More on this book
Community
Kindle Notes & Highlights
by
Eric Ries
Read between
November 1 - November 24, 2017
“How do you know that the work you do every day creates value for somebody else?” A huge percentage of people can’t answer that question. Think of the workers hand-wiring the twenty-nine buttons on the microwave.
PART TWO A ROAD MAP FOR TRANSFORMATION
Part Two tackles the real and difficult details of how to move an organization to a more effective, entrepreneurial way of working, and it answers three fundamental questions: What, exactly, are the systems and structures we need to implement? How, exactly, do we convince managers and employees alike to try something different than what they’ve known their whole careers? (Remember, even in a hypergrowth startup, most employees were not present for the founding of the company.) When, exactly, is a company ready to make this transformation happen?
Phase One is about laying the foundation through experimentation, adaptation, and translation. It is about preparing for the moment when decisive change becomes possible by building a critical mass of success stories and demonstrating that the new way of working is not only viable but preferable.
Phase Two, which is for rapid scaling and deployment. All the resisters and objectors come out of the woodwork. The transformation either develops its own political heft or it dies.
Phase Three: dealing with the deep systems of the corporation.
PHASES AND SCALES
1. Start Small
2. Build Dedicated, Cross-Functional Teams
Convincing leaders not only to build truly cross-functional teams but to make them fully dedicated is one of the most significant challenges I typically face when I work with companies of any size.
Susana Jurado Apruzzese, head of the innovation portfolio at Telefónica, says that one of the biggest challenges her company faces is the transfer of knowledge in innovation projects from the innovation area to the business unit for commercialization. To take the success of a project to the next step—namely, to market—the project must be transferred to sales and marketing. Jurado Apruzzese finds that including the business side on her team early on makes getting their buy-in much easier. It’s also an ideal way to make sure that sales and marketing are well versed in the knowledge around the
...more
3. Wield the Golden Sword
The team presents an offer to the senior leaders, saying here is what you get: faster cycle time, more insight into what’s happening on the ground, and a promise to solve the problem fully and control spending along the way. And here is what it will cost you: air cover, secure funding, and cross-functional collaborators. From the point of view of most executives, that’s a true bargain. Greater accountability, greater confidence that the team will deliver real results, and all for the low, low cost of some political maneuvering, which is one thing they excel at.
4. Design a Good Experiment
In order for an experiment to tell us what we need to know, i.e., whether it’s worth continuing, it needs to have certain features. Teams don’t do experiments just to see what might happen (if they did, they’d always be successful because something will always happen!). They do them to gain knowledge by measuring customer actions, not just what customers say. Every experiment should have:
A clear falsifiable hypothesis. Without a clear vision of what is supposed to happen, we can’t judge success or failure. And if you cannot fail, you cannot learn. An obvious next action. Build-measure-learn is a cycle, which means every experiment should lead directly to a follow-on action. One experiment is never enough to draw the necessary conclusions. Only a series of experiments can reveal the truth. Strict risk containment. What’s the worst that can happen? is usually a question we ask flippantly. But here we really need to know the answer—and make sure we can live with it. The goal
...more
This highlight has been truncated due to consecutive passage length restrictions.
When experimenting with business models,9 here are a few things to keep in mind: Whose balance sheet should the product be on? Does it really make sense to force a small business or a consumer to pay cash up front? Why make a distinction between product and service? If products are designed to require periodic maintenance, why not take responsibility for providing it? Should a company profit from a product that may not actually fulfill the customer’s needs? By charging only when the product performs—per-use or performance-based compensation—the company stays fully aligned with the customer’s
...more
In Lean Startup terms, we understand cycle time as build-measure-learn, which means that sometimes nontechnical parts of the cycle can be collapsed through business model changes. One product team I worked with kept running into a problem. After their product was designed and built, it took a year or more for customers to be given the chance to buy the new model. Why? Because distributors had to purchase the new product and revamp their showrooms in order to display it. Many distributors found this expensive and had little incentive to do it frequently. In markets with few distributors, there
...more
5. Create New Ways to Meas...
This highlight has been truncated due to consecutive passage length restrictions.
In this first phase of transformation, the organization is setting up cross-functional teams and doing experiments. But how do the teams know if they’re succeeding? Through the use of l...
This highlight has been truncated due to consecutive passage length restrictions.
