Bringing Science to the Art of Strategy



Reprint: R1209C

Many managers feel doomed to trade off the futile rigor of ordinary strategic planning for the hit-or-miss creativity of the alternatives. In fact, the two can be reconciled to produce novel but realistic strategies. The key is to recognize that conventional strategic planning, for all its analysis, is not actually scientific—it lacks the careful generation and testing of hypotheses that are at the heart of the scientific method.

The authors outline a strategy-making process that combines rigor and creativity. A team begins by formulating options, or possibilities, and asks what must be true for each to succeed. Once it has listed all the conditions, it assesses their likelihood and thereby identifies the barriers to each choice. The team then tests the key barrier conditions to see which hold true.

From here, choosing a strategy is simple: The group need only review the test results and choose the possibility with the fewest serious barriers.

This is the path P&G took in the late 1990s, when it was looking to become a major global player in skin care. After testing the barrier conditions for several possibilities, it opted for a bold strategy that might never have surfaced in the traditional process: reinventing Olay as a prestigelike product also sold to mass consumers. The new Olay succeeded beyond expectations—showing what can happen when teams shift from asking “What is the right answer” and focus instead on figuring out “What are the right questions?”

Photography: Sylvain Deleu
Nuala O’Donovan,
Teasel, Exposed Interior, 2011, porcelain, unglazed, 57 x 35 x 25 cm

Strategic planners pride themselves on their rigor. Strategies are supposed to be driven by numbers and extensive analysis and uncontaminated by bias, judgment, or opinion. The larger the spreadsheets, the more confident an organization is in its process. All those numbers, all those analyses, feel scientific, and in the modern world, “scientific” equals “good.”

Yet if that’s the case, why do the operations managers in most large and midsize firms dread the annual strategic planning ritual? Why does it consume so much time and have so little impact on company actions? Talk to those managers, and you will most likely uncover a deeper frustration: the sense that strategic planning does not produce novel strategies. Instead, it perpetuates the status quo.

One common reaction is to become explicitly antiscientific—to throw off the shackles of organized number crunching and resort to off-site “ideation events” or online “jam sessions” intended to promote “out of the box” thinking. These processes may result in radical new ideas, but more likely than not, those ideas cannot be translated into strategic choices that guide productive action. As one manager put it, “There’s a reason we keep those ideas outside the box.”

Many managers feel they are doomed to weigh the futile rigor of ordinary strategic planning processes against the hit-or-miss creativity of the alternatives. We believe the two can be reconciled to produce creative but realistic strategies. The key is to recognize that conventional strategic planning is not actually scientific. Yes, the scientific method is marked by rigorous analysis, and conventional strategic planning has plenty of that. But also integral to the scientific method are the creation of novel hypotheses and the careful generation of custom-tailored tests of those hypotheses—two elements that conventional strategic planning typically lacks. It is as though modern strategic planning decided to be scientific but then chopped off essential elements of science.

The approach we’re about to describe adapts the scientific method to the needs of business strategy. Triggered by the emergence of a strategic challenge or opportunity, it starts with the formulation of well-articulated hypotheses—what we term possibilities. It then asks what would have to be true about the world for each possibility to be supported. Only then does it unleash analysts to determine which of the possibilities is most likely to succeed. In this way, our approach takes the strategy-making process from the merely rigorous (or unrealistically creative) to the truly scientific. (See the exhibit “Seven Steps to Strategy Making.”)

Step 1: Move from Issues to Choice

Conventional strategic planning is driven by the calendar and tends to focus on issues, such as declining profits or market share. As long as this is the case, the organization will fall into the trap of investigating data related to the issues rather than exploring and testing possible solutions.

A simple way to get strategists to avoid that trap is to require them to define two mutually exclusive options that could resolve the issue in question. Once you have framed the problem as a choice—any choice—your analysis and emotions will focus on what you have to do next, not on describing or analyzing the challenge. The possibilities-based approach therefore begins with the recognition that the organization must make a choice and that the choice has consequences. For the management team, this is the proverbial crossing of the Rubicon—the step that starts the strategy-making process.

