A Refresher on Discovery-Driven Planning22 February, 2017 / Articles
You’re working on a new venture and you know you’ve got to create a plan to execute it. So you look at past projects, gather and analyze relevant market data, make predictions about how much revenue you’ll be able to generate, decide what resources you’ll need, and set milestones to reach your targets. Right? Not so fast. That process might work for conventional or ongoing business lines, but new ventures, which are less predictable, require a different set of planning and control tools. That’s where the discovery-driven planning (DDP) process comes in.
I talked with Rita McGrath, a professor at Columbia Business School, who together with Ian MacMillan, of the University of Pennsylvania’s business school, developed this classic methodology for planning innovation. Their goal was to help entrepreneurs and those inside established companies adopt a new approach, “one better suited to high-potential projects whose prospects are uncertain at the start.” Discovery-driven planning has since become a staple in business schools’ entrepreneurship curricula and a go-to technique for those who manage innovation.
Where Did Discovery-Driven Planning Originate?
As McGrath and MacMillan explained in a 2014 article, the idea started in the mid-1990s while reviewing the projects Rita tracks in her “flops” file — her collection of failed growth projects that had lost their parent company at least $50 million. In an effort to understand those failed innovations (perfume from the people who make cheap plastic pens and vegetable-flavored Jello are some examples they cite), they saw a few patterns: “Untested assumptions, taken as facts. Linear plans. Too much funding up front. Little opportunity to redirect when new information was found. And, often, senior executives so besotted with the project that they simply refused to take in information that might have called its direction into question,” McGrath explains.
In short, too many firms used conventional planning to manage their new ventures. To help managers avoid these huge mistakes, McGrath and MacMillan came up with a disciplined process to systematically uncover, test, and (if necessary) revise the assumptions behind a venture’s plan. They called this new approach “discovery-driven planning” and introduced it in their 1995 HBR article of the same name.
Since then, they have taught the concepts to thousands of students and managers. They’ve also written about it in their coauthored book, Discovery-Driven Growth, and in McGrath’s latest book, The End of Competitive Advantage.
What Is Discovery Driven-Planning?
It’s a technique that any manager can use when developing and launching a new venture. As McGrath and MacMillan explained in their original article, “conventional planning operates on the premise that managers can extrapolate future results from a well-understood and predictable platform of past experience.” But new ventures are uncertain from the start. The assumptions you make at the outset aren’t likely to hold up as new information emerges, requiring substantial adjustments to the plan along the way. “In today’s popular ‘lean’ terminology, these adjustments are called ‘pivots,’” says McGrath.
Discovery-driven planning offers a lower-risk way to move a product forward in the face of “what is unknown, uncertain, and not yet obvious to the competition” so that firms can “learn as much as possible as cheaply as possible” while pursuing new ventures.
Ultimately, discovery-driven planning is a set of disciplines and tools that includes the following five steps.
Step 1. Define success. Before the venture is launched, decide what success will look like in concrete terms. You start by creating a “reverse income statement” for the project (you can see an example in the original article). Instead of estimating the venture’s revenues and then assuming profits will come, you determine the profit margin required, which should be at least 10%. You then calculate the revenues needed to deliver that profit.
The key here, says McGrath, is to have a “clear frame” for the venture and a “specific profit model.” You’re trying to answer the question: What would make the venture worthwhile for you (if you’re an entrepreneur) or your company? The reverse income statement forces you to articulate what success would look like and “allows you to see if you’re getting off track really early.”
There are cases where success can’t be measured in financial terms. McGrath says you should still articulate success “in terms of things like number of users, size of network, uptake of a new solution, and so on,” and link these outcomes to what might drive them. Ask yourself “what would need to be true to achieve the outcomes you are seeking.”
Step 2: Do benchmarking. The next step is to “figure out how realistic your reverse income statement is,” says McGrath. Here you “benchmark the key revenue and cost metrics in your business against the market and against firms offering the most-comparable products.” This will help you quickly assess whether you’re being realistic.
McGrath recalls the painful experience of working with a big chemical company that wanted to diversify its business. One of the growth vectors they identified was moving into the production of high-end apparel that would be sold in department stores. “They had had some success trademarking material that went into clothing, so they thought they might move up the value chain,” she says.
The team had specified what size they anticipated the market would be by year five, a target number spelled out in the spreadsheets at the back of the plan. But what they found after McGrath helped them do some market sizing and benchmarking “should have killed the project immediately,” she says. “One out of every six garments sold in the U.S. would have had to be produced by this company.” The project leader was reluctant to stop the project since they’d already put $10 million into it. “He basically argued that advantages in rapid product launch cycles and superior materials technology would ensure that it had a competitive advantage, even in light of the clearly daunting challenge in an industry new to the firm.”
The company launched the venture and “after three months of racking up even more losses, they declared it a flop and withdrew from the market,” McGrath explains. “The sad thing about this story is that the company could have learned that the business was unrealistic and stopped it long before spending what they did.”
