Kevin Cushnie is an Engineering and Technology leader bridging the gap between strategy and execution.
A 2025 Gartner survey of over 500 technology leaders found that 72% of organizations are breaking even or actively losing money on their AI investments. Adoption is no longer the problem; the problem is production and scale.
We are living through an output crisis. According to Hype’s 2025 State of Corporate Innovation Report, “87% cite turning ideas into business outcomes as the top pipeline obstacle.” The workshops run. The sticky notes accumulate. The decks circulate. And yet almost nothing ever ships.
There are four disciplines I’ve seen that separate organizations that build genuine innovation capability from those who are performing innovation theater.
Frameworks Collapse Where Production Begins
Consultant-designed innovation models are prescriptive. They treat innovation as a series of sequential gates: ideate, validate, prototype, pilot and scale. While these work on a whiteboard, they don’t translate to production.
Production environments are living systems. They collide with legacy infrastructure, organizational politics and constraints that frameworks typically don’t anticipate. Research into innovation leadership highlights a core tension in this space: Current incentive structures reward decisiveness and staying in lanes, while genuine innovation demands detective-like thinking, building an evidence base from the ground up and following where data leads.
I’ve watched this pattern play out repeatedly when building rapid prototyping disciplines with engineering teams. The framework looks elegant during the planning phase. Then the plan is derailed by a subsystem that wasn’t documented, a data model nobody fully understands or an approval chain that adds three weeks to every deployment. The teams that ship build their timelines around these collisions; they expect the unexpected.
Fixing this means letting the people closest to the code and the customer define their own implementation path, with iterative discovery funded as a first-class activity.
The Six-Week Discipline
The strategic shift that matters is moving from “building the most” to “learning the fastest.” AI-augmented development now makes it possible to compress the build cycle dramatically, with senior engineers using AI to parallelize code generation, test scaffolding and environment preparation. The goal is reaching real users and real data faster, because market intelligence gathered in production outweighs months of research and speculative planning.
A case study from Tech-Stack illustrates this discipline. MenuReady, an AI-powered food photo enhancement tool for independent restaurants, reached production in exactly six weeks. The team bypassed traditional subscription models, launching with a frictionless pay-per-photo approach. That early entry provided three months of market intelligence that they would not have had if they’d remained tied to a six-month roadmap.
Another critical enabler is platform engineering. Without a stable foundation handling infrastructure, security and data management, every innovation team has to reinvent the wheel. Google Cloud research found that 86% of organizations consider platform engineering essential to realizing the business value of AI. The platform absorbs complexity so that product teams can focus purely on value creation and speed to deployment.
Measure What Ships, Kill What Doesn’t
Innovation theater thrives on vanity metrics: ideas generated, workshops conducted or patents filed. A 2025 analysis of innovation KPIs reveals a quiet revolution in how success is being defined. Since 2015, patent and IP metrics have declined by 7%, while real-time tracking of commercialization and time-to-market has increased by 150%. Companies using formal KPI systems achieve 2.1 times higher innovation ROI than those without them.
Two metrics deserve particular attention. First, new-products-to-margin conversion, which measures how well sales of new products translate into actual gross margins. Revenue without profitability is just expensive activity. Second, the product kill rate. High-performing organizations actively track what they stop. A healthy kill rate signals that resources are being redirected toward viable projects and that “zombie projects” (prototypes that are neither launched nor killed) are being cleared from the portfolio.
One caution: According to the 2025 study linked above, approximately 41% of employees report measurement fatigue. Leaders who pile on dashboards without distinguishing between visible output and invisible value, such as peer code reviews or architectural decisions that prevent future debt, risk sabotaging the culture they’re trying to build.
Systems That Learn From Shipping
The compounding advantage of practitioner-led innovation is that implementation experience accumulates. Traditional frameworks start every project from zero, whereas practitioner-led systems treat each deployment as training data for the next one.
In building rapid prototyping disciplines with engineering teams, I’ve seen this compounding effect firsthand. The first six-week cycle is a grind. The second is noticeably smoother. By the fourth or fifth, teams have built muscle memory: reusable platform components, deployment playbooks refined through production friction and pattern recognition for the failure modes that derail early-stage products. Manufacturing data confirms this principle at scale, with digital twin workflows delivering 50% reductions in development costs and 30% faster time-to-market.
The Mandate
The era of innovation theater is reaching its natural conclusion. Ideation without implementation is a liability. Organizations that reward shipping over storytelling, measure deployed impact over workshop attendance and treat every production deployment as a chance to compound capability can build structural advantages that frameworks alone cannot deliver.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
