by Will Larson
- pg 32: Managers should support six to eight engineers
- pg 32: Managers supporting fewer than four engineers usually act as TLMs, taking on design and implementation work.
- pg 32: Managers supporting more than eight engineers usually act as coaches and safety nets for problems.
- pg 32: Managers-of-managers should support four to six managers
- pg 33: On-call rotations want eight-engineers
- pg 33: This is for a two-tier 24/7 rotation.
- pg 33: Shared rotations are not long-term solutions because it is stressful for people to be on-call for components they're unfamiliar with.
- pg 33: Small teams (fewer than four members) are not teams
- pg 34: To reason about their deliery you must know about each on-call shift, vacation, and interruption.
- pg 35: A team is falling behind if each week their backlog is longer than it was the week before.
- pg 35: A team is treading water if the team can get critical work done, but cannot start paying down technical debt or start major new projects.
- pg 36: A team is repaying debt if they can start to pay down technical debt and are beginning to see the benefit of that repayment snowball.
- pg 36: A team is innovating when technical debit is sustainably low, morale is high, and the majority of work is satisfying new user needs.
- pg 37: When the team is falling behind, hire more people to move to treading water.
- pg 37: Avoid re-assigning other people to the team, as people are not fungible and generally folks end up in useful places.
- pg 37: When the team is treading water, consolidate efforts to finish more things, and reduce concurrent work until you move to repaying debt.
- pg 37: When the team is repaying debt, add more time, thereby allowing the value of paying down that debt to compound.
- pg 37: When the team is innovating, maintain slack to sustain quality, but ensure your team isn't doing science projects that will lead to it being defunded.
- pg 39: If most teams are falling behind, then hire onto one team until it's staffed enough to tread water, and then move onto the next.
- pg 39: Adding new members to a team requires the team to re-gel, so rapidly grow a team and then stop growing it to minimize the disruption.
- pg 40: Sustained productivity comes from high-performing teams, and disassembling such a team leads to significant loss of productivity, even if its members are retained.
- pg 40: Moving individuals across teams can reset the clock on re-gelling, and so preserving a team's gelled state requires restrained growth.
- pg 41: Teams put spare capacity to great use by improving areas within their aegis, in both incremental and novel ways.
- pg 42: Instead move scope between teams, preserving the teams themselves. This works better because it avoids re-gelling costs, and preserves system behavior.
- pg 42: Undoing this also entails less disruption than undoing a staffing change would.
- pg 44: If you're doubling every six months and it takes six to twelve months to ramp up, then untrained engineers outnumber trained ones, and each trained engineer is devoting much of their time to training efforts.
- pg 47: Most system-implemented systems are designed to support one to two orders-of-magnitude growth from the current load.
- pg 47: If your traffic doubles every six months, then your load increases by an order of magnitude every 18 months.
- pg 47: Under these conditions you'll have to re-implement your system twice every three years.
- pg 47: The real productivity killer is not system rewrites but the migrations that follow, which expand its consequences to the entire surrounding organization.
- pg 48: You only get value from projects when they finish: to make progress, above all else, you must ensure that some of your projects finish.
- pg 49: To minimize the impact of interruptions, funnel them into an increasingly small area and then automate that area as much as possible: require tickets, create chatbots to create tickets, create runbooks, etc.
- pg 50: The best solution to frequent interruptions is a culture of documentation, documentation reading, and functional documentation search.
- pg 51: Avoid the gatekeeper pattern, which create odd social dynamics and are rarely a great use of a person's time.
- pg 51: When faced with urgent project requests when underwater, learn to say no in a way that is appropriate to your company's culture.
- pg 51: Organizational debt is a sibling of technical debt, and includes things like biased interview processes and inequitable compensation.
- pg 52: The volatile subset of organizational debt that is likely to come abruptly due is called organizational risk.
- pg 52: With so much organizational risk, identify a few areas to improve, ensure you're making progress on those, and give yourself permission to do the rest poorly.
- pg 52: Stabilize team-by-team and organization-by-organization. Only delegate solvable risk; if something isn't likely to go well, you should hold the bag.
- pg 53: First, figure out what you do: examine your role in meetings, your calendar, your recurring processes, how you support individuals, your emails, and your to-do lists.
- pg 55: From your list, identify individuals who can immediately take over work. For items without someone who is ready today, identify individuals who could potentially take over.
- pg 55: For the remaining items, separate into easy and risky gaps, and write up a plan to close all of the easy gaps and one or two of the risky gaps.
