Innovation Isn't a Department
Companies that want to stay competitive often create innovation labs: dedicated teams, separate from core operations, charged with building the next breakthrough. Give smart people room to experiment, the thinking goes, and good things will follow. Sometimes they do. One lab we worked with built a solid prototype that could have redefined the way the company works. They validated it with users and it was ready for integration. But the product teams had a full roadmap. The prototype used a different tech stack, so no team could own it. A year later, the innovation lab turned into a regular team, maintaining their own products instead of building new ones. The instinct to separate innovation makes sense. Clayton Christensen's Innovator's Dilemma showed why successful companies struggle with disruption: they're too good at serving existing customers. When a disruptive technology emerges, pursuing it looks like a bad business decision. The rational response, focusing on profitable customers, becomes the fatal one. Christensen proposed a solution: create separate units to explore disruptive innovations. Many companies followed this advice, but most failed anyway. Separate innovation units can work well for pure research and early-stage development. The model breaks down when those innovations need to be adopted by the core organization. Either keep the unit truly independent as a separate business, or embed innovation capacity across existing teams. Harvard Business Review research found that the vast majority of innovation labs don't deliver on their promise. Sometimes labs respond by shipping products themselves, but this creates its own problems. Innovation and maintenance require different skills and mindsets. Once a lab starts maintaining production systems, it becomes just another team focused on operations rather than experimentation. Why separation fails Cultural isolation. Innovation labs develop different values, risk tolerance, and ways of working. When projects need to return to the main organization, they face antibodies. The engineers who stayed behind resent the special treatment, and middle managers see threats to their budgets. "Not invented here" becomes "not wanted here." Misaligned metrics. The core business measures quarterly revenue and margin. The lab measures experimentation velocity and learning. Neither understands the other's definition of success. When budget pressure hits, labs get defunded because they can't show traditional ROI. Research on organizational ambidexterity by O'Reilly and Tushman found that companies need tight executive integration between exploratory and core units, but most never achieve it. Trapped capability. Google's Project Aristotle found that psychological safety, the belief that you can take risks without punishment, predicts 43% of the variance in team performance. Siloed labs might have it internally, but the broader organization usually doesn't. Innovative thinking stays trapped in the lab, and the rest of the company never learns to think differently. Executive attention drift. When the core business hits trouble, leadership focus shifts away from the lab. Budgets get frozen, timelines extend, and the best people get pulled back to "real" projects. Labs need sustained executive sponsorship to survive, and that sponsorship evaporates the moment quarterly numbers disappoint. Survival Kodak invented digital photography in the 1970s. Nokia saw smartphones coming years before the iPhone. Both companies had innovation happening somewhere in the organization. They had the ideas but couldn't execute on them. Research comparing Nokia and Kodak's failures found that both exhibited "cultural rigidity, misaligned leadership strategies, and reluctance to cannibalize profitable core businesses." They knew what was coming but couldn't act on it. HealthManagement.org's analysis found that at Nokia, fear drove dysfunction: top managers feared external competitors, middle managers feared internal politics. Nobody felt safe enough to push for radical change. Kodak conducted a study in 1981 showing they had about ten years to prepare for digital photography. They buried it to protect the film business. The capability existed, but the culture to use it didn't. Embedding innovation in the organization BCG research found that companies promoting innovation culture across the organization are 60% more likely to be innovation leaders. O'Reilly and Tushman studied ambidextrous organizations, companies that integrate exploration with exploitation under unified leadership. These succeed at breakthrough innovation 90% of the time. Siloed approaches succeed less than 25% of the time. Create space for experimentation Traditional metrics strangle experimental work. 3M lets employees spend 15% of their time on self-directed projects, which is where Post-it Notes came from. Google used to have 20% innovation time, which produced Gmail and AdSense. Create different success criteria for experimental work, but apply them across the organization. Make failure risk-free When failure carries career risk, people stop trying new things. Organizations that want innovation need to celebrate learning from failed experiments and promote people who took smart risks that didn't pan out. This requires changing how the whole organization responds to uncertainty. Rotate people and ideas When innovation lives in a lab, capability concentrates rather than spreads. Rotate engineers and product people through exploratory projects. Everyone should know how to run experiments, test assumptions, and kill ideas that aren't working. Bet on many small innovations Most corporate innovation gets one shot, but venture capitalists expect most bets to fail. The same portfolio thinking works for internal innovation: run many small experiments, kill the ones that aren't working, and double down on what shows promise. McKinsey's research found that top innovators have adopted agile practices organization-wide at eight times the rate of lagging companies. That means actual agility in how decisions get made, not just adopting the methodology. Culture eats labs for breakfast Kodak and Nokia both had the technology. They saw what was coming and built prototypes years ahead of competitors. What they lacked was an organization willing to bet on its own inventions. Innovation embedded in your culture can sustain itself over time because the whole organization learns to think differently. Innovation isolated in a lab rarely survives long-term.Lumia Labs works with organizations navigating technical and cultural change. If you're rethinking how your engineering teams approach innovation, we'd like to hear from you.