses have exactly one plausible customer, which is a government agency. That is a contracting business, not a venture-scale one, and it should be financed accordingly.
Cost targets that assumed volume never reached. A satellite priced on the assumption of building five hundred is not viable if forty are built, and the volume assumption often depended on demand that did not materialise.
Competing with a vertically integrated incumbent. A launch startup competing on price against a provider with high cadence and captive demand faces the structural disadvantage described in The launch market.
Exits
| Path | Frequency | Character |
|---|---|---|
| Acquisition by a prime or larger operator | Most common | Consolidation of capability; often modest returns |
| Acquisition by a customer | Common | A constellation operator buying a supplier it depends on |
| Public offering | Rare, conventional route | Requires sustained revenue |
| SPAC | A 2020 to 2022 phenomenon | Largely unavailable now |
| Sustained private operation | Uncommon | Requires either profitability or a patient owner |
| Wind-down | Frequent | Assets sold, team dispersed |
Acquisition dominates, and it is often defensive: a large operator buying a supplier it cannot afford to lose, or a prime buying capability it failed to build. Those transactions clear the market but rarely produce the return multiples that venture funds require.
The realistic assessment
Private capital has genuinely changed the industry. Reusable launch, large constellations and commercial cargo and crew all resulted from private investment alongside government anchor customers, and none would have happened on agency budgets alone.
At the same time, the returns have not matched the narrative. The companies that succeeded had exceptional capital access, a government anchor customer through the development years, or both. The generic thesis that space is a large emerging market has not been validated by returns, and the most reliable businesses have been the least ambitious ones.
The useful conclusion for anyone assessing a space venture is to ask three questions: who is the customer today, not in the projection; how much capital is required before revenue; and does the business survive if the market grows more slowly than assumed.
References
- Bryce Tech, Start-Up Space: Update on Investment in Commercial Space Ventures, published annually.
- Weinzierl, M. and Sarang, M. “The Commercial Space Age Is Here”, Harvard Business Review, 2021.
- Klausner, M., Ohlrogge, M. and Ruan, E. “A Sober Look at SPACs”, Yale Journal on Regulation, 39:228, 2022.
- Space Capital, Space Investment Quarterly, for segment-level investment tracking.
- Company regulatory filings, which are the only reliable source for actual revenue against projection.