This is the third article in my series on how to create a vision and a strategy, prioritise your work and create a roadmap for new and existing products.

TLDR; The key is to move from high uncertainty to low uncertainty as quickly and cheaply as possible.

There’s a sneaky problem with innovation: it always looks exciting and easy at the beginning. Any idiot can come up with an idea, and the beautiful moment of genesis is a heady time of unlimited opportunity, unblemished by the scars of commercial reality.

Innovation, however, isn’t creativity. The innovation of new businesses and new product lines must demonstrate at least two critical components of success. Many teams, founders and executives stop at the first; identifying a problem which people wish to be solved.

But being able to solve a customer problem isn’t close to being sufficient to make a good business. All too frequently, innovators don’t answer the second question quickly enough; will anyone be motivated to pay to keep a business running? It is a question that will inevitably be answered even if it’s when the business or product fails.

In the startup world, this has been regularly called out and subject to high visibility research. A well known piece of analysis from CB Insights suggests that the most common reasons that startups fail are commercial realities that should have been avoided. These reasons should have been obvious before even getting close to burning cash on engineering; no market need, problems with pricing or even just having boundless optimism in place of a business model.

The existential test failed by these young businesses is the collision between uncertainty — of assumptions made about business models or justifications for creating a product — and the cold reality of turning the product into sufficient revenue. Despite the religion of Lean Startup, the rush to engineer a solution (beware: even a Minimum Viable Product counts as engineering) is too great a lure for founders to avoid.

I’ve been fortunate to work with many startups and larger businesses and the refrain is the same whatever the scale — the siren song of building product for (apparently) obvious customer needs is too tempting for startup founders and corporate business leaders alike.

In my experience large companies are as bad as, if not worse than, startups when delivering their innovation projects (and remember it’s a common claim that startups have a 1 in 10 chance of survival). Startup failure is a matter of public record and obvious to observers; the business simply ceases to exist. Failures in large corporate innovation projects become rounding errors on programme budgets, with many dying products left withering with no hope of a merciful end.

The CB Insights data publicly exposes a problem which also exists (albeit privately) inside large businesses that need to develop new product lines or diversify away from their core — what McKinsey might call ‘Horizon 2’ and ‘Horizon 3’.

My frustration watching company after company (and startup after startup) rushing to build products with too little thought led me to obsess over identifying a simple, repeatable process for gating investment in new products.

The elusive quality missing from many innovation processes seemed to be a basic rightness that allowed them to apply equally to startup businesses, new product features, corporate innovation programmes and even innovation outside industry. At its core there seemed to be one simple requirement for a universal process: drive down uncertainty before increasing investment.

Thinkers and Doers

Sadly, typical business structures often split people into those who have ideas (the thinkers, often executives), and those who implement them (the doers, the teams throughout the business). This leads to a deep disconnect — the thinkers are quickly frustrated as the doers struggle through the hard reality of implementing an idea that has had insufficient consideration.

For the thinker their beautiful spark, still bright and fraught with possibility, has been ruined by the doers who once again killed something beautiful and boundless. The doers, on the other hand, are left frustrated and bemused as they struggle to build something that is increasingly removed from being reasonably achievable or desirable.

This represents all kinds of wrong-thinking, the most insidious being the classification and separation of management ‘thinkers’ and team member ‘doers’ into some sort of Taylorist silos. In addition there is the classification of innovation as the preserve of the most senior employees, separating the responsibility for ideation and delivery, siloing innovation teams away from the core lines of business or, most telling of all, confirming that there is something mystical about innovation by hiring a ‘Chief Innovation Officer’.

The underlying assumption that we have been programmed to accept about a new idea is that it has a fixed chance to succeed dependent on its quality — that good ideas will succeed and bad ideas won’t. Typically the chance of success (or validity of the idea) is considered to be greater the more the originator gets paid.

This is bullshit.

Not all ideas are created equal, of course, but even bad ideas can lead to true innovation when they’re given fertile ground and disciplined curation.

The truth is that any idea needs to be tested and challenged. We need to give ideas room to breathe and for teams to test their riskiest assumptions before we commit resource to them. We’re looking to continually drive down our uncertainty as we drive up our investment. This means that we need to look for the quickest and cheapest way to challenge the assumptions in a new idea right at the beginning.

“Failures are cheap if you do them first. Failures are expensive if you do them at the end”

Innovation is not a four box model

I was once tasked with drawing up a new innovation process for a corporate innovation function. The process would be used by teams of highly talented people to work with new ideas and get them into production. It was important that the process combined the best elements of agile development, lean startup and design thinking.

It didn’t really seem to matter what activity was represented by the boxes, but we did spend a lot of time arguing about the names. What was really important, however, was that there were exactly four boxes. No more, and no less. Four.

This thinking is inspired by Scott Edgett and Robert Cooper’s Stage-Gatemodel, a set of proprietary, trademarked project phases that seem too restrictive, too procedural and too prescriptive to adequately marshal the chance, creativity and fluidity of entrepreneurial innovation.

I found the idea that subjecting the delicate structures of innovative concepts to this rigid phasing jarred with my experience. What I was looking for was a simple, understandable, universal process for gating and advancing projects which demanded that individuals and teams take ownership of their own innovation, choose their own tools and justify their own success.

