Глава 1
Designing a Future Worth Choosing: The Science of Intentional Behavior Change
Ever wondered why your best intentions often fall short of actual behavior? Why do we so frequently know what we should do but fail to follow through? Matt Wallaert's "Start at the End" tackles this fundamental human challenge by revealing a systematic approach to behavior change that has transformed organizations from Microsoft to Clover Health. Unlike typical business methodology books, Wallaert's approach begins with a radical premise: start with the behavior you want to create, then work backward to design interventions that make it happen. The book has become required reading in behavioral economics programs at leading universities and has influenced product design across Silicon Valley's most innovative companies. As Arianna Huffington noted, "In a world obsessed with 'nudging,' Wallaert offers something more profound-a comprehensive framework for creating meaningful change at scale."
Глава 2
The Intervention Design Process: A New Framework for Creating Change
The Intervention Design Process (IDP) represents a fundamental shift in how we approach behavior change. Rather than starting with solutions or products, it begins with identifying a behavioral gap-the difference between what people currently do and what they could be doing. This gap-first thinking flips traditional product development on its head, forcing organizations to examine actual human behavior before crafting solutions.
The process starts by carefully documenting current behaviors through multiple research methods: observational studies, data analytics, surveys, and in-depth interviews. This multi-faceted approach ensures a comprehensive understanding of the real behavioral patterns, not just reported ones. For instance, when studying workplace productivity tools, researchers might combine usage analytics with employee interviews to understand both what people do and why they do it.
Consider the iPhone's success. While many attribute it to Apple's mission of "challenging the status quo," Wallaert argues that what truly worked were two crucial questions: "Why would people want to use phones everywhere for everything?" and "Why aren't they doing it already?" These questions reveal the two forces that shape all behavior: promoting pressures (motivations) and inhibiting pressures (barriers). Apple identified that people wanted constant connectivity and access to information (promoting pressures) but were held back by complicated interfaces and poor user experiences (inhibiting pressures).
The IDP formalizes this approach into a systematic process with distinct phases:
1. Identify a behavioral gap through rigorous research
2. Validate findings through multiple data sources and methodologies
3. Create a clear, actionable behavioral statement
4. Map the complete ecosystem of pressures affecting current behaviors
5. Design targeted interventions that address specific pressures
6. Test interventions through controlled experiments
7. Scale successful interventions while continuing to measure impact
At Microsoft's Bing, Wallaert's team demonstrated the power of this approach in education. Initial assumptions about students needing motivation to search more were proven wrong through careful validation. Instead, they discovered that teachers' concerns about inappropriate content, advertising, and classroom management were the primary barriers. The team developed a specialized version of Bing that directly addressed these inhibiting pressures: implementing locked SafeSearch, removing advertisements, and reducing data collection. This evidence-based intervention led to a 40% increase in school-based searches and influenced industry standards, compelling Google to create similar solutions.
The IDP's strength lies in its commitment to evidence over assumptions. Each step builds on validated insights rather than executive opinions or market trends. This methodical approach has been successfully applied across various sectors, from healthcare (increasing medication adherence) to finance (promoting retirement savings) to environmental conservation (encouraging recycling behaviors). By focusing on actual human behavior rather than idealized notions, organizations can create more effective interventions that drive meaningful change.
The process also emphasizes the importance of continuous measurement and iteration. Rather than treating interventions as one-time solutions, the IDP promotes ongoing monitoring of behavioral changes and adjustment of strategies based on real-world results. This adaptive approach ensures that interventions remain effective as contexts and behaviors evolve over time.
Глава 3
Behavioral Statements: The Foundation of Effective Change
Why do so many organizational visions fail to create meaningful change? According to Wallaert, it's because they lack behavioral specificity. A proper behavioral statement articulates the world we're trying to create from an explicitly behavioral perspective, following a specific structure: "When [population] wants to [motivation], and they [limitations], they will [behavior] (as measured by [data])."
This structure forces clarity about who we're targeting, why they would engage in the behavior, what conditions must exist, what measurable action we want them to take, and how we'll know if we've succeeded. The power lies in its binary nature-everything is either satisfied or not, creating clear accountability.