LEADING INDICATORS Leading indicators come in many forms. Their purpose is to track signs that the process is working at the team level. The goal is to show that the probability of something good happening is increasing. For example, one executive I worked with was very focused on cycle time as a leading indicator for success. He was happy if his product teams achieved an order-of-m...
This highlight has been truncated due to consecutive passage length restrictions.
Dropbox’s two basic metrics were: Virality. “We did not want Paper to become a single-user tool. If someone was using Paper to replace [to-do list software] Evernote, we weren’t interested in that. We needed it to spread and be collaborative.” Week-two retention. “We invited someone, they tried it. Did they come back in week two?”
He says, “We used to measure too many damn things that did not have anything to do with the actual business value created.” Now they have a list of four: (1) how fast a team can get a new task done; (2) how many tasks a team can complete in the course of one regular work cycle (however long that is); (3) how long it takes for a task pulled out of the backlog lineup to get into production; and (4) how long a task has been sitting in the backlog, which includes pruning projects that have been made irrelevant by the passage of time.
“The simpler your metrics, the simpler your goals,” Smith says. “If everyone can understand it without a manual, people start getting better at a faster rate.”
Metrics are critical, Lefler says, because “when leadership can’t measure results, the common response is to require all decisions to go up to their level. The first-order effect of better measurements is better decision making. The second-order effect is that leadership gives high-performing teams the autonomy to act faster and focus their attention on the right things.”
6. Work by Exception
Every team that’s working in the Startup Way needs to have someone in company leadership they can call upon, when needed, to resolve the toughest problems they face when interacting with the wider organization. The lack of such a person can be fatal to an internal startup at worst and cause a huge waste of time at best. Without such a sponsor, the team will have to spend preciou...
This highlight has been truncated due to consecutive passage length restrictions.
7. Translate This Way of Working into Terms the Organization Can Understand
I tell companies to put someone already on their staff directly in charge of this initiative and to give that person the necessary resources. I believe this is the only way to make a change like this permanent. It has to come from within the organization and be seen as an indigenous development. It has to be designed by people who truly understand the company culture and the levers that make it work. It’s fine to have coaches help along the way, but an outsider pushing the organization to change is doomed to failure.
CHAPTER 7 PHASE TWO: SCALING UP
So what does the tipping point of the Startup Way look like? When and how do companies catalyze their early successes into scaled-up deployment?
As with Phase One, there’s no “right” way to go about Phase Two. But there are key patterns and tasks common to organizations that are working their way through this next stage of change. Review and identify challenges faced by Phase One teams and projects. Develop and roll out a widespread system for working in the new way. Identify and make proper use of executive-level champions to reinforce the new methods. Bring internal functions into the transformation process. Create an internal coaching program. Establish growth boards and begin to use metered funding for resource allocation.
As I like to say, what we’re doing is setting up a framework where skepticism will be either proven or disproven. All we really want is the truth.
If someone presented a plan like this one at a Lean Startup Meetup, they’d probably be booed off the stage. And yet, when I coach teams, “Talk to customers” is a piece of advice I almost never give. Founders are stubborn. Most either think they’ve already spent enough time speaking with customers or they’ve already decided it’s not worthwhile. Instead, I saw my job as helping the team run a good experiment that, from their point of view, would confirm their preexisting beliefs. They were already convinced that they’d sell a lot of units at the trade show, so I couldn’t get them to start with
...more
There’s another reason, though, why this style of coaching is especially important for startups. I always tell the teams I work with: I’m going to assume you’re right and I’m wrong about your plan. Let’s design our experiments to prove that.4 In addition to the learning benefits I mentioned above, this approach offers another major bonus: Sometimes the team really is right!
Traditionally, the power of managers in most organizations is measured by the number of their direct reports (or the number of people they influence via matrix management). This creates a tremendous drive for savvy managers to increase their power by arguing for ever-larger budgets and ever-more personnel. But this is a dangerous thing for an initiative that is attempting to be cross-functional. In the early days of transformation, there tends to be resistance from many functional leaders. But once the initiative has experienced its aha moment and shifts into Phase Two, many former resisters
...more
Ed Essey, principle program manager for the Microsoft Garage, notes three problems that tend to come up in this situation: (1) Coaches require a lot of encouragement and support to keep them volunteering. (2) The most motivated coaches tend to leave the company and find a place where they’ll be able to pursue the job more fully. (3) Each coach has a different skill set—marketing, or design, or technology—meaning none of them can represent the lean method fully.