Also by this author

In the late 1990s, when Procter & Gamble was contemplating becoming a major player in the global beauty care sector, it had a big issue: It lacked a credible brand in skin care, the largest and most profitable segment of the sector. All it had was Oil of Olay, a small, down-market brand with an aging consumer base. P&G crossed its Rubicon and laid out two possibilities: It could attempt to dramatically transform Oil of Olay into a worthy competitor of brands like L’Oréal, Clarins, and La Prairie, or it could spend billions of dollars to buy a major existing skin care brand. This framing helped managers internalize the magnitude of what was at stake. At that point P&G turned from contemplating an issue to facing a serious choice.

Step 2: Generate Strategic Possibilities

Having recognized that a choice needs to be made, you can now turn to the full range of possibilities you should consider. These might be versions of the options already identified. For example, P&G could try to grow Oil of Olay in its current price tier or take it upmarket, or it could seek to buy the German company that owns Nivea or pry Clinique out of the hands of Estée Lauder. Possibilities might also exist outside the initial options. For instance, P&G could extend its successful cosmetics brand, Cover Girl, into skin care and build a global brand on that platform.

Constructing strategic possibilities, especially ones that are genuinely new, is the ultimate creative act in business. No one in the rest of the beauty industry would have imagined P&G’s completely reinventing Olay and boldly going head-to-head against leading prestige brands. To generate such creative options, you need a clear idea of what constitutes a possibility. You also need an imaginative yet grounded team and a robust process for managing debate.

Desired output.

A possibility is essentially a happy story that describes how a firm might succeed. Each story lays out where the company plays in its market and how it wins there. It should have internally consistent logic, but it need not be proved at this point. As long as we can imagine that it could be valid, it makes the cut. Characterizing possibilities as stories that do not require proof helps people discuss what might be viable but does not yet exist. It is much easier to tell a story about why a possibility could make sense than to provide data on the odds that it will succeed.

A common temptation is to sketch out possibilities only at the highest level. But a motto (“Go global”) or a goal (“Be number one”) does not constitute a strategic possibility. We push teams to specify in detail the advantage they aim to achieve or leverage, the scope across which the advantage applies, and the activities throughout the value chain that would deliver the intended advantage across the targeted scope. Otherwise it is impossible to unpack the logic underlying a possibility and to subject the possibility to subsequent tests. In the Cover Girl possibility, the advantage would come from Cover Girl’s strong brand and existing consumer base combined with Procter & Gamble’s R&D and global go-to-market capabilities. The scope would be limited to the younger demographic at the heart of the current Cover Girl consumer base, and it would need to build internationally from North America, where the brand was strong. The key activities would include leveraging Cover Girl’s stable of model and celebrity endorsers.

Managers often ask, “How many possibilities should we generate?” The answer varies according to context. Some industries offer few happy stories—there are simply not a lot of good alternatives. Others, particularly ones in ferment or with numerous customer segments, have many potential directions. We find that most teams consider three to five possibilities in depth. On one aspect of this question we are adamant: The team must produce more than one possibility. Otherwise it never really started the strategy-making process, because it didn’t see itself as facing a choice. Analyzing a single possibility is not conducive to producing optimal action—or, in fact, any action at all.

We also insist that the status quo or current trajectory be among the possibilities considered. This forces the team in later stages to specify what must be true for the status quo to be viable, thereby eliminating the common implicit assumption “Worst case, we can just keep doing what we’re already doing.” The status quo is sometimes a path to decline. By including it among the possibilities, a team makes it subject to investigation and potential doubt.

The team at P&G surfaced five strategic possibilities in addition to the status quo. One was to abandon Oil of Olay and acquire a major global skin care brand. A second was to keep Oil of Olay positioned where it was, as an entry-priced, mass-market brand, and to strengthen its appeal to current older consumers by leveraging R&D capabilities to improve its wrinkle-reduction performance. A third was to take Oil of Olay into the prestige distribution channel—department stores and specialty beauty shops—as an upscale brand. A fourth was to completely reinvent Olay as a prestigelike brand that would appeal more broadly to younger women (age 35 to 50) but be sold in traditional mass channels by retail partners willing to create a “masstige” experience, with a special display section. A fifth was to extend the Cover Girl brand to skin care.

The people.

