It's easy to understand the appeal of innovation labs. When bigger companies invite smaller, nimbler players to work alongside them, they expose the business not only to new technology but also to different ways of using it. So why are so many innovation labs struggling to create value?

Innovation leaders like Disney, Target and Coca-Cola have recently shut down significant parts of their innovation efforts, and other corporate players are following suit, stripping investment from innovation labs or closing them entirely.

Sometimes it's hard to find a balance between genuine 'blue-sky' research (which has no defined goal) and applied innovation (which attempts to turn concepts into commercial value). Both require a change in organisational behaviour, and this isn't always practical.

Is it possible for example, to institutionalise the obsessive tenacity required to develop real breakthrough innovation?

Perhaps, but all too often the realities of quarterly earnings get in the way before any meaningful progress is made. One retailer that gets the balance right is Tesco, which asks its lab for a minimum of one breakthrough per year – any more than that is considered a bonus.

The trouble begins when heads of innovation start to believe their own hype, investing time and funds in concepts that are intellectually stimulating, but too far out and disconnected from the business strategy. Innovation depends on the ability to look ahead. But retailers or banks don't need the foresight of NASA to innovate. They need a disciplined program that tracks the context of the environment they are in.

Many companies mistakenly apply a premium on ideas, ploughing too much investment into innovating at the top of the funnel. But it's not feasible to start with 1,000 ideas – the pool is too deep. Labs need the discipline to pick 15 good ideas and reduce it down to five.

That's where the innovation starts. Somewhere along the way we created the myth of impossible things being done by people in white coats. But innovation is not about the art of the impossible – it's about the art of the feasible.

One way for organisations to focus on more achievable goals is to bring the lab closer to the business unit. Harvard Business Review research shows that this is starting to happen. Forty-six percent of innovation lab leaders report that at least some of their funding is supplied by business units, giving these valuable stakeholders "more skin in the game;" a quarter said business units provided the majority of their funds.

The risk of funding models like these, of course, is that ideas come with too many strings attached. The trick is balancing risk and reward.

Another reason labs fail is the mistaken belief that the success of Amazon, Facebook, Google and Apple has somehow changed the rules.

All four companies continue to invest heavily in high-risk innovation projects – and when they fail, are equally unafraid to buy answers to the questions their labs cannot answer.

Most companies are not Google or have bottomless reserves of cash to throw at big ideas. A successful grocery business, fighting for margin, needs discretion not bravery. A cost-focused utility needs judgement more than it needs originality. Vision is always helpful, but efforts to predict the future must be set in the context of the industry, the customer need and the talent at these organisations' disposal.

The rush to close innovation labs is premature, but innovators shouldn't expect to import start-up behaviours into a corporate environment without a defined plan. We see a matrix of three factors governing their future success: discipline, aggression and investment.

Discipline: Innovation labs chase too many ideas. A decent hit rate means starting with the feasible. Innovation works best under constraints. If you shoot for the moon the journey tends to be expensive with no guarantee of success.

Aggression: It's a myth that failure should be celebrated. It should be tolerated, but only as a platform for structured learning – through A or B testing, for example. If you're too aggressive, your hit rate will suffer.

Investment: The question is not about total investment, it's about spending money in the right way. What type of lab do you want to build? There's a big difference in value between thinking of things and making things. It's the combination of ideas and applied innovation that adds value. Good developers are expensive, but you can move much faster if you prove as you build.

Dee Burger is the Head of Digital Services for North America at Capgemini. Prior to his current role, Dee held the position of Head of Digital Services and he has also served as the Head of Capgemini Consulting and its Innovation & Digital Services (IDS) unit in North America, in addition to acting as the company's global sector lead for the Telecom, Media and Entertainment practice. In his career span of over 20 years, Dee has also held the position of vice president in Ernst & Young's Telecommunications practice before joining Capgemini.