The innovation process is a structured sequence of steps that transforms a raw idea into something that delivers real value, whether that’s a new product, service, or way of doing business. Most frameworks break it into five core stages: generating ideas, screening them, experimenting with the strongest ones, preparing them for market, and rolling them out at scale. The specifics vary by organization, but that basic arc stays remarkably consistent.
Understanding this process matters because innovation without structure rarely works. The late Harvard Business School professor Clayton Christensen estimated that 95% of new product launches fail. A clear process won’t guarantee success, but it dramatically improves the odds by forcing teams to test assumptions early, kill weak ideas before they drain resources, and focus investment on what customers actually want.
The Five Core Stages
While different companies use different terminology, most innovation processes follow five stages that move from wide-open creativity to disciplined execution.
1. Idea Generation
This is the stage where quantity matters more than quality. The goal is to surface as many potential ideas as possible before filtering kicks in. Effective ideation uses techniques like rapid ideation (everyone writes down as many ideas as possible within a set time limit), starbursting (systematically asking who, what, when, where, and why about a problem), and figure storming (imagining how a specific person would solve the problem). Changing the physical setting or medium, like switching from a laptop to sticky notes on a wall, can shift how people think and unlock ideas that wouldn’t surface in a standard meeting room.
Once ideas exist, they need to move. Mobilization is the step where an idea travels from the person who had it to the team, department, or partner who can actually develop it. Many promising ideas die here simply because there’s no mechanism to carry them forward.
2. Screening and Evaluation
Not every idea deserves development resources. During screening, teams evaluate each idea’s potential benefits and problems, then make a judgment call about whether it moves forward. This is where you compare ideas against strategic priorities, market size, technical feasibility, and competitive landscape. The key is being honest: an idea that sounds exciting in a brainstorm may collapse under basic scrutiny, and that’s the point. Killing weak ideas early is one of the most valuable things a good process does.
3. Experimentation
Surviving ideas get tested, typically through prototypes or pilot programs. A prototype doesn’t need to be a finished product. It can be a simple model that tests whether a design concept works or whether users can navigate an interface. Prototypes exist mainly to test the viability of ideas, letting internal teams or selected users react and suggest adjustments before large-scale production begins. They’re generally inexpensive because they use simplified models focused on specific elements rather than full functionality.
4. Commercialization
Commercialization is where a validated idea gets packaged for the real world. This stage focuses on the idea’s potential market impact: clarifying how and when it can be used, bundling it with complementary features, and using data from experiments to demonstrate benefits. It’s essentially the bridge between “this works in testing” and “people will pay for this.” Teams define their go-to-market strategy, identify external partners for technology development, and in some cases form startup companies around the innovation.
5. Diffusion and Implementation
The final stage has two sides. Diffusion is the organization-wide acceptance of the innovation, getting buy-in from the teams who will support, sell, or maintain it. Implementation is the operational work of setting up everything needed to produce or deliver the innovation at scale. A product can pass every previous stage and still fail here if the organization isn’t aligned behind it.
How Stage-Gate Adds Decision Checkpoints
Many large companies use a variation called the Stage-Gate process, which adds formal decision points (gates) between each stage. At every gate, senior managers or a steering committee review the project’s progress and make one of four calls: approve it to continue, kill it entirely, put it on hold, or send it back for fundamental revision.
The Stage-Gate model typically includes a discovery phase followed by five stages: scoping (sharpening the product definition), building a business case, product design, testing and validation, and product launch. What makes it distinctive is that nothing advances without explicit approval. This prevents the common problem of “zombie projects,” initiatives that keep consuming resources long after the evidence suggests they won’t succeed. The tradeoff is speed. Gate reviews add time, and in fast-moving markets that delay can be costly.
Design Thinking and Lean Startup
Two widely used methodologies offer different lenses on the same innovation challenge. Design thinking is inquiry-based and open-ended. It forces you to put yourself in the customer’s shoes through sustained interaction, because you don’t know when the critical insight will come. You develop prototypes to gather qualitative feedback about what people actually need, not just what they say they want.
Lean startup makes that process more rigorous. Instead of relying on qualitative impressions, it frames every assumption as a hypothesis and tests it at the lowest possible cost. You build a “minimal viable product,” see what’s working, and double down on those elements. If the data invalidates your hypothesis, you pivot: keeping some elements fixed while changing others. Stanford’s Graduate School of Business teaches both approaches together, starting with design thinking to discover the customer problem and then applying lean startup to validate solutions. The empathy you gain through design thinking helps you identify where to pivot when data tells you to change direction.
Prototypes vs. Minimum Viable Products
These two terms get used interchangeably, but they serve different purposes. A prototype tests whether an idea is viable. It may not function completely and is typically shared only with internal teams or selected stakeholders. A minimum viable product (MVP) tests whether a market exists. It has to work well enough for real users to experience it and provide feedback on actual usage patterns, not just design preferences.
Building an MVP requires more investment than a prototype because it needs to fulfill its core function reliably. For example, an MVP for an e-commerce platform might include only a shopping cart and payment processing, nothing else. An MVP can develop directly from a prototype: you take the feedback and testing insights from your prototype, incorporate the essential functions, and produce something ready for early users. The progression from prototype to MVP to full product is how risk gets reduced incrementally rather than all at once.
Why Innovation Processes Fail
The most common barriers aren’t technical. They’re human. Research on innovation barriers consistently finds that interaction-specific challenges, tensions between people and teams who need to collaborate, are the most prevalent obstacle. Hierarchical organizational structures make this worse. When every idea needs to travel up a chain of command for approval, speed drops and the people closest to the problem lose influence over the solution.
Financial constraints matter too, but not always in the way you’d expect. Global corporate R&D spending hit nearly $1.3 trillion in 2024, a historic high, yet real growth (adjusted for inflation) was only about 1%. Global R&D growth slowed to 2.9% in 2024 and is projected to fall further to 2.3% in 2025, the weakest expansion in over a decade. Companies are spending more in absolute terms but getting less growth per dollar, which puts pressure on innovation teams to demonstrate results faster.
Measuring Innovation Performance
You can’t improve a process you don’t measure. The most commonly tracked innovation metrics fall into four categories. Input metrics track what goes into the pipeline: how many ideas are submitted, how much is spent on R&D. Process metrics measure efficiency: how long it takes from idea to launch, what percentage of ideas convert from one stage to the next. Output metrics count what comes out: prototypes built, patents filed, products launched. Outcome metrics capture what actually matters to the business: revenue from new products, return on innovation investment, and new customers acquired through innovation.
One particularly telling metric is the ratio of total sales from products developed in the last three years. It reveals whether a company’s innovation process is generating meaningful commercial results or just producing incremental updates to existing offerings. The right metrics depend on your innovation strategy and how mature your process is, but time-to-market and idea-to-launch ratio are useful starting points for nearly any organization.