The group tasked with dreaming up strategic possibilities should represent a diversity of specialties, backgrounds, and experiences. Otherwise it is difficult to generate creative possibilities and to flesh out each one in sufficient detail. We find it useful to include individuals who did not create, and therefore are not emotionally bound to, the status quo. This usually implies that promising junior executives will participate. We also find that individuals from outside the firm, preferably outside the industry, often lend the most original ideas. Finally, we believe it’s crucial to include operations managers, not just staff members, in the process. This not only deepens practical wisdom but also builds early commitment to and knowledge of the strategy that is ultimately chosen. If you show us a company where the planners are different from the doers, we will show you a company where what gets done is different from what was planned.

Optimal group size varies among organizations and their cultures. Companies with a culture of inclusion, for example, should assemble a large group. If you go this route, use breakout groups to discuss the specific possibilities; a group larger than eight or 10 people tends to be self-censoring.

It’s usually not a good idea to have the most senior person serve as the leader; she will have a difficult time convincing the others that she is not playing her usual role as boss. Instead, choose a respected lower-level insider who is not perceived as having a strong point of view on which course should be chosen. Or tap an outside facilitator who has some experience with the firm.

The rules.

Once selected, the possibility generators must commit themselves to separating their first step—the creation of possibilities—from the subsequent steps of testing and selecting. Managers with critical minds naturally tend to greet each new idea with a long list of reasons why it won’t work. The leader must constantly remind the group that ample time for skepticism will come later; for now, it must suspend judgment. If anyone persists with a critique, the leader should require him to reframe it as a condition and table it for discussion in the next step. For example, the critique “Customers will never accept differential pricing” becomes the condition “This possibility requires that customers accept differential pricing.” It’s particularly important that the leader not shoot down possibilities early. If that happens, it’s open season on all possibilities. And removing an option about which a particular team member feels strongly may cause that person to withdraw from the process.

Many management teams try to generate strategic possibilities in a single off-site brainstorming session. Such sessions are useful, especially if they are held at an unusual location that gets people out of their accustomed routines and habits of mind. But we have also seen teams benefit from spreading the possibility-generation process over some time so that individuals have an opportunity to reflect, think creatively, and build on ideas. It is perhaps most effective to start by asking each person to spend 30 to 45 minutes sketching out three to five (or more) stories. The stories do not need to be detailed; they should truly be sketches. After this exercise the group (or breakout groups) fleshes out the initial possibilities.

Conventional strategic planning is not actually scientific. It lacks the creation of hypotheses and the careful generation of tests.

Possibility generation centers on creativity, and many techniques purport to boost creativity. We’ve found three kinds of probing questions to be especially useful. Inside-out questions start with the company’s assets and capabilities and then reason outward: What does this company do especially well that parts of the market might value and that might produce a superior wedge between buyer value and costs? Outside-in questions look for openings in the market: What are the underserved needs, what are the needs that customers find hard to express, and what gaps have competitors left? Far-outside-in questions use analogical reasoning: What would it take to be the Google, the Apple, or the Walmart of this market?

You will know that you have a good set of possibilities for further work if two things prove to be true. First, the status quo doesn’t look like a brilliant idea: At least one other possibility intrigues the group enough to make it really question the existing order. Second, at least one possibility makes most of the group uncomfortable: It is sufficiently far from the status quo that the group questions whether it would be at all doable or safe. If one or both of these don’t hold, it is probably time for another round of possibility generation.

The uncomfortable possibility for P&G was the fourth option described above. It involved transforming a weak, low-end brand into a more desirable player that could compete with upmarket department store products and then creating an entirely new masstige segment that mass retailers would enthusiastically support.

Step 3: Specify the Conditions for Success

The purpose of this step is to specify what must be true for each possibility to be a terrific choice. Note that this step is not intended for arguing about what is true. It is not intended to explore or assess the soundness of the logic behind the various possibilities or to consider data that may or may not support the logic—that comes later. Any consideration of evidence at this point detracts from the process.

The importance of this distinction cannot be overstated. When the discussion of a possibility centers on what is true, the person most skeptical about the possibility attacks it vigorously, hoping to knock it out of contention. The originator defends it, parrying arguments in order to protect its viability. Tempers rise, statements become more extreme, and relationships are strained. Meanwhile, little of either opponent’s logic is revealed to the other.

