The failure case Mara is selling against is narrow and specific: the jammer is on and the quadcopter keeps flying, because its control link is a fiber spool or because it never needed a link at all. Mara says electronic warfare is becoming less reliable exactly as those two features spread [4]. The company's answer is geometry rather than transmit power. Co-founder and CEO Daniel Kofman says most defenses concentrate both replacement cost and line of fire in one place, while Spike spreads small, cheap nodes that can each aim and shoot in multiple directions at once [6].
The division of labour is plain enough. A "spotter" node watches with cameras and microphones, with radar as an option, and the software fuses those feeds to identify likely drones, track them and work out where they are heading [7]. A "seeker" then leaves a tube and flies itself into the target [8]. Mara claims the distributed launchers let it assign separate interceptors to several drones in the same moment, where guns and directed-energy mounts generally hold one line of fire per weapon and must work through a queue [10]. The nodes are meant to be attritable, which is the vendor's way of saying you are allowed to lose one [11].
Run the physics on the interceptor and the design logic falls out. At 250 grams and 200 kph, the seeker carries roughly 386 joules at impact [14]. That is a plausible budget for the kill mechanism the company describes, which is breaking a propeller, motor or arm rather than detonating a warhead [12].
What crews are actually rationing on a bad day is how many objects can be engaged inside the same ten seconds, and how many rounds are left in the tube [10]. That is the number that decides this procurement, and it is also the number nobody has published. Mara's material puts no price on a seeker; the only cost statement on the record is comparative, from Khosla Ventures partner Nicole Fraenkel, who says the system cost is orders of magnitude below anything else she has seen [13][15]. Set that against the attacker's economics, where an FPV airframe runs a few hundred dollars and the defense against it can cost many multiples once EW and missile interceptors are counted [3].
The forcing function for anyone shopping this year comes down to cost per intercept including the consumable, not the capital price of the mount, and the number of simultaneous engagements before a queue forms. Neither is hit probability. Multiply the first by the largest raid you think you will see, compare it with a few hundred dollars per airframe on the other side [3], and if your figure is larger by a factor your budget cannot absorb monthly, the problem is arithmetic, and no amount of capability language closes an arithmetic gap. Fraenkel's own description places Spike at user testing, which is a stage, not a fielded answer [13].