Step 3: Define operational requirements. Next, McGrath says, “you have to think very critically about what has to be true” to realize the profit goals. Lay out all the activities required to produce, sell, and deliver the new product or service to customers. How many salespeople do you need? How many calls do they have to make? How many sales do they have to close and in what time period? McGrath and MacMillan call the investments to perform these activities the venture’s “allowable costs.”
“Mac always liked to ask his students to specify how they were going to get their ‘first five sales,’ rather than put grandiose projected revenue numbers in their spreadsheets,” they write. Once you’ve determined these costs, you subtract them from the required revenues and see whether the venture will deliver significant returns. McGrath warns that this is a step that people “often gloss over,” but it’s critical for determining whether your new venture is worth the risk.
Step 4: Document assumptions. This step is an essential difference between conventional and discovery-driven planning; it’s also “where a lot of companies go wrong,” according to McGrath. With new ventures, leaders often don’t see that they are basing decisions on big assumptions — it’s a “huge learning disability” for most companies. To avoid falling into this trap, get everyone together who is working on the venture and list all of the assumptions behind your profit, revenue, and allowable costs calculations. In the beginning, stick to a few of the most critical assumptions, such as what customer problem or need your project addresses. As the cost and risk of the project increase, also increase the thoroughness of your plans to test assumptions. Identifying an assumption now that turns out to be false can save you a lot of pain (and money) later on.
Step 5: Plan to key checkpoints. Now it’s time to lay out a plan. But don’t create a document that covers today through to launch — “plan only as far out as you have knowledge,” McGrath says. Identify a series of checkpoints (which she and MacMillan originally called “milestones” but have since renamed) at which you’ll determine whether your assumptions are holding true or need to be redefined. These should be points in time right before your company decides to invest more time and money, so that you can either stop the project or redirect it based on what you’ve learned so far. If your assumptions need to be adjusted, update your reverse income statement and operational requirements.
What Common Mistakes Do People Make When Using Discovery-Driven Planning?
Since people have been using the technique for over two decades, I asked Rita about common mistakes she sees.
The first one is the well-intentioned instinct for “acolytes” of the technique to apply it to all sorts of projects or issues. But, she says, it’s “not applicable to everything. You can’t use one tool to handle all problems. For example, I wouldn’t want to build a $2 billion semiconductor plant with DDP,” she says. DDP is most relevant to ventures where there is a lot of uncertainty, and you need to make a lot of assumptions and then test and convert learning quickly. At some point with a new venture, once you’ve proven its viability, it may make sense to switch to a conventional plan.
The second mistake is to go through the five steps listed above and think you’re done. Essential to the methodology is the continual updating of your assumptions and checkpoints. It’s a living plan that has to be revisited regularly.
McGrath says that managers need to adapt their mindset when using discovery-driven planning. “People are very afraid of being wrong — and you can’t blame them,” she says. Without the willingness to figure out that you were wrong about some or all of your assumptions, you can undermine the whole process, warns McGrath. “DDP becomes Kabuki theater, where you’re just going through the motions.” You should be planning to learn rather than planning to show that you’re right. That’s why she and MacMillan encourage people to be careful with their language. They prefer “assumption,” “guess,” and “hypothesis,” which show that you’re still learning, rather than “projection,” “target,” or “goal,” which “suggest that you’ve learned all that you need to and can move on.”
How Has Discovery-Driven Planning Evolved Since Its Inception?
The technique has proved to be “remarkably durable,” says McGrath, but she and MacMillan have made several enhancements, which they explained in 2014. First, they integrated the need to take a close look at future competition, both in generating assumptions and designing checkpoints that “test whether and when brand new competitors are emerging, thus better anticipating disruption.” Second, they now suggest that you create “assumptions about when competitive attacks and erosion of profits will begin, and design checkpoints as indicators that this is happening so that the next advantage stage can be launched at the optimal time.” Third, they suggest that companies more quickly stop pursuing ventures when they turn out to be flawed. “Companies should stop throwing good money and resources after bad once it has become obvious that a venture isn’t going anywhere,” she explains. Last, they’ve shortened the duration of the plan, suggesting that managers shouldn’t think past more than four checkpoints, and ask themselves, “Do we have enough money to get through the next three checkpoints?”
These last two additions are a result of how things have changed over the last 20 years. “The velocity today is so much faster than it was then,” says McGrath. Companies need to make decisions more quickly. This also means that managers can be less specific in their early estimates. More people are using ranges instead of precise numbers, and rather than assuming the exact size of an opportunity, it’s sufficient to say that they know it’s “big,” but not how big.
Discovery-driven planning may be more relevant now than it was 20 years ago. After the original idea was published, in 1995, it was picked up by Steve Blank, and then by Eric Ries, and became the foundation of the lean startup movement. “It’s really core to how we think about innovation today,” says McGrath. It’s easy to see how DDP has influenced several lean concepts, including the minimum viable product and rapid prototyping. While McGrath and MacMillan didn’t talk explicitly about either of those concepts, she says they were implicit in the iterative nature of their technique.