Microsoft's original vision statement-"a computer on every desk running Microsoft software"-illustrates both the power and potential pitfalls of behavioral statements. While revolutionary at the time, it focused on computer ownership rather than meaningful usage. This led Microsoft to prioritize sales metrics over user experience, creating feature-bloated software while neglecting everyday users. When Google Docs emerged with a usage-focused approach, Microsoft had to pivot dramatically, creating Office 365 with more accessible features.
Even worse than choosing the wrong behavior is choosing no behavior at all. Vague goals like "make customers love our product" fail because love isn't measurable or observable. Without clear metrics, you can't evaluate whether interventions are working or compare their effectiveness. This creates organizational chaos as different teams justify contradictory initiatives using the same unmeasurable goal.
Effective behavioral statements should be bold rather than timid. Uber's statement described a world where people would always use Uber when traveling from point A to point B-not sometimes, but always. This ambitious framing creates measurable outcomes and clear direction. If Apple had designed the iPod based on the Discman's limitations, it would have merely evolved music-listening rather than fundamentally transforming it.
When implemented effectively, behavioral statements create transparency and accountability throughout organizations. Each person's statement should be as large as possible while remaining something they can be individually responsible for, creating a direct connection between individual work and organizational goals.
Глава 4
Understanding Pressures: The Hidden Forces Behind Behavior
At its core, behavior change requires understanding and modifying the pressures that create distance between current reality (Point A) and desired reality (Point B). Rather than directly controlling behavior, we must influence these underlying forces. The fundamental model is simple: promoting pressures make behaviors more likely, while inhibiting pressures make them less likely. Our actual behavior results from the net balance of these competing forces.
M&M's provide a perfect illustration of competing pressures. The obvious promoting pressure is taste, which explains why Mars has created over forty flavors. But other factors also drive consumption: vibrant colors trigger evolutionary preferences for bright foods, create identity connections, and influence consumption patterns (people eat more from bowls with multiple colors). Counterintuitively, calories function as both promoting and inhibiting pressures-midday blood sugar crashes make calorie-rich snacks appealing, which Snickers has successfully leveraged in decades of advertising.
Despite these promoting pressures, we don't constantly eat M&M's because of powerful inhibiting forces. Physical availability is paramount-we eat dramatically more candy when it's within arm's reach versus across the room. Psychological availability matters too; studies show people consume significantly more candy from clear containers than opaque ones (which Google leveraged to reduce employee candy consumption by 3.1 million calories over seven weeks).
When mapping pressures, humans show predictable biases. When trying to increase behaviors, we focus almost exclusively on promoting pressures (like rewards). When trying to decrease behaviors, we emphasize inhibiting pressures (like punishments). This explains why Mars keeps creating new M&M's flavors rather than focusing on availability. The predictable error creates untapped opportunity, as Uber demonstrated by focusing on reducing inhibiting pressures (cost, wait time, payment friction) rather than enhancing promoting ones. Their success came from recognizing that the desire to travel between points was already strong-they just needed to remove barriers.
Effective pressure mapping requires examining both sides of the equation and validating pressures through research rather than making assumptions. The intervention design process demands that we draw both arrows, populate both sides with research-backed pressures, and overcome our natural biases.
Глава 5
From Ideas to Action: Designing and Selecting Interventions
Intervention design translates validated pressures into actionable solutions, but our natural bias toward novelty can derail effective behavior change. The key is to resist rushing toward creative solutions that sound good but don't address underlying pressures. Many organizations fall into the trap of implementing flashy programs or technology solutions without first understanding the core behavioral pressures they need to address. Instead, we must recognize that pressures and interventions have complex relationships-a single pressure can inspire many interventions, while one intervention might address multiple pressures simultaneously.
When designing interventions, focus on generating volume rather than immediate practicality. Research shows that the first ideas are rarely the best ones, and breakthrough solutions often emerge after generating 30-40 initial concepts. Facilitators should introduce specific techniques to maximize creative output: scheduled breaks to prevent mental fatigue, flipping behavioral statements to consider opposite approaches, examining individual pressures in isolation, and creating artificial limitations like time constraints or resource restrictions. These methods help teams push beyond obvious solutions and generate diverse intervention possibilities-all while maintaining clear connections to validated pressures.
