The Digital Transformation Playbook: Rethink Your Business for the Digital Age

Step 4: Observe
The iterative development of ideas for your innovation begins with obser-
vation. Observation informs and provides the insights you need to solve 
the next stage of the problem you are working on. The goal of observation 
is to both deepen your understanding of the problem itself and broaden the 
range of ideas you bring to bear in finding a solution.
You should focus first on observing the customer’s context—to better 
understand the problem you’re trying to solve. Learn everything you can 
about the customer, the nature of the problem, and the context into which 
your solution needs to fit.
In addition, look for ideas from further afield. Look at other mar-
kets (how other customers deal with the same issue) and other industries 
(benchmarking from beyond direct competitors in your industry). You 
can also look to ideas generated in previous innovation efforts. IDEO, for 
example, maintains a “Tech Box” in each design studio, where prototypes 
and product ideas that were intriguing but ultimately not completed can 
be stored away for future inspiration. Rummaging in past ideas that didn’t 
quite make it may lead to unexpected discoveries for your current project.
Step 5: Generate More than One Solution
The next step is to generate ideas to solve the defined problem. This is the 
stage where your own intuition plays its proper role in innovation: to help 
create new ideas and possible solutions (not to evaluate them, which should 
be done by the customer).
There have been numerous books written on creativity and effective 
idea generation techniques. If you do not already have an ideation process 
developed within your company, I would highly recommend you read a 
few, incorporating the tools and processes that you find most helpful into 
your practice. Some of my favorite books include Bernd Schmitt’s Big Think 
Strategy, Luke Williams’s Disrupt, Drew Boyd and Jacob Goldenberg’s Inside 

---

I N N O V A T E B Y R A P I D E X P E R I M E N T A T I O N

151
the Box, William Duggan’s Creative Strategy, and Rita Gunther McGrath 
and Ian MacMillan’s Marketbusters.
The only strict rule here does not cover how you generate ideas; rather, 
it requires that you generate more than one idea. Your aim is not to run 
an intense brainstorming process and conclude with a single favored solu-
tion to the problem (perhaps after heated debate among the team members 
about the relative merits of others). Rather, your goal in ideation should 
always be to generate multiple viable ideas. (Recall the three very different 
solutions that Intuit initially proposed for the Indian farmers.) You will 
then, in subsequent steps, experiment on these ideas and use market feed-
back to determine which one to pursue and how to develop it.
Step 6: Build an MVP
By now, you should have some promising new ideas. But even brilliant 
ideas are not enough. “If you build it, they will come” may have worked 
for Kevin Costner in the move Field of Dreams, but in business innovation, 
great ideas are just the start of the process.
In this step, you need to translate your ideas into prototypes. In the start-
up world, the focus is on a minimum viable product, often an early website or 
app launched publicly so customers can start using it, responding to it, and 
identifying bugs or missing features. For an established enterprise, where it 
may not be appropriate to share early design ideas in public, I prefer the term 
minimum viable prototype. Either one can be abbreviated as MVP.
The most important point is that your MVP should absolutely not be a 
full-blown or finished product. The most common way to inflate innovation 
budgets is to overdevelop prototypes (through long and expensive technical 
development) before validating them with real customers. Scott Cook says an 
MVP should have “just enough features to allow for useful feedback from early 
adopters.”
34
Recall the makeshift prototype used to test Intuit’s Fasal service. 
The team didn’t build a software platform that could scale to millions of Indian 
farmers. They sent two employees into markets to gather data in person and 
had a third sit at a desk and manually send text messages to farmers to see if 
they used the data and if it actually helped them earn more money. This is a 
perfect example of the goals of an MVP: minimal cost + maximum learning.
If an MVP is successful, it will be followed by further iterations. As you 
progress, your successive prototypes should evolve from lesser to greater 
fidelity (e.g., from a sketch to a model to a working product) and from 

---

152
I N N O V A T E B Y R A P I D E X P E R I M E N T A T I O N