A grand unified theory for developing new propositions

A big experiment in understanding the innovation process was my experience with Monday Labs, our internal incubator at reed.co.uk. In many ways, Monday Labs benefited from the evolution of the Cooperesque investment gating which Reed used centrally for our programme management process.

Monday Labs was fertile ground for three young teams to work as internal startups: Metrodesk, Reed Commercial and Startup Startup. These teams of 4–5 people from across the business were subject to three month investment gates; they were required to produce business plans documenting a target three months in the future, and left to their own devices to achieve them.

The young teams in Monday Labs burned brightly, far outperforming the traditional business in innovation and delivery speed. However, a key learning from this process was that fixing gates around set milestones didn’t work. The teams and their projects were too different and it was clear that a regular, fixed cadence wouldn’t work. Instead, I needed something more flexible and fluid, but still disciplined.

The more I struggled with procedural phasing and fixed milestones, the more I began to believe that there was just one underlying process; a single, unified methodology for gating investment, providing the correct resource and ensuring that the team were being pushed to deliver.

The process asks a number of questions of an innovator, instead of asking for a business case for a prescribed solution.

“We’ll consider giving you some resource to progress your idea”, we say, “if you answer these questions:”

• What outcome do you expect from running the process? What assumption are you testing?
• What team and other resources do you need to successfully deliver the outcome?
• How long will it take to reach the outcome (good or bad?)

It’s the responsibility of the innovator to answer these questions and justify how they will drive down uncertainty. They can use any tools or team at their disposal, but must be always aware that the lower cost, more brief experiments have a greater probability of being pursued. This balance of effectiveness and efficiency produces tight learning loops of innovation.

The innovator should consider what tools and processes they will use to test the idea; it might be good, old-fashioned boots on the ground market-research to bootstrapping businesses and Wizard of Oz tests in WordPress. It might also be even more disciplined and curated, like Jake Knapp’s Design Sprints or Bill Aulet’s Disciplined Entrepreneurship.

Wherever possible, the team are encouraged to avoid anything resembling development (or real physical engineering) until absolutely the last moment due to the massive increase in cost that even the most simple engineering requires.

After the team propose an activity and receive approval, they simply need to come back with the answers to the next question, “What did you just learn?”, and then, “Based on what you now know, should we proceed?

If the answer to that question is yes, we ask the same set of familiar questions again:

• What outcome do you expect from running the process?
• What team and other resources do you need to successfully deliver the outcome?
• How long will it take to reach the outcome (good or bad?)

As each stage progresses, more organisational knowledge will be developed. How did we segment the market? Why did we choose the target we did? Where else might be ripe for development? What ideas did we park along the way? This knowledge should be captured and saved for later, expediting each subsequent innovation cycle.

“We never throw an idea away because you never know when someone else may need it”

— Art Fry, co-inventor (with Spencer Silver) of Post-It Notes

The art of thinking small

For leaders, and those holding budget, this process lowers costs by reducing investment in projects that will be ultimately unsuccessful. In large organisations, the mistake is often that innovation initiatives increasingly run towards larger, and more expensive ‘moonshot’ gambles. In a fascinating article for MIT Sloan Management Review, Corstjens, Carpenter, and Hasan compared the relative R&D spending and growth between P&G (a large R&D spender) and Reckitt Benckiser (a relatively small spender).

Startlingly, Reckitt Benckiser demonstrated three times greater CAGR sales performance, for less than half of the relative R&D investment. Corstjens et al point out that there is a marked difference in the type of innovation in both businesses — P&G expect big returns from big investments, where Reckitt Benckiser’s projects are less risky and far less costly than other firms.

“Innovation is about getting many base hits and occasionally hitting the home run. You very rarely win a baseball game just by hitting home runs.”

Bart Becht, CEO of Reckitt Benckiser from 1999 to 2011

Embedding the discipline

This simple model is univeral; before undertaking an activity — from considering a new product feature to investing in a multi-year enterprise software rollout — simply stop and consider how exposed you are to the unknown. Have you really been thoughtful about the assumptions that you’re making. Would someone else’s perspective help you make a better decision? Should you spend some time trying to learn more before progressing, and what process will help you get data to make better decisions?

“Become a learning organization through relentless reflection (Hansei) and continuous improvement (Kaizen).”

— Toyota Principle #14

We often race to DO because doing is a proxy for results. Activity looks like something that will lead to an outcome. Sometimes it’s best to stop and think about whether doing is really a good idea. Instead we should seek to unseat our assumptions; at each stage of a learning innovation process we can, with good curation, create organisational knowledge with each turn through the cycle, regardless of whether the assumption was proved or disproved. Each activity delivers a wealth of data and understanding which can inspire more innovation and reduce the time for future experiments.

Each project that runs through this discipline turns a disproved assumption into organisational success, always enriching the organisation with greater learning. It naturally enhances agility by favouring more test and learn cycles which encourage greater success from the subsequent longer, more expensive cycles.

Whether you’re a startup making a bet on a new business or a large corporate investing in innovation programmes and new product lines, always focus on how you can move from high uncertainty to low uncertainty as quickly and cheaply as possible.