- pg 62: An accumulation of changes, such as the number of trained managers in your company, is called a stock. They are the memory of changes over time.
- pg 62: A change to a stock is called a flow, and each flow is either an inflow or an outflow.
- pg 65: Product management is an iterative elimination tournament, with each round consisting of problem discovery, problem selection, and solution validation.
- pg 66: Taking the time to evaluate which problem to solve is one of the best predictors of a team's long-term performance.
- pg 66: Themes of analysis include users pain, users purpose, benchmarking, competitive advantages, competitive moats, and compounding leverage.
- pg 66: Moats represent a sustaining competitive advantage, which makes it possible for you to pursue offerings that others simply cannot.
- pg 67: Identify tasks that don't seem important enough to prioritize, but whose compounding value makes the work possible to prioritize.
- pg 67: To curate a problem portfolio, consider what's needed to survive this round, the next round, and to win a round, different time frames, industry trends, and return on investment.
- pg 68: Disagreement about what to work on is often rooted in different assumptions about what time frame to optimize for.
- pg 68: Usually people under-prioritize quick, easy wins.
- pg 68: Effective techniques include writing a customer letter, identifying prior art, finding first users, experimenting, building cheaply, and ensuring low switching costs.
- pg 69: Prefer experimentation over analysis, as it's far easier to get good at cheaply validating than it is to get great at always picking the right solution.
- pg 70: Strategies are grounded documents that explain trade-offs and actions to address a specific challenge.
- pg 70: Visions are aspirational documents that enable individuals who don't work closely together to make decisions that fit together cleanly.
- pg 71: An effective structure for good strategy has three sections: diagnosis, policies, and actions.
- pg 71: The diagnosis describes the challenge, calling out the factors and constraints that define it. It is a very thorough problem statement.
- pg 72: Policies describe the general approach you'll take to address the challenge. Bad guiding policies entrench the status quo, and good ones take a clear stance on competing goals.
- pg 72: Actions are the steps to implement policies. Good actions make you feel both uncomfortable and hopeful, while bad actions shirk changing anything.
- pg 73: An effective vision helps people think beyond the constraints of their local maxima, and lightly aligns progress without tight centralized coordination.
- pg 73: A good vision defines a vision statement, value proposition, capabilities, solved constraints, future constraints, reference materials, and a narrative.
- pg 73: The vision_statement is a one- or two-sentence aspirational statement to summarize the rest of the document.
- pg 73: The value proposition defines how you will be valuable to your users and to your company, and what kinds of success you will enable them to achieve.
- pg 74: The solved constraints explain what you're limited by today but will no longer be constrained by in the future.
- pg 74: The narrative is a one-page summary of everything that is easy to digest. This typically immediately follows the vision statement.
- pg 74: A vision is succeeding when people reference it to make their own decisions, and is struggling when decisions keep happening that don't fit its direction.
- pg 75: Have one vision for every distinct area. Having two articulated visions in one place is worse than having zero.
- pg 76: Defining goals – good goals are a composition of four specific kinds of numbers
- pg 76: A target states where you want to reach.
- pg 76: A baseline indicates where you are today.
- pg 76: A trend defines the current velocity.
- pg 76: A time frame sets bounds for the change.
- pg 76: Investments and baselines
- pg 76: Investments describe a future state that you want to reach, and baselines describe aspects of the present that you want to preserve.
- pg 77: Baseline metrics are useful for narrowing the solution space that you explore in order to accomplish your investment goals.
- pg 77: Plans and contracts
- Specify as few investment goals as possible, maybe three, and focus discussion on those. You should identify more baseline goals.
- pg 78: Metrics can be an extremely effective way to lead change with little or no organizational authority.
- pg 79: First identify where the levers for change are, then dive deep on building a mental model around them.
- pg 79: Diving deep also kicks off a relationship between you and the teams who you'll want to partner with most closely.
- pg 79: Build a system of second-degree attribution for metric performance, thereby allowing you to build data around teams using the platform.
- pg 79: Once you have attribution data, begin benchmarking, which is particularly powerful because it automatically adapts to changes in behavior.
- pg 80: Nudge teams by emailing teams whose metrics has changed recently, both in terms of absolute change and their benchmarked performance.
- pg 80: If nudges aren't effective, work with key teams to agree on baseline metrics.
- pg 80: This ensures baselines are top-of-mind and gives them a powerful tool for negotiating priorities with stakeholders.