If, instead, the dialogue is about what would have to be true, then the skeptic can say, “For me to be confident in this possibility, I would have to know that consumers will embrace this sort of offering.” That is a very different sort of statement from “That will never work!” It helps the proponent understand the skeptic’s reservations and develop the proof to overcome them. It also makes the skeptic specify the exact source of the skepticism rather than issue a blanket denunciation.

We’ve developed a framework for surfacing the conditions that have to be true for a possibility to be an attractive strategy (see the exhibit “Assessing the Validity of a Strategic Option”). The conditions fall into seven categories relating to the industry, customer value, business model, and competitors. Begin by clearly spelling out the strategic possibility under consideration. Then move to a two-stage discussion process:

Generate a list.

In the first stage of discussion, the aim is to enumerate all the conditions that need to hold true for everyone in the room to be able to honestly say, “I feel confident enough to make this possibility a reality.” The conditions should be expressed as declarative rather than conditional statements—for example, “Channel partners will support us,” not “Channel partners would have to support us.” This helps paint a positive picture of the possibility, one that will be inviting to the group if the conditions actually hold.

You must make sure that the individual who proposed the possibility under review does not dominate this conversation. Any condition that is put forward should be added to the list. The person putting it forward should simply be asked to explain why that condition would be necessary for him to be confident; he should not be challenged about the truth of the condition.

When each member of the group has had a chance to add conditions to the list, the facilitator should read the list aloud and ask the group, “If all these conditions were true, would you advocate for and support this choice?” If everyone says yes, it’s time to move to the next step. If any members say no, they must be asked, “What additional condition would enable you to answer yes?” This line of questioning should continue until every member replies affirmatively.

Once again, during this step expressing opinions about whether or not conditions are true should be strictly prohibited. The point is simply to ferret out what would have to be true for every member of the group to feel cognitively and emotionally committed to each possibility under consideration.

It is important to treat the current strategy in this way as well. We recall one discussion a number of years ago about the status quo option. Toward the end, the president of the company leaped out of his seat and sprinted from the room. When he returned, 10 minutes later, his colleagues asked whether he was OK. He explained that the discussion had made him see how logically weak the status quo was. The reason he had raced out was to cancel a multimillion-dollar initiative in support of the status quo—the go/no-go deadline was that very day.

Weed the list.

The previous exercise typically overshoots, and the list of conditions crosses the line between “must have” and “nice to have.” After finishing the list of conditions, the group should take a break and then review the items, asking, “If every condition but this one held true, would you eliminate the possibility or still view it as viable?” If the answer is the former, the condition is a must-have and should be maintained. If it is the latter, it is a nice-to-have and should be removed.

The goal here is to ensure that the list of conditions is truly a binding set. To this end, once you’re finished reviewing, you should ask, “If all these conditions were true, would you advocate for and support this choice?” If any member says no, then the group needs to return to the first-stage discussion and add any necessary conditions that were initially overlooked or mistakenly removed.

After arriving at a full set of possibilities and ensuring that all must-have conditions are attached to each, the group needs to bring its options to the executives whose approval will be required to ratify the final choice and to any other colleagues who might stand in the way. For each possibility, the group needs to ask these people the same questions it asked its members: “If these conditions were shown to hold true, would you choose this possibility? If not, what additional conditions would you include?” The goal is to make sure that the conditions for each possibility are well specified in the eyes of everyone with a say in the choice— before analysis ensues.

Step 4: Identify the Barriers to Choice

Now it’s time to cast a critical eye on the conditions. The task is to assess which ones you believe are least likely to hold true. They will define the barriers to choosing that possibility.

Begin by asking group members to imagine that they could buy a guarantee that any particular condition will hold true. To which condition would they apply it? The condition they choose is, by inference, the biggest barrier to choosing the possibility under consideration. The next condition to which they would apply a guarantee is the next-biggest barrier, and so on. The ideal output is an ordered list of barriers to each possibility, two or three of which really worry the group. If there is disagreement about the ordering of particular conditions, you should rank them as equal.