“They began presenting their ‘plan’ as a series of experiments to be tested, rather than a traditional ‘execution’ plan,” Blank explains. “They were essentially acknowledging the inherent uncertainty of their proposed ideas, while simultaneously providing a plan for reducing their uncertainty through experiments.”
Techstars, a tech accelerator program, takes its coaching role so seriously that it published its own manifesto5
Be Socratic. Expect nothing in return (you’ll be delighted with what you do get back). Be authentic/practice what you preach. Be direct. Tell the truth, however hard. Listen, too. The best mentor relationships eventually become two-way. Be responsive. Adopt at least one company every single year. Experience counts. Clearly separate opinion from fact. Hold information in confidence. Clearly commit to mentor or do not. Either is fine. Know what you don’t know. Say “I don’t know” when you don’t know. “I don’t know” is preferable to bravado. Guide, don’t control. Teams must make their own
...more
If teams feel entitled to funding, it’s almost impossible to generate the energy and focus that startups require. Innovation without constraints is no blessing—startup mortality rates are unusually high for overfunded projects, with many infamous examples.
The benefits of metered funding when transposed into a corporate context are many: Scarcity mindset Changes the calculus of who’s to blame if a project fails Allows managing a set of projects as an explicit portfolio, along with portfolio metrics Greatly reduced political burden on teams Greatly enhanced focus on “what do I have to provably learn in order to unlock more funding?” More conducive to cross-functional collaboration (because everyone is paid out of a common budget) Reduces middle manager interference (because no resources are borrowed from the parent company)
As project teams produce results, “additional amounts of resources are triggered by additional evidence of social impact.” This approach differs from traditional philanthropy, which often funds nonprofits to do specific projects, but monitors activities to measure success, rather than outcomes or impact.
To follow the SafeBoda example (though the company is still in the pilot-program stage as of this writing), the company might collect data on measurable changes in head injuries as they begin to get a larger market share and others begin to copy them. “If in Stage One your money is 80 percent for learning and 20 percent for outcome, in Stage Two it’s more like 50/50,” says Zwane.
The only way to give this change staying power is to use the early successes and their attendant institutional clout to tackle the deep systems of the company—namely, its incentives structure, how people are held accountable, how resources are allocated. Think about government procurement and how deeply entrenched it is. In most organizations, these systems are considered untouchable by most employees. To make changing them a prerequisite for embarking on the transformation is a nonstarter. But change them we must.
My proposal was simple: create a one-page guidance document that laid out, in plain English, a series of parameters within which innovation teams would be pre-cleared to work. If you’re doing a tiny MVP experiment with fewer than X customers possibly affected, total liability of Y, and a cost of Z, you’re pre-approved. If the experiment is a success and you want do a “scaling MVP” of a little bit greater complexity and bigger numbers for X, Y, and Z, it’s deemed approved in advance as long as: (a) it’s built on an initial MVP and (b) you have managerial sign-off. If you want to exceed these
...more
The typical team has an idea, and they hope to come up with some great numbers to put into their business plan. In order to produce vanity metrics, they think they need to show the experiment to thousands of customers. But when they look at this sheet of guidelines and discover that doing so would require them to call legal—well, who wants to do that? They realize it’s actually easier for them to start small—say, by showing one hundred customers instead of seeking permission from legal to show ten thousand. When teams start to think this way, their behavior changes. Legal is now a part of the
...more
Eventually, the team settled upon a traditional lean technique called set-based concurrent engineering (SBCE)6 that could bring the initial MVP to market in less than eighteen months.
When you go from a system of ratings like the one GE previously had, in which employees were rated annually using one of five ranked labels,8 to a system that treats learning, honesty, and outcomes as the signs of success, it’s hard to envision how that translates into career progress.
The team chose three metrics in particular to measure: Managers’ self-reported ability to effectively plan, prepare, and differentiate for salary and bonuses 77 percent said their salary planning was either the same or simpler than it had been before PD@GE, and the findings were the same for those using ratings versus no ratings. The average merit and bonus increase both for employees participating in the no-ratings test and those who were still given a rating These remained unchanged, indicating that the new system allowed for differentiation in rewards regardless of ratings or no ratings.
...more