- pg 81: The fact that something stops working at significantly increased scale is a sign that it was not originally over-designed.
- pg 82: Migrations are usually the only available avenue to make meaningful progress on technical debt.
- pg 82: Migrations occupy the awkward territory of reduced immediate contribution today in exchange for more capacity tomorrow.
- pg 83: De-risk
- pg 83: After you have buy-in on a new design, embed into the one or two most challenging teams and work with them to build and migrate to the new system.
- pg 83: If you leave one migration partially finished, then people will be exceedingly suspicious of participating in the next one.
- pg 83: Enable
- pg 83: Instead of generating tracking tickets for teams to implement, build tooling to programmatically migrate the easy 90 percent.
- pg 83: After programmatically migrating as much as possible, write out the self-service tooling and documentation for the remaining teams.
- pg 84: Finish
- pg 84: By ensuring that all newly written code uses the new approach, time becomes your friend because you are making progress by default.
- pg 84: Generate tracking tickets for teams that need to migrate.
- pg 84: If a team isn't working on a migration, it's typically because their leadership has not prioritized it.
- pg 84: And for the remaining work, finish it yourself, and reserve celebration and recognition for when it is completed successfully.
- pg 85: The best kind of re-org solves a structural problem. The worst kind of re-org is done to avoid a people management issue.
- pg 87: Only re-org if the problem it will solve already exists, because it's remarkably hard to predict future problems.
- pg 87: Use the historical hiring rate to project this.
- pg 89: Ensure that you can write a crisp mission statement for each team, define their areas of ownership, and define the interfaces between them.
- pg 89: Put teams that work together (especially poorly) as close together as possible. This close proximity helps fill harmful information gaps between them.
- pg 89: Avoid implicitly creating holes of ownership, instead explicitly defining such holes (i.e. defining an unstaffed team).
- pg 90: Accidentally missing someone is the cardinal sin of a reorganization.
- pg 91: Organizational change is very resistant to rollback, and so you must be collectively committed to moving forward with it in the face of adversity.
- pg 91: A good roll-out requires explaining the reasoning driving the re-org, documenting how each person and team is impacted, and availability and empathy to impacted individuals.
- pg 91: To execute, discuss with heavily impacted individuals in private, prepare managers and key individuals, send out an email, be available for discussion, and have more skip-level one-on-ones.
- pg 92: Controls are the mechanisms that you use to align with other leaders you work.
- pg 92: Controls include metrics, visions, strategies, organization design, roadmaps, performance reviews, and more.
- pg 93: For whatever controls you pick, you must then agree on the degree of alignment for each one.
- pg 93: Levels of alignment include do it yourself, preview, review, notes, no surprises, and let me know if something comes up.
- pg 93: If you do something yourself, it's better to be explicit and avoid confusion on responsibilities.
- pg 93: The no surprises level only requires updates to keep your mental model current. This is important when you are evaluated on your ability to stay on top of new problems.
- pg 94: Analyzing your distribution across levels can reveal whether you are micromanaging or not.
- pg 94: Most of us are always at the furthest point in our career, each change a step into the unknown, with limited visibility into upcoming opportunities.
- pg 95: Climbing the career ladder isn't bad, but it has the side-effect of funneling most folks toward a constrained pool of opportunity.
- pg 96: Managers have a strong sense of the business's needs, allowing them to find the intersection of your interests and the business's priorities.
- pg 96: Your goals, once aligned to your company's priorities, should become a central artifact for how you and your manager collaborate on your career evolution.
- pg 97: Three rules for speaking with the media:
- pg 97: Answer the question you want to be asked, re-framing a difficult or challenging question into one you're comfortable with.
- pg 97: Stay positive, by finding a positive frame and sticking to it.
- pg 97: Speak in threes, narrowing down your message to three concise points that you make your refrain.
- pg 98: One of the trickiest, but most common, leadership scenarios is leading without authority.
- pg 98: When needing to effect change, it's a hard sell to convince someone that your personal experience should invalidate their personal experience.
- pg 99: To model a behavior, just begin doing it. Frame it as a short experiment, and keep iterating on it until you're confident it works.
- pg 99: Then document the problem you set out to solve, the learning process you went through, and how another team would adopt the practice.
- pg 99: Finally share your documented approach in a short email, without asking anyone to adopt the practice or lobbying for change.
- pg 99: A mandate assumes it's better to adopt a good-enough approach quickly.