Pay close attention to the member who is most skeptical that a given condition will hold true; that person represents the greatest obstacle—and, in the case of a problematic possibility, an extremely valuable obstacle—to the selection and pursuit of the option. Members must be encouraged to raise, not suppress, their concerns. Even if only one person is concerned about a given condition, the condition must be kept on the list. Otherwise he would be within his rights to dismiss the final analysis. If the skepticism of every member is drawn out and taken seriously, all will feel confident in the process and the outcomes.

When the P&G beauty care team reviewed the nine conditions it had come up with for the Olay masstige possibility, the members felt confident that six would hold: The potential consumer segment was big enough to be worth targeting; the segment was at least as structurally attractive as the current mass-market skin care segment; P&G could produce the product at a cost that would permit a somewhat lower price than those of key lower-end prestige players; it was capable of building retailer partnerships (if retailers liked the idea); prestige competitors would not copy the strategy; and mass competitors could not copy the strategy. However, three conditions worried the team, in descending order: that mass-channel consumers would accept a new, significantly higher starting price point; that mass-channel players would be game to create a new masstige segment; and that P&G could bring together prestigelike brand positioning, product packaging, and in-store promotion elements in the mass-retail channel.

Step 5: Design Tests for the Barrier Conditions

Once you’ve identified and ordered the key barrier conditions, the group must test each one to see whether it holds true. The test might involve surveying a thousand customers or speaking to a single supplier. It might entail crunching thousands of numbers or avoid any quantifiers at all. The only requirement is that the entire group believe that the test is valid and can form the basis for rejecting the possibility in question or generating commitment to it.

The member who is most skeptical about a given condition should take the lead in designing and applying the test for it. This person will typically have the highest standard of proof; if she is satisfied that the condition has passed the test, everyone else will be satisfied. The risk, of course, is that the skeptic might set an unachievable standard. In practice this does not happen, for two reasons. First, people demonstrate extreme skepticism largely because they don’t feel heard. In a typical buy-in process, concerns are treated as roadblocks to be pushed aside as quickly as is feasible. The possibilities-based approach ensures that individuals with concerns both feel and actually are heard. Second is the specter of mutually assured destruction. Though I may have serious doubts about possibility A, I quite like possibility B. You, on the other hand, have few doubts about possibility A but have serious qualms about choosing possibility B. I get to set the tests for the barrier conditions for possibility A, but I do so with the knowledge that you will be setting the tests for possibility B. If I set too high a bar, you will surely do the same. Being fair and sensible is, then, the smartest approach.

Step 6: Conduct the Tests

We typically structure this step according to what we call “the lazy man’s approach to choice,” testing conditions in the reverse order of the group’s confidence. That is, the condition the group feels is least likely to hold up is tested first. If the group’s suspicion is right, the possibility at hand can be eliminated without any further testing. If that condition passes the test, the condition with the next-lowest likelihood of confirmation is tested, and so on. Because testing is often the most expensive and time-consuming part of the process, the lazy man’s approach can save enormous resources.

At $18.99, consumers were willing to cross over from prestige department and specialty stores to buy Olay in discount, drug, and grocery stores.

Typically, at this step you bring in people from outside the strategy team—consultants or experts in relevant functional or geographic units, who can help fine-tune and conduct the tests you have prioritized. It is important to ensure that they concentrate solely on testing. You are not asking them to revisit the conditions. In fact, one beauty of the possibilities-based approach is that it enables you to focus outside resources that may be costly and time-consuming.

This approach differs profoundly from the process followed by most strategy consultants, who conduct a relatively standard suite of analyses in parallel. That generates a lot of (expensive) analysis, much of which turns out to be not essential or even useful in decision making. Furthermore, depth is sacrificed for breadth: Analyses are a mile wide and an inch deep, because the cost of deep analysis across the board would be prohibitive. To generate choice and commitment, we need analysis that is an inch wide and a mile deep—targeting the concerns that could prevent the group from choosing an option and exploring those areas thoroughly enough to meet the group’s standard of proof. The possibilities-based approach permits this.