The brainstorming phase requires suspending judgment. Teams must resist the common urge to dismiss ideas as unworkable or impractical, remembering that testing will determine effectiveness, not initial impressions. Even seemingly outlandish ideas can contain valuable elements that can be refined into practical solutions. For example, a suggestion to "pay people $1000 to exercise" might evolve into a more feasible points-based reward system.
Intervention selection is the most subjective part of the IDP, requiring careful judgment calls about which ideas to pilot. While we might refer to a complex intervention like "a flu shot clinic in a historically black church on a Sunday with a reminder about protecting the congregation" as one intervention, it actually combines multiple pressures and smaller interventions - community trust, convenience, social responsibility, and timing. The goal is achieving meaningful behavior change, not isolating which specific component drives results.
Real-world examples demonstrate the power of well-designed interventions. Meetup's counterintuitive success with adding a "pledge to create real, face-to-face community" checkbox illustrates this principle perfectly. While conventional wisdom might suggest that adding friction to the group creation process would reduce participation, this intervention actually reduced spam while increasing successful meetup creation by 16%. The checkbox worked by strengthening promoting pressures, reminding potential organizers of their authentic motivation to build community. Similarly, a hospital increased hand hygiene compliance by 45% not through additional training or reminders, but by redesigning workflow to place hand sanitizer stations directly in caregivers' natural path of movement.
The key to successful intervention design lies in maintaining a clear focus on addressing validated pressures while remaining open to unexpected solutions. Teams should prioritize interventions that are feasible to implement, scalable across the organization, and likely to create sustainable behavior change rather than temporary compliance.
Глава 6
The Ethics of Changing Behavior
Behavior change carries an inherent stigma despite being central to respected roles like teachers, doctors, and parents. This paradox stems from self-serving bias-we attribute our good actions to our character but blame our bad behaviors on environmental factors, while reversing this logic for others. For instance, we might view our healthy eating as willpower but blame poor food choices on stress, while judging others' unhealthy choices as character flaws. In reality, behavior change itself is neither ethical nor unethical-the ethics lie in what behaviors we change and how we change them.
The ethical rule for behavior change is straightforward: if your outcome behavior doesn't honor any of the population's motivations, it's unethical. Consider a workplace wellness program - it's ethical when it aligns with employees' desires for better health and work-life balance, but unethical if it solely serves to reduce insurance costs while ignoring worker preferences. For the intention-action gap (wanting to do something but failing to act), this is addressed by including motivation in the behavioral statement. For example, someone might want to exercise but struggle to start - the ethical intervention helps bridge this gap rather than forcing unwanted behavior.
The intention-goal gap (wanting an outcome but rejecting the necessary behavior) presents more complex ethical challenges. A person might want to save money but resist budgeting, or desire career advancement while avoiding networking. Ethical interventions must respect both the desired outcome and the resistance to specific methods, finding alternative pathways that align with individual values.
A complete ethical standard requires that the benefit of the behavior outweighs costs to alternative motivations, and transparency about both behavioral statements and interventions allows for ethical evaluation. Companies should push for maximum transparency through internal ethics committees, expert consultation, and public sharing of interventions. For example, Google's AI ethics board and public AI principles demonstrate this commitment to transparency.
Organizations like Uber and Facebook demonstrate the dangers of unethical behavior design. Uber's surge pricing algorithm initially encouraged drivers to operate in dangerous conditions, while Facebook's infamous emotional contagion study manipulated users' newsfeeds without consent to influence their emotional states. These ethical failures not only harm users but can cost companies billions in market value - Facebook's various privacy scandals have led to massive financial penalties and trust erosion. The ethical standard is clear: if your behavior doesn't reflect population motivations, if benefits don't outweigh costs, or if you're unwilling to publicly describe your interventions, it's unethical.