- pg 100: A model, document, and share approach assumes it's better to adopt a great approach slowly.
- pg 101: We create a centralized, accountable group whenever a problem becomes truly acute.
- pg 101: A positive freedom is the freedom to do something, such as pick a programming language that you prefer.
- pg 101: A negative freedom is the freedom from something happening to you, such as not having to support a programming language, even if someone prefers it.
- pg 102: When ownership is extremely diffuse, a group should increase net positive freedom without greatly reducing negative freedom.
- pg 102: Group design considers influence, interaction modes, size, time commitment, identity, selection process, length of term, and representation.
- pg 102: If a group is six or fewer individuals, then they are able to gel and shift a portion of their individual identities to the team.
- pg 103: The best default is fixed terms for members, while allowing current individuals to remain eligible, and without enacting term limits.
- pg 104: Failure modes for groups are being domineering, bottlenecked, status-oriented, and inert.
- pg 104: A domineering group is most common when decision-makers are abstracted from the consequences of their decisions.
- pg 104: A bottlenecked group attempts to do too much, thereby burning out its members, and slowing down folks who need their authority.
- pg 104: An inert group doesn't do much of anything; their members have not gelled or are too busy.
- pg 107: Senior leaders pierce into areas to debug problems. Instead of hiding problems, use them as an opportunity to explain your plans to address them.
- pg 108: A playbook is tie topic to business value, establish narrative, make ask, show data, present decision-making principles, present actions with timeline, and reiterate ask.
- pg 108: You must tie the topic to business value to answer "Why should anyone care?"
- pg 108: The historical narrative should be two to four sentences explaining how things are going, how we got here, and the next planned step.
- pg 108: Present enough raw data to allow people to follow your analysis. Providing only an analysis without data is evasive.
- pg 108: Explain your mental model and principles that you're applying against the diagnosis; it should be clear how your actions follow from them.
- pg 109: You have to prioritize long-term success over short-term quality, even if it's personally unrewarding to do so.
- pg 111: Finish small, leveraged things: Each thing you finish should create more bandwidth to do more work, and it's rewarding to finish things.
- pg 111: Stop doing things: If you're very underwater, alert your team and management chain that you're not doing some work. But don't drop work silently.
- pg 111: Delegate handling exceptions: By holding onto the authority to handle exceptions, you lose much of the system's leverage.
- pg 112: Decouple participation from productivity: Don't assume that attending every meeting is valuable.
- pg 112 :Hire until you are slightly ahead of growth: Having a clear organizational design allows hiring folks before their absence is crippling.
- pg 113: For a community, facilitate, don't lecture, brief presentations and long discussions, use breakout groups, bring learnings to the full group, choose known topics, encourage tenured folks to attend, provide optional pre-reads, and start by checking in.
- pg 113: Folks want to learn from each other more than they want to learn from a single presenter.
- pg 113: Successful topics are ones that people have already thought about, typically because these concepts are core to their daily work.
- pg 115: Encourage the most senior folks to attend, because there is so much for them to teach newer folks.
- pg 119: Environments that tolerate frequent exceptions are not only susceptible to bias but also inefficient.
- pg 120: Every policy you write is a small strategy, built by identifying your goals and the constraints that bring actions into alignment with those goals.
- pg 121: A bad policy is one that does little to constrain behavior.
- pg 121: Applying policy consistently is challenging because of:
- pg 121: Accepting reduced opportunity space: This is incurred by establishing good constraints.
- pg 121: Local sub-optimality: Sometimes teams must deal with challenging circumstances to support a broader goal that they experience little benefit from.
- pg 122: Granting exceptions undermines people's sense of fairness, and sets a precedent that undermines future policy.
- pg 122: Organizations spending significant time on exceptions are experiencing exception debt, and they must instead work the policy.
- pg 122: Once you've collected enough escalations, revisit your constraints, merge in challenges discovered from applying the policy, and either preserve the existing constraints or generate new ones that handle escalations more effectively.
- pg 123: Folks who communicate a no are able to convincingly explain their team's constraints and articulate why the proposed path is unattainable or undesirable.
- pg 124: When articulating your constraints, the two most frequent disagreements are over velocity and prioritization.
- pg 124: When folks want you to commit to more work than you believe you can deliver, your goal is to provide a compelling explanation of how your team finishes work.
- pg 125: Once you can explain your constraints and how time is being spent, then you're having a useful conversation about whether you can shift time from other behaviors toward your constraints.
