!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> Streamline Training & Documentation: Brainstorming McKinsey Style

Monday, December 03, 2007

Brainstorming McKinsey Style

Recently, the wide-open style of brainstorming — encouraging people to range broadly in generating ideas, with no premature filtering — has been called into question (e.g., see these earlier posts).

An example of this rethinking can be found in the December issue of the Harvard Business Review, where Kevin P. Coyne, a McKinsey alum, and Patricia Gorman Clifford (pdf) and Renée Dye, current McKinsey consultants, describe the semi-structured approach to brainstorming that they find more productive than a wholly unstructured approach.

In the authors' experience, it turns out that
if you systematically constrain the scope of their thinking (but not too much), people are adept at fully exploring the possibilities, and they can regularly generate lots of good ideas — and occasionally some great ones. Setting the right constraints is a matter of asking the right kinds of questions: ones that create boxes that are useful, but different, from the boxes your people currently think in.
The authors argue that
The most fertile questions focus the mind on a subset of possibilities that differ markedly from those explored before, guiding people to valuable overlooked corners of the universe of possible improvements.
By way of example, the authors provide a list of "21 Questions for Developing New Products." The questions are sorted into six categories:

"De-average" buyers and users.
Which customers use or purchase our product in the most unusual way?

Do any customers need vastly more or less sales and service attention than most?

For which customers are the support costs (order entry, tracking, customer-specific design) either unusually high or unusually low?

Could we still meet the needs of a significant subset of customers if we stripped 25% of the hard or soft costs out of our product?

Who spends at least 50% of what our product costs to adapt it to their specific needs?

Explore unexpected successes.
Who uses our product in ways we never expected or intended?

Who uses our product in surprisingly large quantities?

Look beyond the boundaries of your business.
Who else is dealing with the same generic problem as we are, but for an entirely different reason? How have they addressed it?

What major breakthroughs in efficiency or effectiveness have we made in our business that could be applied in another industry?

What information about customers and product use is created as a by-product of our business that could be the key to radically improving the economics of another business?

Examine binding constraints.
What is the biggest hassle of purchasing or using our product?

What are some examples of ad hoc modifications that customers have made to our product?

For which current customers is our product least suited?

For what particular usage occasions is our product least suited?

Which customers does the industry prefer not to serve, and why?

Which customers could be major users, if only we could remove one specific barrier we've never previously considered?

Imagine perfection.
How would we do things differently if we had perfect information about our buyers, usage, distribution channels, and so on?

How would our product change if it were tailored for every customer?

Revisit the premises underlying your processes and products.
Which technologies embedded in our product have changed the most since the product was last redesigned?

Which technologies underlying our production processes have changed the most since we last rebuilt our manufacturing and distribution systems?

Which customers' needs are shifting most rapidly? What will they be in five years?

In addition to the sample questions cited above, the article also provides practical advice on how to run the brainstorming session:
  • Bound the range of acceptable ideas, then select and tailor the questions accordingly.

  • Select participants who can produce original insights (as opposed to selecting people simply on political grounds).

  • Ensure that everyone is fully engaged, e.g., by setting up a contest for the best idea or group of ideas.

  • Structure the meeting to ensure social norms work for you, not against you. Most importantly, this means using small groups for the initial discussion of your chosen questions to make it more natural for everyone to contribute actively. The authors also recommend putting the most pushy people in a single group so they do not inhibit contributions from the shyer participants (who are dispersed among the other groups).

  • Focus every discussion using your preselected questions. The authors explain that this typically results in a dynamic in which "[t]he first five minutes of each session sound like any other brainstorming meetings. But then the participants return to the better ideas and refine them. Thoughtful variants emerge. The interplay results in complex, multilayered notions that have a higher likelihood of growing into a true killer idea."

  • Do not rely solely on a single brainstorming session. A second session may be useful, there may be value in having participants do some prework, and/or it may be helpful to obtain information after the session from knowledgeable individuals who did not participate.

  • Narrow the list of ideas down right away to the ones you will seriously investigate. This is important to give participants confidence that their creative work will lead somewhere.
Since the article is informative and a model of clarity, I encourage reading it in its entirety.