Modern examples of ethical behavior design include fitness apps that allow users to set their own goals and choose their preferred activities, or financial services that offer multiple pathways to saving money, respecting individual preferences while supporting desired outcomes. These approaches demonstrate how behavior change can be both effective and ethical when designed with user autonomy and transparency in mind.
Глава 7
Testing, Learning, and Scaling What Works
The progression from pilot to test to scale represents increasing certainty about an intervention's effectiveness, with each stage involving larger sample sizes, more durable designs, and greater organizational investment. Pilots are intentionally small-scale and "operationally dirty"-designed for minimal organizational impact and quick implementation (ideally within two weeks). This approach allows teams to rapidly test hypotheses without getting bogged down in operational complexity or requiring extensive resources.
Statistical validation is necessary because humans are neither perfectly predictable nor completely unpredictable. For a pilot, we might send 100 people a letter (treatment group) and do nothing for another 100 (control group). These initial sample sizes are chosen to balance statistical power with operational feasibility. Two key metrics emerge: effect size (how much behavior changed) and p-value (the confidence that our intervention caused the change). While academia demands p < 0.05, for behavior change work, even p = 0.2 can be acceptable at the pilot stage if the intervention has minimal downsides. This more lenient threshold acknowledges the real-world constraints and the need to balance rigor with practicality.
After test validation, we can make a "juice/squeeze statement" that follows this format: "We are [confidence] that [intervention] will [direction] [behavior] (as measured by [data]). Scaling this requires [effort] and will result in [change]." For example: "We are very confident that sending personalized flu shot letters based on member health motivations will increase the rate of getting a flu shot (as measured by flu shot claims). Scaling this requires about ten hours and $5,500 and will result in about five hundred additional flu shots." Additional examples might include: "We are moderately confident that implementing appointment reminders via text message will reduce no-show rates by 15% (as measured by appointment attendance data). Scaling requires $3,000 in setup costs and will save approximately $50,000 annually in lost revenue."
Unlike academic publishing or typical business practices, the IDP requires documenting all results-successful, marginal, or failed. This avoids the "file drawer problem" where negative or inconclusive results are hidden. Even scaled interventions eventually stop working, requiring continuous monitoring to check the health of our entire behavior-change portfolio. This monitoring might include quarterly effectiveness reviews, regular A/B testing of variations, and systematic documentation of environmental changes that might impact intervention success. Teams should maintain intervention libraries that catalog all attempts, including those that didn't work, along with hypothesized reasons for failure and potential conditions under which they might succeed.
The scaling process itself requires careful attention to maintaining intervention fidelity while adapting to larger populations. This might involve creating detailed implementation guides, training materials, and quality control measures. Teams should also establish clear criteria for when to sunset interventions that have lost effectiveness, allowing resources to be reallocated to more promising approaches.
Глава 8
Identity as the Ultimate Behavioral Driver
Identity functions as the most powerful behavioral pressure, requiring careful ethical consideration in its application. Our identities are remarkably flexible and multifaceted, continuously shifting to accommodate circumstances both consciously and unconsciously. Rather than using simplistic personas, we should view identity as hierarchical-with roles at the top representing collections of values and associated behaviors. For instance, a "parent" identity encompasses values like nurturing and protection, which manifest in behaviors ranging from preparing healthy meals to researching school districts.
Identity can be leveraged through three distinct intervention techniques: priming (activating existing identity-behavior connections), moderation (strengthening or weakening these connections), and mediation (creating new connections between identity and behavior through intermediate values). Each technique serves a specific purpose in behavior change strategies.
Priming works by activating direct connections between identity and behavior that already exist within someone's self-concept. The classic example of Asian women performing better on math tests when primed with their Asian identity versus their female identity has become less reliable as these stereotyped associations have weakened over time. However, other priming effects remain robust - such as environmental activists making more sustainable choices when their environmentalist identity is made salient, or healthcare workers showing increased hand hygiene compliance when their professional identity is emphasized. For priming to be effective, the target behavior must be directly connected to a well-defined identity someone willingly expresses and values.