- pg 125: The only two ways to add capacity are by moving existing resources to the team (away from their current work) or creating new resources (usually through hiring).
- pg 125: To communicate your priorities: document all your incoming tasks, develop guiding principles for selecting work, and then share subsets of selected tasks.
- pg 126: If a stakeholder has a request that is more important than your current work, shift priorities at your next planning session.
- pg 126: If you've explained your velocity and priorities but your perspective isn't resonating, then you likely have a relationship problem to address.
- pg 127: Remember that you leave a broad wake, and that your actions have a profound impact on those around you.
- pg 127: Almost every internal problem can be traced back to a missing or poor relationship, and that great relationships allow us to come together and solve almost anything.
- pg 128: Process is a tool to make it easy to collaborate, and the process that the team enjoys is usually the right process.
- pg 128: Instead of avoiding the hardest parts, double down on them. As a leader you can't run from problems, so you must engage them.
- pg 129: Do the right thing for the company, then do the right thing for the team, and then do the right thing for yourself.
- pg 129: While this puts you last, it is a reminder to "pay yourself": Give as much as you can sustainably give, and draw the line there.
- pg 130: The best theory of management evolves as it comes into contact with reality. The worst theory is to not have one at all, and the second worst is one that doesn't change.
- pg 130: At a mid-size company, parts of it are growing quickly with an emphasis on execution, and other parts have stabilized with ideas becoming the more valued currency.
- pg 131: What folks in the growth plates need is help reducing and executing the existing backlog of ideas, not adding more ideas to be evaluated.
- pg 131: It is extremely hard to do the basics well in the growth plates, because you don't have enough time to do them well.
- pg 132: All slow-growth environments used to be high-growth environments, meaning a sufficiently effective executor evolved them into a slow-growth environment.
- pg 132: Leadership is matching appropriate action to your current context, and it's uncommon that any two situations will flourish from the same behaviors.
- pg 132: New managers only manage down, only manage up, never manage up, optimize locally, assume hiring never solves problems, don't spend time building relationships, define their role too narrowly, and forget their manager is human.
- pg 133: If you never manage up, then excellent work goes unnoticed because it was never shared.
- pg 133: If your business is growing quickly, then eventually you hire or burn out.
- pg 133: Managers are the glue in their team, filling any gaps – and that may mean doing things you don't want to do in order to set a good example.
- pg 133: More experienced managers do what worked at their previous company, spend too much time building relationships, assume that more hiring can solve every problem, abscond rather than delegate, and become disconnected.
- pg 133: Pause to listen and foster awareness before "fixing" things, or else you may fix problems that don't exist with tools that may not be appropriate.
- pg 134: Managers at all levels of experience mistake team size for impact, mistake title for impact, confuse authority with truth, don't trust the team enough to delegate, let other people manage their time, and only see problems.
- pg 134: Authority lets you get away with weak arguments and poor justifications, which causes people to leave quickly.
- pg 135: To partner with your manager you must know a few things about them, they must know a few things about you, and you should occasionally update the things you know about each other.
- pg 136: Tell your manager what you're trying to solve, that you're making progress, what you prefer to work on, how busy you are, your professional goals and growth areas, and how you believe you're being measured.
- pg 136: Success hinges on finding the communication mechanism that works for your manager.
- pg 136: If a manger doesn't seem to care, it's likely they do care but are too stressed to participate in successful communication.
- pg 137: Ask your manager their current priorities, how stressed they are, what you can do to help, their management's priority, and what their goals are.
- pg 138: To grow we should pursue scope – not enumerating people, but taking responsibility for the success of increasingly important and complex facets of the organization and company.
- pg 139: This means can can always find ways to increase your scope and learning, even if your company doesn't have room for more directors or vice presidents.
- pg 140: When things go poorly, you're swamped with more direction and input than you can absorb. When things are going well, you must supply your own direction and that of your team.
- pg 140: The first step to setting direction is to cast the widest possible net for ideas.
- pg 141: From the ideas form a strategy, define its key pivots, and let that set your direction.
- pg 141: No one will have the time to soak up all the detail of your overly precise document, so distill it down to three of four bullet points.
- pg 142: Upon transitioning to management, your ability to put your head down and solve gritty, important problems is now the wrong answer to most of your problems.
- pg 142: For every problem that comes your way, either close out, solve, or delegate.
- pg 142: Closing out means making a decision and communicating it to all involved participants.