For the P&G beauty care team, the most challenging condition for the Olay masstige possibility related to pricing. The test of the condition showcased the ability of a truly scientific, hypothesis-driven approach to generate strategies that are both unexpected and successful. Joe Listro, Olay’s R&D manager, explains how it went. “We started to test the new Olay product at premium price points of $12.99 to $18.99 and got very different results,” he says. “At $12.99, there was a positive response and a reasonably good rate of purchase intent. But most who signaled a desire to buy at $12.99 were mass shoppers. Very few department store shoppers were interested at that price point. Basically, we were trading people up from within the channel. At $15.99, purchase intent dropped dramatically. At $18.99, it went back up again—way up. So $12.99 was really good, $15.99 not so good, $18.99 great.”

The team learned that at $18.99, consumers were crossing over from prestige department and specialty stores to buy Olay in discount, drug, and grocery stores. That price point sent exactly the right message. For the department store shopper, the product was a great value but still credibly expensive. For the mass shopper, the premium price signified that the product must be considerably better than anything else on the shelf. In contrast, $15.99 was in no-man’s land—for a mass shopper, expensive without signaling differentiation, and for a prestige shopper, not expensive enough. These differences were quite fine; had the team not focused so carefully on building and applying robust tests for multiple price points, the findings might never have emerged.

It is important to understand that tests cannot eliminate all uncertainty. Even the best-performing possibility will entail some risk. That is why it is so crucial to set testable conditions for the status quo: The team then clearly sees that the status quo is not free of risk. Rather than compare the best-performing possibility with a nonexistent risk-free option, the team can compare the risk of the leading option with the risk of the status quo and reach a decision in that context.

Step 7: Make the Choice

In traditional strategy making, finally choosing a strategy can be difficult and acrimonious. The decision makers usually go off-site and try to frame their binders of much-discussed market research as strategic options. With the stakes high and the logic for each option never clearly articulated, such meetings often end up as negotiations between powerful executives with strong preconceptions. And once the meetings are concluded, those who are skeptical of the decision begin to undermine it.

With the possibilities-based approach, the choice-making step becomes simple, even anticlimactic. The group needs only to review the analytical test results and choose the possibility that faces the fewest serious barriers.

Often a strategy chosen in this way is surprisingly bold and would most likely have been strangled at birth in the traditional process. Consider the Olay case. P&G ended up deciding to launch an upmarket product called Olay Total Effects for $18.99. In other words, the brand once dismissed as “Oil for Old Ladies” was transformed into a prestigelike product line at a price point close to that of department store brands. And it worked. Mass-retail partners loved the product and saw new shoppers buying at new price points in their stores. Beauty magazine editors and dermatologists saw real value in the well-priced, effective product line.

The masstige strategy succeeded beyond expectations. P&G would have been happy with a billion-dollar global skin care brand. But in less than a decade the Olay brand surpassed $2.5 billion in annual sales by spawning a series of “boutique” product lines—starting with Total Effects and following with Regenerist, Definity, and Pro-X—that attracted more prestige shoppers and commanded prices eventually exceeding $50. Laid out neatly on paper, the possibilities-based approach sounds easy. But many managers struggle with it—not because the mechanics are hard, but because the approach requires at least three fundamental shifts in mind-set. First, in the early steps, they must avoid asking “What should we do?” and instead ask “What might we do?” Managers, especially those who pride themselves on being decisive, jump naturally to the former question and get restless when tackling the latter.

Second, in the middle steps, managers must shift from asking “What do I believe?” to asking “What would I have to believe?” This requires a manager to imagine that each possibility, including ones he does not like, is a great idea, and such a mind-set does not come naturally to most people. It’s needed, however, to identify the right tests for a possibility.

Finally, by focusing a team on pinpointing the critical conditions and tests, the possibilities-based approach forces managers to move away from asking “What is the right answer?” and concentrate instead on “What are the right questions? What specifically must we know in order to make a good decision?” In our experience, most managers are better at advocacy of their own views than at inquiry, especially about others’ views. The possibilities-based approach relies on and fosters a team’s ability to inquire. And genuine inquiry must lie at the heart of any process that aims to be scientific.

A version of this article appeared in the September 2012 issue of Harvard Business Review.


Source link

Still have questions?
We will help to answer them