Why Most Innovation Labs Fail, and What Actually Works

The launch event was excellent. There was a striking space with beanbags and whiteboards, a charismatic head of innovation and a demo that made the board nod. Two years later, the lab has produced a dozen prototypes, a handful of awards and no meaningful change to the way the company earns money. This pattern is so common that it has a name: innovation theatre. Corporate innovation management does not fail because organisations lack ideas. It fails because ideas have no route from the lab to the business.

Five Reasons Innovation Labs Fail

1. The lab is disconnected from the strategy.If the lab chooses its own projects based on what is interesting, it will produce interesting work that nobody in the core business needs. Innovation without a link to strategic priorities is a hobby.
2. There is no path to scale. Prototypes are easy to build and hard to adopt. Business units have targets, budgets and established processes. Unless someone owns the transfer, a promising pilot ends on a shelf.
3. The metrics reward activity. Counting ideas generated, workshops held or patents filed measures effort, not value. Teams optimise for what is counted, and the lab fills with busywork.
4. Talent and incentives are misaligned.Staff in the lab are often rewarded for novelty, while staff in the business are rewarded for stability. Neither side is irrational. They are responding to different incentives.
5. Leadership patience is erratic.Funding surges after a successful pitch and vanishes when quarterly results disappoint. Stop-start investment destroys the learning that makes later projects better.

Our article on why digital transformation is about people and processes makes a related point: new capabilities fail when the organisation around them does not change.

The Right Question: What Is the Lab For?

Before debating structure, define the purpose. A lab can serve very different goals.

• Exploring the edge. Testing technologies or business models that may matter in five to ten years.
• Extending the core. Improving existing products and processes faster than the main business can.
• Building new businesses. Incubating ventures that could grow into separate revenue streams.

A useful way to think about this is the Three Horizons model, described by Mehrdad Baghai, Stephen Coley and David White in The Alchemy of Growth. It separates the core business (Horizon 1), emerging opportunities (Horizon 2) and longer-term possibilities (Horizon 3). Each horizon needs different goals, timescales and measures. Judging a Horizon 3 experiment by quarterly revenue guarantees that it will be killed. Judging a Horizon 1 improvement by how radical it is will encourage unnecessary risk.

What Actually Works

The organisations that turn innovation into results tend to share six habits. None is glamorous.

Anchor every project to a business sponsor. Require a senior leader in the relevant business unit to co-own each project, with a real stake in the outcome. A sponsor who will eventually run the new product has every reason to make it work.

Fund in stages, tied to learning. Release money in small tranches. Each stage should answer a specific question, such as "Will customers pay for this?" Funding follows evidence, not enthusiasm.

Write the transfer plan first. Before a pilot begins, agree who will adopt it, what resources they need and what would make them say yes. If no one volunteers, that is useful information.

Measure learning, then adoption, then value. Early metrics should capture how fast the team is testing assumptions. Later ones should track customer adoption and commercial results. Using the wrong measure at the wrong stage causes most misjudgements. Our guide to decision-making under uncertainty provides methods for deciding when evidence is incomplete.

Set kill criteria in advance. Agree what results would end a project. Teams that expect to stop when thresholds are missed are more honest about progress.

Make it safe to fail well. Experiments should sometimes fail, or they are not experiments. Teams need to report disappointing results without fear. The research summarised in our article on psychological safety in the workplace explains why open reporting of problems is a feature of high-performing teams.

To see how this works, picture a manufacturer testing a predictive maintenance service for its customers. In stage one, a small team spends eight weeks interviewing ten customers to learn whether unplanned downtime is a genuine pain. Funding for stage two is released only if at least some customers agree to a paid pilot. In stage three, the service line sponsor commits two engineers to run the pilot, and kill criteria are written down: if fewer than agreed customers renew, the project stops. The example is illustrative, but the pattern is the point: each tranche buys an answer.

Alternatives to the Standalone Lab

A dedicated lab is only one option. Depending on your goals, consider:

• Embedded innovation teams inside business units, which reduce the transfer problem.
• Corporate venturing, where the company invests in or partners with external start-ups.
• Customer-led pilots, where a business unit invites a young company to solve a defined problem.
• Internal ventures with their own leadership and profit-and-loss responsibility.

Founders who build lean, technology-enabled businesses are reshaping what is possible, as our article on the AI-native startup explains. Large organisations can learn from their speed without copying their structure.

Learn from People Who Have Done It Before

Experience matters in innovation. Our piece on the hard-earned lessons that give serial entrepreneurs a measurable edge offers insights that corporate teams can adapt: how seasoned founders test assumptions early, spend cautiously and recover quickly from setbacks. Consider appointing someone with start-up experience to advise lab leaders, and treat their scepticism as an asset.

Link Innovation to a Wider Excellence System

Innovation works best as part of a coherent management system. Organisations using a business excellence framework examine strategy, leadership, processes and results together. That integrated view helps ensure that innovation priorities reflect strategy, that processes can absorb new ideas and that results are measured consistently.

Build the Bridge Before the Building

An innovation lab is not a solution. It is a tool, and its value depends on the bridge that connects it to the rest of the organisation. Define the purpose, fund learning, secure sponsors, plan the transfer and protect honest reporting. Leaders who do so will find that innovation stops being an event and becomes a habit. Explore more on this theme in our Leadership Edge section.

Frequently Asked Questions

Why do corporate innovation labs fail?

Common reasons include weak links to strategy, no route to scale, metrics that reward activity, misaligned incentives and inconsistent funding. These are organisational problems, not creative ones.

What is the Three Horizons model?

Described by Baghai, Coley and White in The Alchemy of Growth, it separates the core business, emerging opportunities and longer-term possibilities, each requiring different goals, timescales and measures.

How should innovation be measured?

Match the measure to the stage. Early on, track how quickly the team tests assumptions. Later, track customer adoption and commercial results. Counting ideas or events measures activity rather than value.

Do companies need a separate innovation lab?

Not always. Embedded teams, corporate venturing, customer-led pilots and internal ventures can work better, depending on the purpose and the company's ability to adopt new ideas.