- pg 142: Solving means designing a solution such that you won't need to spend time on this particular kind of ask again in the next six months.
- pg 142: Delegate means passing it to someone who has the specialized skills to close it or solve it, or who can work in the system.
- pg 147: An inclusive organization is one in which individuals have access to opportunity and membership.
- pg 147: Opportunity is having access to professional success and development.
- pg 147: Membership is participating as a version of themselves that they feel comfortable with.
- pg 148: The most effective way to provide opportunity to members of your organization is structured application of good process.
- pg 148: To create opportunity, employ rubrics, have a structured approach to choosing leads, have explicit budgets, nudge involvement, and create education programs.
- pg 148: It's effective to reach out to people and encourage them to apply to opportunities or use available resources. Even more powerful is showing where they are on a distribution.
- pg 149: To measure opportunity, focus on retention, who leads critical projects, level distribution, and time at level.
- pg 149: Retention is the first thing you pay attention to, but it is the slowest to show change.
- pg 150: Impactful programs include recurring weekly events, employee resource groups, team offsites, coffee chats, and team lunches.
- pg 151: To measure membership, focus on retention, referral rate, attendance rates for recurring events, and the number of completed coffee chats.
- pg 151: Balancing opportunities for membership across a large population is tricky; success requires a broad portfolio of options and a willingness to balance concerns across events and time.
- pg 152: The number of people who can lead critical projects both measures the company's ability to execute and the extent that its members have access to growth.
- pg 152: For a critical project define its scope and goals with selection criteria, announce it, nudge good candidates, select a lead, find an advisor to sponsor it, notify whoever wasn't selected, and kick it off.
- pg 153: This is a great learning opportunities for the sponsors, who aren't used to teaching others how to lead large and ambiguous projects.
- pg 153: When notifying those who weren't selected, provide feedback on why they weren't selected. Saying you want to create room for another person to learn is reasonable.
- pg 154: Done well, selecting project leads like this can be the cornerstone to your efforts of creating an inclusive organization.
- pg 154: The best collaboration happens when teams had managers willing to disappoint the teams they managed in order to help their peers succeed.
- pg 155: Members of such a team have shifted their first team from the folks they support to their peers.
- The ingredients for this are:
- pg 155: Awareness of each other's work: A member cannot optimize for their team if they're not familiar with other members' work.
- pg 155: Evolution from character to person: Spending time together, often at an offsite, will transform strangers into people.
- pg 155: Refereeing deflection: If someone acts in self-interest, then other team members will do the same. A manager must referee and ensure good behavior.
- pg 155: Avoiding zero-sum culture: No one will coordinate in such cultures, where perceived success depends on capturing scarce resources like headcount.
- pg 156: The more fully you embrace optimizing for your collective peers, the closer your priorities will mirror your manager's.
- pg 156: A first team provides a community of learning, where your team's rate of learning is the sum of everyone's challenges – not just your own experiences.
- pg 157: Fair consideration doesn't mean that we prefer internal candidates, but that there is a structured way for them to apply, and for us to consider them.
- pg 157: Evaluating candidates tests partnership, execution, vision, strategy, spoken and written communication, and stakeholder management.
- pg 158: To test these categories, collect written feedback, see a 90-day plan, see a strategy/vision doc, and them have them present it to peers and present it to an executive.
- pg 158: When collecting feedback, lean into controversial feedback and listen to the concerns of would-be dissenters.
- pg 158: The 90-day plan is a great opportunity to understand their diagnosis of the situation, and having them incorporate your feedback is an opportunity to try out working together.
- pg 159: This process creates intentional practice: Folks get to take risks in their plans, and you get to give direct feedback without micromanaging.
- pg 160: Each positive freedom we enforce strips away a negative freedom, and each negative freedom we guarantee eliminates a corresponding positive freedom.
- pg 160: We ramp toward negative freedom to keep the times rolling, and ramp toward positive freedom to help adapt to new and challenging circumstances.
- pg 161: Unless people aren't trying hard, the mantra "work harder" only breeds hero programmers whose heroic ways make it difficult for non-heroes to contribute meaningfully.
- pg 161: Eventually you have a cadre of burned-out heroes, everyone else is alienated, and your project is still totally screwed.
- pg 163: You must either kill the environment that breeds the hero programmers, or kill the hero programmer by burnout.
- pg 163: Working at a frenetic pace for weeks or months may feel like a major outpouring of effort and energy, but you cannot quickly counteract a deficit created over months of poor implementation or management choices.
