The small satellite industry is bracing for a decade of explosive growth, with a new forecast from Novaspace predicting 70,000 satellites will be launched over the next ten years—a sevenfold increase from the previous decade.
Insights into the SmallSat bottlenecks came from the Downlink to Value: Emerging Markets for Smallsat Data & Connectivity panel at the SmallSat Symposium in Silicon Valley, CA organized by Satnews Events on Thursday, Feb. 12, 2026.
The panel was moderated by Nathan De Ruiter (Partner & Managing Director, Novaspace) and panelists included Ali Younis (VP of Sales & Business Development of Mynaric), John Malsbury (CEO & Founder, AnySignal), David Proulx (Chief Product Officer, SkyWatch) and Hirokazu Mori (Group Chief Strategy Office, US CEO, Warpspace).
Downlink bottleneck
This surge represents a $134 billion investment according to De Ruiter, but as experts warned during the “Downlink to Value” panel, the hardware is no longer the main challenge. The real battleground has shifted from launching mass to orbit, to extracting actionable value from the data once it gets there. “Data stuck in space is meaningless,” cautioned Ali Younis of Mynaric, underscoring the industry’s urgent need to solve the connectivity bottleneck.
The consensus is that the current downlink infrastructure is archaic. Younis compared the current state of satellite comms to having to manually schedule a specific cell tower just to make a phone call. The solution lies in automation and a hybrid approach. John Malsbury of AnySignal argued for a network where radios function like T-Mobile phones—”you walk out the store, [it] works right now”—without the user ever needing to think about which ground station or band they are connecting to. Meanwhile, Hirokazu Mori of Warpspace predicted an exponential increase in optical ground stations (OGS) over the next few years to handle the massive data loads that RF signals struggle to carry.
Commodity discussion
Another theme of the panel was whether the space industry has truly reached the era of commoditization—where hardware and data become standardized, cheap, and interchangeable. The panel said “no,” though with different nuances across the value chain.
In speaking about Earth Observation data, David Proulx of Waterloo based SkyWatch said, “I don’t think EO data is a commodity yet in the strict definition of the term, because it is not like for like, replaceable or substitutable across providers, nor is there a surplus of it within the key dimensions of temporal resolution… And there’s actually sort of a relatively few viable providers at each layer of the resolution stack.”
Hirokazu Mori of Warpspace agreed, noting that while standardized formats are improving, customers ultimately value the “insight” derived from the data rather than the raw pixels themselves. “In that way, I think data itself is not commoditized. It’s not commodity yet. But some of the Insight coming from especially optical data is… not commodity.”
On the hardware side, John Malsbury of Any Signal painted a vision of a future where satellite manufacturing mimics the automotive industry, but admitted the supply chain isn’t there yet. “I believe in the world where a satellite bus and all of its content should be about the price of a Kia Soul… [which] sells for $21,000 MSRP… I think there’s a lot of juice to squeeze out of commoditization of the busses themselves.”
Ali Younis of Mynaric pointed out a paradox in the hardware market: rapid innovation prevents products from settling into a commoditized state. Just as a standard begins to emerge, customer demand shifts to the next level of performance. “What I’m noticing is the innovation cycles and the demand for the next faster terminal are becoming so much shorter that… it’s not allowing the industry to fully commoditize to the point of COTS [Commercial Off-The-Shelf]… because now 100 gigs is becoming the standard, and people are already asking for a terabit.”
Defense vs. commercial
The panel described an industry in transition, where defense contracts currently pay the bills, but commercial scalability remains the ultimate goal. Ali Younis of Mynaric said that government contracts currently account for 70-80% of the company’s revenue, largely driven by the Space Development Agency’s (SDA) standardization efforts, which have kickstarted mass manufacturing. However, he predicts this split will shift to a 60/40 balance in favour of commercial markets as private constellations come online.
Other panelists took a “commercial-first” approach to avoid the slow pace of government procurement. John Malsbury of AnySignal noted that his company focused on commercial partners initially because he “didn’t want to start my company with servers and wait for six months to hear back.” Similarly, David Proulx of SkyWatch noted that while they started with a commercial focus, the geopolitical climate is pulling them into “sovereign ISR frameworks,” where governments need commercial data to complement their own national assets.
Orbital data centers: Hype or inevitability?
When asked if “Orbital Data Centers” were inevitable or overhyped, the panel largely agreed on inevitability, but with caveats regarding the timeline.
Mynaric’s Ali Younis argued that the term “data center” might be misleading. Instead, he sees “Physical AI” as the real trend, where every satellite eventually adopts AI processing, effectively becoming a data node by default.
Hirokazu Mori of Warpspace suggested a pragmatic middle ground. He predicts that pre-processing tasks—like cloud removal—will move to space within the next few years to save bandwidth, while heavy computation will remain on the ground.
While agreeing on the destination, John Malsbury offered a reality check on the speed of adoption. Citing the ambitious timelines often seen in the industry, he suggested applying a “factor of 3x” to current predictions for when true orbital data centers will be operational.