The ambition to build a digitally-connected, all-domain targeting web across the UK’s defence estate is bold. It’s the kind of transformative aspiration we expect to see in Strategic Defence Review 2025. But let’s be honest, trying to deploy a massive, integrated targeting web across all services and platforms, from day one, isn’t just bold. It’s symptomatic of the high-effort, high-waste procurement models we’re trying to move beyond.
Whilst we must recognise the geopolitical urgency, the UK isn’t fighting a hot war today. There’s no excuse for boiling the ocean. Instead, we should take a page from the startup playbook, start small, move fast, iterate faster. We should apply the innovation strategies that actually work when building networks, ecosystems, and platforms. From Silicon Valley to Whitehall, the principles are the same: find atomic value, deliver it fast, and scale from there.
The Startup Approach to Defence Innovation
Solve Atomic Problems First (The Cold Start Principle)
As Andrew Chen argues in The Cold Start Problem, the most successful networks didn’t begin as sprawling global ecosystems. They began with an atomic network, a small, dense, high-value group of users with clear, immediate needs. Facebook launched at Harvard. Uber started with black cars in San Francisco. Tinder with sororities and fraternities at the University of Southern California. The same principle applies to defence targeting webs.
We don’t need a universal targeting web. We need working localised targeting threads, delivering value now, in a real mission context. Whether it’s counter-piracy operations off the Gulf of Guinea, SO19’s tactical coordination, or force protection at RAF Brize Norton, these are live operational problems that demand (better) targeting networks today. That’s where the targeting web should start.
Build MVPs, Not Mega-Prototypes
The literature is clear: innovation doesn’t come from planning monoliths. In Four Steps to the Epiphany, Steve Blank describes how startups succeed by rapidly validating hypotheses with MVPs (minimum viable products). In defence, that means getting small targeting web components fielded, tested, and refined, fast.
Forget giant prototypes. Focus on dozens of micro-procurements, each tackling one slice of the targeting web problem. Test in real operational conditions. Learn. Iterate. Scale what works.
Deliver Value with Real Tempo
The most important startup insight? Speed matters. The UK’s COVID vaccine programme worked because procurement adapted to the tempo of the crisis: competitive, open, rapid. The same approach can transform defence innovation.
We need a “first-to-field” mindset, allowing any capable team (startup, prime, internal MoD unit, No.10 Innovation Fellows) to field a component that solves an atomic targeting problem today. The procurement system should be pulling these teams in, not keeping them out.
Scale
We don’t need to pause for a multi-stakeholder workshop on interoperability standards before we begin. Integration follows value, not the other way around. As atomic targeting webs demonstrate tangible impact in real missions, the need for alignment will emerge organically. That’s when standardisation becomes useful, not theoretical.
We can start small: one sensor, one decider, one effector, solving a real operational pain today. From there, we can add. Additional sensors, decision-makers, effectors, data sources, each plugged in when they’re useful, not when they’re written into a plan. This isn’t about building the entire targeting web from scratch or owning the architecture. It’s about catalysing momentum within a broader system. Prove what works. Surface what’s worth connecting. Make integration worth the effort. Until then, the priority is delivery. If we’re still focused on how to coordinate agreement, rather than how to deploy capability, we’re solving the wrong problem.
What Might an Atomic Targeting Web Look Like?
Here are just a few candidate missions where an atomic targeting web could deliver immediate value:
- SO19 urban operations support
Fuse intel, overwatch, and tactical comms into a responsive, mission-driven targeting web. - Counter-piracy tasking in the Gulf of Guinea
Link distributed assets, drones, ships, satellites, into a live targeting loop. - Base protection during airfield deployments (e.g., Estonia, Cyprus)
Build a rapid-deploy targeting web package for deployed expeditionary operations. - RAF Brize Norton perimeter defence
Integrate drones, sensors, and response units for automated threat detection and interdiction.
These aren’t hypotheticals. These are real problems, with real tempo, that the UK faces right now. If you’re a mission owner, or know one, we want to hear from you.
From Atomic to Strategic: Scale Comes Later
If we do this right, each successful MVP becomes the foundation of a larger, scalable targeting web. Atomic networks connect, lessons compound, and before long we have the basis for a national targeting network, not because we planned it from the top down, but because we let the right solutions grow from the ground up.
Just as every great platform begins with a single vibrant node, atomic networks don’t remain isolated. When MVPs are deployed into real mission contexts, they generate operational insights, build trust with users, and uncover new integration points. Over time, these networks begin to interlink, sharing data, protocols, and processes, forming the backbone of something much bigger. It’s not about imposing a grand architecture from day one; it’s about enabling organic convergence. This is how Facebook grew from Harvard dorms to global scale, or how Uber’s city-by-city rollout built a worldwide logistics engine. Defence must follow suit: link what’s working, replicate where useful, and let scale emerge from success. The targeting web won’t arrive in a single procurement cycle, it will emerge, node by node, through the compounding value of mission-proven atomic networks.
Let’s be the Facebook of digital targeting webs, not the Friendster.
Let’s build a defence innovation ecosystem that works like the best startups do.
Let’s start now.