- pg 164: If you set the original direction and have the leverage to change directions, then stand up and taking the bullet for the fiasco you're in.
- pg 164: Without policy, step back and allow the brokenness to collapse under its own weight. Conserve your energy for the system that comes next.
- pg 170: Thriving in a company requires finding a way to succeed in each new era and successfully navigating the transitional periods.
- pg 170: Transitions allow raising the floor by building competencies in new skills, and stable periods allow raising the ceiling by developing mastery in skills that the era values.
- pg 172: To interview candidates: be kind, establish requirements, understand the signal you're looking for, come prepared, express interest in candidates, create feedback loops, and optimize.
- pg 173: You can't conduct a kind, candidate-centric interview process if you're interviewers are tightly time-constrained.
- pg 173: Unkind interviewers are either suffering from burnout from doing so many interviews, or are so busy with other work that they view interviewing as a burden rather than a contribution.
- pg 174: Each skill should be covered by two different interviews to create some redundancy in signal detection, in case one interview doesn't go cleanly.
- pg 174: You must also ensure that the interviewer and the interview format actually ecpose that signal.
- pg 175: Being unprepared is the cardinal sin of interviewing, because it shows a disinterest in the candidate's time, your team's time, and your own time.
- pg 175: Interview preparedness is much more company-dependent than it is individual-dependent.
- pg 176: Whenever you extend an offer to a candidate, have every interviewer send a note to them saying that they enjoyed the interview.
- pg 176: Since your goal is to create a consistent experience for your candidates, shadow interviews are as important for new hires who are experienced interviewing elsewhere as they are for new college grads.
- pg 177: If you own or design the interview loop, the best place to get feedback is from the candidate and from the interview debrief session.
- Referrals come with two major drawbacks:
- pg 180: Your personal network will always be quite small, especially when you consider the total candidate pool.
- pg 180: By hiring within these circles, it's easy to end up with a company whose employees think, believe, and sometimes even look similar.
- pg 181: A concise, thoughtful invitation to discuss a job opportunity is an opportunity, not an infringement, especially for those who are on career networking sites like LinkedIn.
- pg 183: When sending someone a message on LinkedIn, customization doesn't matter, because people mostly choose to respond based on their circumstances, not on the quality of your message.
- pg 184: When scheduling phone calls or coffee chats, interact with each person as if they'll be providing feedback on whether or not to hire you at your next job.
- pg 184: When you meet, explain why you're personally excited about the company and the role, how the interview process works, and then leave ample time for them to ask questions.
- pg 185: Coordinate reaching out; having multiple folks from the same company reach out around the same time can paint a picture of chaos.
- pg 185: Be cautiously concerned if an engineering manager spends more than an hour of week on sourcing (excluding follow-up chats, which take up more time).
- pg 186: The four major steps are identifying candidates, motivating them to apply, evaluating them, and then closing them on joining.
- To motivate candidates:
- pg 187: Have the people the candidate would work with spend time with them.
- pg 187: Clearly define the role. Always give an accurate description of the work, but try to find the best frame for describing that work.
- pg 188: When evaluating a candidate, one of the worst outcomes of the hiring funnel is that the candidate is no longer interested in being a part of your company.
- pg 189: Put your effort where there is the most room to improve, but understand that the reasonable upper bound for each stage of your funnel is not obvious.
- pg 189: Extend your funnel with onboard, impact, promote and retain stages:
- pg 189: To measure the effectiveness of your onboarding process, measure how long it takes new hires to reach P40 productivity.
- pg 190: To measure the impact of your hires, a good proxy is looking at the performance rating distribution for new hires based on time since hiring.
- pg 190: Measuring how long it takes individuals to get promoted is useful for understanding whether they have access to opportunity within the organization.
- pg 190: Measuring retainment of employees is a lagging indicator, since it typically takes years for folks to leave.
- pg 190: If you want to shape your company's culture, inclusion, or performance, this is your most valuable entry point.
- pg 191: Most approaches to performance management are composed of career ladders, performance designations, and performance cycles.
- pg 192: Any ladder with more then 10 individuals should probably be fully fleshed out, but smaller functions can survive with something rough.
- pg 192: Crisp level boundaries provide those on a ladder with a useful mental model of where they are on their journey, who their peers are, and whom they should view as role models.
- pg 193: If you can only invest in one component of performance management, make it the career ladders; everything else builds on this foundation.
- pg 193: It is a cause of concern and debugging if the explicit designation doesn't match the implicit signals someone has received.
