Use smart experimentation to innovate.
Experimentation sits at the very core of startup’s ability to innovate.
Experimentation sits at the very core of every startup’s ability to innovate. This is nothing new. For centuries, the systematic way of testing ideas has been enabling companies to go from the inception of a product to its creation. De facto, every service or product was once just an idea that was later shaped and refined through a process of experimentation.
The commercial success of new products and projects requires running thousands of experiments. Such extensive testing was not a long time ago a painstaking and costly strategy, but today, it is entirely possible, thanks to rapid prototyping, computer simulations, and combinatorics. These new technology-driven experiments are becoming not only cheap but also easy to run at any startup.
Over the past few years, while building new products, I have gathered a few insights into the process of experimentation. Here are some tips that might prove helpful for startup founders.
1. Avoid the good ideas trap.
We all have at least once been a victim of the good ideas trap. We spent hours brainstorming, discussing the best ideas that, at the end of the process, seemed radical — but, when it finally came to their execution, turned out to be too costly, or just did not work.
As Guy Kawasaki argues in “Reality Check: The Irreverent Guide to Outsmarting, Outmanaging, and Outmarketing Your Competition”, startups that consistently encourage and celebrate good ideas tend to become advancement underachievers in the long term.
This happens because, as much as good ideas are generated as a result of debates and analyses, they also, too often, become a central point of a team’s innovation effort. This is the crux of the good ideas trap. Good ideas have no meaning to a business if they are not tested.
Tip: To avoid the good ideas trap, approach every innovation like a scientist, and spend less time on identifying and developing good ideas and more time on testing relevant hypotheses.
2. Normalise failure.
I am not trying to say that failure per se is a useful objective, far from it. Without providing new insights, failure might lower morale, infuriate customers, or damage reputations. But failure, if approached well, is the key to innovation and fostering an agile startup culture.
The wisdom of learning from failure is incontrovertible. Indeed, startups cannot possibly undertake the risks necessary for growth if they are not accustomed to the idea of failure. But what sounds like an easy act — failure, in practice, often turns out to be a barrier that is hard for most of us to overcome. The disappointment that comes from failure, the fear of underachievement, our egos, and the conflicting incentives and reward structures, are all obstacles that inhibit creativity, and therefore innovation.
So even if people understand that they can and should fail, in reality, they do everything possible to avoid it.
To foster a culture in which people aren’t afraid of failing, the key is to focus on the communication of the value created by failure and not on the failure itself. In order to shift the current mindset, make sure to realign your team’s incentives first. To achive that, some startups create incentives that reward those that learn the most and those that have the courage to share their lessons from failures; These startups openly distribute “failure reports”, set up short debriefing meetings that celebrate the latest flops, and share the defeats tableaux. Taking an informal approach to failure can be very effective.
Tip: Experiments shouldn’t just prove whether the original idea was correct. They should place stress on all assumptions, break them apart, and result in many different outcomes that might be stronger and better than what was initially assumed. In order to break the “playing it safe” habit: make sure that you familiarise your team with the idea of failure, by repeatedly emphasising its positive value.
3. Test often, fail fast.
Michael Shrage, an author of “The Innovator’s Hypothesis” claims: Ideas are cheap. What startups should be doing is experimenting frequently to validate or invalidate good ideas, minimising risk through a series of quick tests, and iterating on the learnings generated by manageable failure.
Make sure that you start framing a team’s innovation proposals as soon as possible in the form of testable hypotheses. The moment when you make an assumption while going through a process, check whether this assumption could be tested and verified now, not later. By doing this, you will not only be able to move forward with a more and more realistic product or project but also make sure that ideas that would normally be killed off on whiteboards and in follow-up meetings, will actually see daylight. —
Creating an environment where it is easy to try new ideas and actively generate actionable data by running experiments means that it becomes easier to keep everyone in the team motivated, as everyone gets to see all their ideas live in the real world.
Tip: Keep the team running the experiment small to ensure agility and adaptability. (3–4 person team made up of an engineer (a technical expert), a developer (an implementor), and a business analyst can be the most effective.
4. Play with both old & new.
The latest research by Marco Iansiti of Harvard Business School has found that, in many industries, the ability to integrate both old and new technologies is crucial to developing superior products and services. His research indicates that new experimentation technologies might only be effective 60–70% of the time. However, by combining new and established experimentation methods, startups can avoid this performance gap, while taking advantage of cheap experimentation.
IDEO, for example, still largely advocates the development of rough prototypes that people are invited to instantly test and modify — a process that, it claims, ultimately leads to better products.
In fact, technologically advanced models and the increased automation of experimental processes might never fully remove the human element in innovation.
Tip: Whether you decide on using an advanced analytics tool, a regression model, or just a piece of cardboard should be considered carefully, do not be swayed by the hype of the latest technologies. Be open to various media of experimentation to suit your specific needs and get you to the most realistic answer.
___
Gaining new information from experiments is a powerful tool that adds to a startup’s competitive advantage. It not only significantly improves the team’s ability to assess new ideas, but also helps to develop outstanding products. This is because unpromising components, design features, and stretched assumptions can be assessed and eliminated before hundreds of thousands of dollars are invested in their development.
Test often and learn from failures, not to create new rules, but to push your team into ever more exciting experiments. When you adopt this practice of continuous experimentation and curiosity, business growth will follow.
__
by Agata Nowicka — a serial entrepreneur and investor based in London, UK.
follow me on ->Twitter
follow my ventures on -> Pixel | Female Foundry