Discovery Driven Planning: Converting Assumptions Into Knowledge

Discovery Driven Planning: Planning to Learn
Rita Gunther McGrath • [email protected] • www.ritamcgrath.com
What do Revlon’s Vital Radiance line of cosmetics, AOL’s failed $850 million acquisition of Bebo,
the disastrous $1 Billion 2010 launch of the Microsoft Kin phone and Church and Dwight’s
since-discontinued “small size” condoms all have in common? All these flops share a pattern:
Untested assumptions, taken as fact, few low-commitment tests prior to making a major
commitment, little appetite for redirection and when failure loomed, no strategy in place to
recoup or benefit from the failure. In short, the companies behind these failed launches
planned their way to disaster.
The problem starts with a deceptively simple ratio, the assumption : knowledge ratio. When
that ratio is low, in other words, when you are planning with a great deal of information and
can confidently predict what is likely to happen, conventional planning works like a charm.
When that ratio starts to get larger, however, conventional planning can lead you astray,
sometimes disastrously. An alternative, popular with entrepreneurs, corporate ventures, and
even not-for-profits and internal projects, is to use discovery driven planning. Discovery driven
planning is an approach that recognizes that in an uncertain situation, you can only plan
confidently to the limits of your existing knowledge. So it gives you permission to do that! As
you reach that limit, the approach suggests that you stop, revisit your critical assumptions,
update them, and redirect your efforts based on the new information you now have. In this
way, you can proceed into highly uncertain activities while containing your risk. As the graph
suggests, you structure your investment so that you ramp up investment only as you decrease
the range of uncertainties you face.
Uncertainty / Cost Profile over time
Range of
Uncertainties
(Risk)
Checkpoints
Investment
Profile
(Costs)
A discovery driven plan proceeds in five steps: 1) determining what a successful outcome
would be; 2) doing some benchmarking; 3) specifying operationally what you would have to do
to accomplish the targets; 4) documenting and testing assumptions and 5) planning to key
checkpoints.