- pg 195: Calibrations fall soundly in the category of things that are terrible but have no obvious replacement.
- pg 195: Useful rules for calibrations are adopt a shared quest or consistency, read and don't present, compare against the ladder and not others, and study but don't enforce the distribution.
- pg 195: Steer managers in a calibration away from anchoring on the designations they enter with, and toward shared inquiry.
- pg 195: To reduce bias, everyone read the manager review instead of allowing managers pitch their candidates to the room.
- pg 196: Waiting for performance designations to deal with performance issues is typically a sign of managerial avoidance.
- pg 196: The most important factor for effective performance cycles is forcing folks to practice.
- pg 197: You want to change the cycle at most once every second time, thereby allowing everyone to adapt fully and allowing you to observe how well changes work.
- pg 197: Designation momentum is the natural tendency of a performance process to consistently produce the same evaluations of the same people despite performance changes.
- pg 198: If you're not happy with the direction of your designation momentum, and your manager isn't prescribing a path to high performance, then you must be an active participant.
- pg 198: If your manager is pushing back on your goals' ambition, this is probably their way of saying that their peers will challenge their difficulty.
- pg 199: Tit-for-tat favor trading, via silence by participants instead of raising concerns, pushes all responsibility on the calibration referee.
- pg 199: A company experiencing frequent level expansion is a sign that progression, compensation, or recognition is overly tied to the level system.
- pg 199: Levels added at the top introduce level drift by downward pressure on existing levels, so that expectations at a given level decrease over time.
- pg 200: Time-at-level limits are typically used as a backstop for situations in which performance management seems appropriate but is not occurring.
- pg 200: A level split of the career level into two halves extends the runway of career progression, but tends to solidify the moat guarding post-career levels.
- pg 201: Retention-driven performance designations reset expectations permanently in ways that sacrifice long-term usefulness of the system in order to manage short-term difficulty.
- pg 202: Folks often look at new roles as less important, framing them as service roles to absorb work they're not interested in.
- pg 203: Moving from generalized roles toward specialists introduces more single points of failure; this is especially acute in organizations with frequent structural changes.
- pg 203: Many individuals will offload challenging, difficult, or uninteresting tasks, leading to the new roles being immediately underwater.
- pg 204: Requiring the first hires to be strong role models leads to a proliferation of requirements until it's impossible for any candidate to pass the bar.
- pg 204: If you can't find an executive sponsor, it's a sign that leadership doesn't believe the new role will have a good return on invested energy.
- pg 204: Every role to recruit for has a high fixed cost, so new roles can make it difficult for recruiters to hit the targets that their performance is measured against.
- pg 205: Don't frame roles as auxiliary support functions. You must frame the role's work without referencing other existing roles for it to succeed long-term.
- pg 205: It's not possible for a role to be valued or evaluated coherently without a thoughtful career ladder.
- pg 206: Specialized efficiency can lead to more throughput without increasing head count. This is the most important for teams where financial resources are the limit on growth.
- pg 206: With specialists, you can add exactly the capacity you are missing, which is very powerful for effectively solving constraints.
- pg 206: Creating a new role allows you to consider a new pool of candidate sin your hiring funnel.
- pg 207: Some questions to decide whether to create the new role include:
- pg 207: Do you have a plan for changing how the compnay values the work? Creating a new role won't inherently change how the company values this work.
- pg 207: Who will pay the maintenance costs for the new role? And if it's you, who will take up the torch when you leave?
- pg 207: Always create a new role if it immediately covers 20 individuals; reluctantly create one if it would cover 20 individuals within two months; and otherwise, be skeptical.
- pg 208: Do not start designing a new interview loop until you've instrumented your metrics funnel.
- pg 208: To understand the current loop's performance, look at funnel performance, employee trajectory after hiring, and perform candidate debriefs.
- pg 210: Identify skills to test for from role models and a career ladder, and then for each one design a test where the candidate shows their strength instead of tells you about it.
- pg 210: If you find it difficult to identify a useful rubric for a test, then you should look for a different test that is easier to assess.
- pg 211: Avoid designing by committee, and instead prefer a working group of one or two people that is then tested against a bunch of candidates for feedback.
- pg 211: Don't hire for potential, which is a major vector for bias and should be strictly avoided.
- pg 212: To summarize the approach: Avoid reusing stuff you know doesn't work, creatively approach the matter from first principles, and keep iterating on it.