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

Step 10: Share Learning
Whether your experiment led to a successful solution that you are prepar-
ing to launch or failed to solve the defined problem, it is vital to preserve 
the learning that came through your process. It is therefore important to 
have a formalized process for capturing, sharing, and accessing the learning 
from any divergent experiment. This includes archiving or documenting 
prototypes you developed, solutions you tried (which may not have worked 
but could inform others), and lessons you learned.
You can find a list of sample questions to use in capturing and sharing 
learning from any divergent experiment with your team in the Tools sec-
tion of http://www.davidrogers.biz.
Four Paths to Scaling Up an Innovation
So you’ve developed a successful innovation. Now what?
One of the ways that the digital revolution has changed innovation 
is in defining its end point. Innovation used to focus on a finished, pol-
ished product for launch into the market. Now, with the addition of data 
and software to nearly every offering, businesses have the opportunity to 
continue rapidly experimenting with and evolving their innovations even 
after launch.
Companies like Google are famous for launching products as an explic-
itly incomplete beta to get user feedback on how to finalize the design. 
Pierre Omidyar launched eBay after coding the first version of its website in 
three days. This is a classic example of the start-up philosophy of launching 
a minimum viable product directly to consumers—in essence, running the 
process of experimentation in the public eye.
But launching an MVP is not an option for every company or every 
innovation. If you are Ford Motor Company, you can’t put an MVP for 
a new car on the road for customers to buy while you are still testing its 
market fit. Apple has good reason for maintaining the secrecy surrounding 

---

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

155
its products before unveiling them rather than releasing product betas for 
early adopters.
There are four general paths for scaling up an innovation to a full 
release. To understand which path you should take, you need to answer 
two questions:
r
 Can you iterate this offering quickly after launch? For software products, 
iteration is generally easy via online updates. For services, iteration is 
also often possible (e.g., launching a new sales process that you can 
adapt based on feedback). However, for physical products or physi-
cal designs such as retail environments, rapid iteration after launch is 
rarely an option. If your innovation is heavily dependent on partners or 
constrained by regulations, you may also not be able to iterate quickly.
r
 Can you limit your rollout to stages, or does the innovation have to be 
released to all customers at once? You may be able to limit the rollout 
of an innovation to specific locations (e.g., a retail design or a local 
service). You may be able to limit it to a subset of customers (e.g., by 
invitation only). You may be able to limit the duration of a new offering 
(e.g., a holiday menu item or a limited prerelease of your next video 
game). For other projects, though, it is will be necessary to offer your 
innovation immediately to anyone who is interested.
Your answers to these two questions will place you in one of four 
quadrants (see figure 5.4). Let’s look at the requirements for successfully 
scaling up an innovation in each quadrant.
Polished
roll-out
Cannot iterate
quickly after
launch
Can iterate
quickly after
launch
MVP
roll-out
Increased
pressur
e