Moderation changes the strength of existing connections between identity and behavior through sustained messaging and experience. The #LikeAGirl campaign by Always demonstrates this technique by weakening the association between "girl" and "terrible running." The commercial shows adult women performing stereotypical "running like a girl" movements, then contrasts this with actual girls who run confidently. Similar successful moderation campaigns include the evolution of "nerdy" from a negative to potentially positive identity, and the reframing of "gamer" identity to be more inclusive of diverse players.
Mediation creates entirely new motivational pathways by introducing a waypoint between identity and behavior. The author's mother, a nurse who initially feared computers, eventually worked in nursing informatics after hospitals created a new association: computing equals care. This wasn't priming (reminding her she was a nurse) or moderation (changing existing associations), but mediation-linking computing (behavior) to care (value) so her existing identity could incorporate the behavior. Other examples include linking environmental conservation to religious stewardship identities, or connecting exercise to professional success for business executives. Mediation is particularly powerful when the desired behavior seems initially incompatible with someone's core identity.
Understanding these three mechanisms allows for more strategic and ethical behavior change interventions that work with, rather than against, people's existing identities. The key is to recognize which approach best fits the situation and to implement it in a way that respects individual autonomy and authenticity.
Глава 9
The Power of Inhibiting Pressures: An Untapped Opportunity
Despite trying to remain neutral in earlier sections, Wallaert confesses his bias toward inhibiting pressures. They're ethically straightforward since they only affect people already motivated to act, and they possess special properties making them abnormally effective. Inhibiting pressures are often overlooked, creating fertile ground for impactful interventions in an uncrowded space-unlike the saturated world of promoting pressures where standing out requires enormous resources.
Inhibiting pressures have several unique advantages: they're homogeneous (applying universally across populations), they have longevity (remaining relevant far longer than promoting pressures), they're predictable (often coming with measurable units), and they benefit from prospect theory (Kahneman and Tversky's Nobel Prize-winning insight that equivalent losses hurt worse than equivalent gains feel good).
The "penny gap" demonstrates this-people who wouldn't pay even one cent for a t-shirt will happily take it for free, showing how completely eliminating an inhibiting pressure can trigger behavior change. When combined, these special properties make inhibiting pressures extraordinarily powerful despite a world heavily geared toward promoting pressures.
Real-world examples demonstrate this power. In Darley and Batson's famous Good Samaritan study, seminary students encountered an actor in respiratory distress while walking between buildings. Neither religious beliefs nor the topic of their upcoming speech determined whether they stopped to help. The overwhelming inhibiting pressure was time-students who felt rushed rarely stopped, with only one of ten helping in the "late" condition.
Similarly, the GetRaised.com project addressed women's pay inequality by focusing exclusively on removing inhibiting pressures that prevent women from asking for raises. The site calculates underpayment using Bureau of Labor Statistics data, recommends an optimal raise percentage, and generates a formal letter. With minimal promoting pressure (just email follow-ups), more than 80% of women who submitted the letter received raises averaging over $7,000.
These examples demonstrate that meaningful behavior change doesn't require massive resources-just focus on the right pressures. By understanding and systematically addressing the forces that shape human behavior, we can create interventions that actually work, making the world "a whole lot better" one behavior at a time.
![[start-with-why-by-simon-sinek:c0000] Start With Why p1-1](/_next/image?url=https%3A%2F%2Fd1y2du6z1jfm9e.cloudfront.net%2Fassets%2Fpodcast%2Fpurple.png&w=1024&q=50)
![[start-with-why-by-simon-sinek:c0001] Start With Why p1-1](/_next/image?url=https%3A%2F%2Fd1y2du6z1jfm9e.cloudfront.net%2Fassets%2Fpodcast%2Fgreen.png&w=1024&q=50)
![[start-with-why-by-simon-sinek:c0002] Start With Why p1-1](/_next/image?url=https%3A%2F%2Fd1y2du6z1jfm9e.cloudfront.net%2Fassets%2Fpodcast%2Fyellow.png&w=1024&q=50)
![[start-with-why-by-simon-sinek:c0003] Start With Why p1-1](/_next/image?url=https%3A%2F%2Fd1y2du6z1jfm9e.cloudfront.net%2Fassets%2Fpodcast%2Fblue.png&w=1024&q=50)
