The staff engineer's path

Most engineering career advice stops at senior. This book covers the messy leap to staff: influence without authority, cross-team execution, and organizational strategy. Reily, a veteran staff engineer, gives concrete frameworks like the Three Maps, RAPID, and the 50/50 rule. Read it if you're a senior engineer eyeing staff, a new staff engineer drowning in meetings, or a manager trying to define the role.

1. 1Chapter 1

   Four archetypes: pick your flavor of staff

   Reily lays out four distinct archetypes. The Tech Lead guides a team's execution day to day. The Architect designs critical systems that outlast any single project. The Solver tackles the hardest, most ambiguous problems nobody else wants. The Right Hand acts as a trusted deputy to a senior leader, extending their reach.

   The idea

   The title is a container, not a job description. Companies hand out 'staff engineer' to people doing four fundamentally different jobs, and confusion sets in when you compare your role to someone else's. Knowing which archetype you are, or want to be, tells you what to say yes to and what to decline. It also explains why a peer's week looks nothing like yours and why that's fine.

   Try this: Write down which archetype your current work most resembles, then list three tasks from last week that don't fit it.

2. 2Chapter 2

   Three maps: read the org before you move it

   Reily hands readers three maps to reconcile the two. The Locator Map is the official answer: your box on the chart. The Topographical Map is the real terrain: who actually sways decisions, which approval paths wind through which offices. The Treasure Map marks where value and opportunity hide, the spots worth scarce attention.

   The idea

   Org charts describe reporting lines, not power. The person above you on paper may route every real decision through someone two levels down, and the highest-value work may sit in a team nobody watches. Staff engineers who map influence and opportunity first spend their political capital where it compounds. Those who don't spend it defending choices nobody made.

   Try this: Sketch the last three decisions that affected your team and trace who actually made each one.

3. 3Chapter 3

   Jake's diagram: one picture ends a year of chaos

   Jake joins a company where the payments system has fragmented into pieces nobody can count. Jake spends his first months drawing a single architectural diagram of the entire system, every path laid out in one picture. It becomes the shared reference point for all future work. Teams align on priorities. Duplicate efforts shrink. The picture does what a year of meetings never could: it gives everyone the same truth.

   The idea

   The highest-leverage first move is often documentation, not code. When a system's complexity lives only in scattered heads, every conversation starts from scratch and every team optimizes its own corner. One accurate, shared picture converts private knowledge into common ground.

   Try this: Pick the system your teams argue about most and start a one-page diagram of how it actually works.

4. 4Chapter 4

   You can do anything, but you can't do everything

   Reily names the trap in one sentence: 'You can do anything, but you can't do everything.' Time is the scarcest resource you have, and the only defense is constant, explicit trade-offs.

   The idea

   Seniority doesn't buy more hours; it buys harder choices. The engineer who treats every ask as equally urgent ends up doing shallow work everywhere and deep work nowhere. Trade-offs only hurt when you make them consciously, which is exactly why they're worth making. A visible no protects the one commitment that matters more than a dozen quiet yeses.

   Try this: List your current commitments, cut one, and tell its owner this week.

5. 5Chapter 5

   Five stages of a big project, and where you matter most

   A big project has five stages, and Reily lays them out plainly: Scoping, Design, Building, Launching, Operating. The staff engineer's role burns hottest in the first two, where ambiguity peaks and the questions are still open.

   The idea

   Ambiguity is where staff engineers earn their title. Early in a project, the questions are still open: what problem, for whom, at what scale. Answers locked in then ripple through months of building and years of operating. Showing up late means inheriting someone else's assumptions. The unglamorous whiteboard weeks matter more than the launch-day demo.

   Try this: For your next project, block calendar time for scoping before anyone writes code.

6. 6Chapter 6

   Unblock the VP: data turns a stalled decision into a one-meeting yes

   The staff engineer gathered the missing data, laid out the options with clear trade-offs, and brought them to a single meeting. Decision made. The project unblocked, and months of waiting collapsed into one well-prepared hour.

   The idea

   Stalled decisions usually aren't stubbornness; they're missing inputs. Leaders defer when the cost of choosing wrong feels higher than the cost of waiting, and only solid data plus honest trade-offs lowers that cost. The staff engineer's job is to make the decision easy to make. Do the homework nobody assigned, and the logjam breaks in an afternoon.

   Try this: Find the decision your project is waiting on and draft a one-page options sheet with trade-offs.

7. 7Chapter 7

   Your actions are amplified

   Reily states the core rule of staff-level role modeling without softening it: 'Your actions are amplified. If you cut corners, your team will cut corners. If you treat people with respect, your team will treat people with respect.'

   The idea

   Power scales your habits before it scales your intentions. A senior engineer's worst day teaches more than their best document, because people copy what they see, not what they say. This cuts both ways: the same amplification that spreads carelessness spreads standards. The discipline to behave well when tired or annoyed is part of the job description, not a bonus virtue.

   Try this: Pick one behavior you'd be happy to see copied and do it visibly every day this week.

8. 8Chapter 8

   Lend your credibility to someone else's idea

   Reily calls this lending your credibility, and it's a quiet superpower.

   The idea

   Influence is transferable. Your reputation took years to build, but you can spend it in minutes to amplify someone else's voice, and the organization gets better ideas adopted faster as a result. This is one of the few staff-level moves that costs you nothing but attention. Done consistently, it also grows the next generation of people whose ideas are worth backing.

   Try this: In your next meeting, publicly endorse a junior colleague's proposal by name.

9. 9Chapter 9

   Titles lie; scope tells the truth

   A 'Staff Engineer' at one company might be a 'Senior Principal' at another, and the same title can mean wildly different things. The real measure is scope of influence: how many people and systems your decisions touch.

   The idea

   Titles are local currency; they don't exchange at par. Two engineers with identical titles may hold tenfold different reach, and two with different titles may do the same job. Fixating on the word on your badge leads to bad moves, like taking a smaller role for a bigger-sounding title. Scope, the set of problems you're trusted with, is the honest yardstick.

   Try this: Write down the scope of your current influence in one sentence, ignoring your title entirely.

10. 10Chapter 10

    Write the job you want: the staff engineer job description

    The key components are Scope, the area of responsibility you own; Archetype, which of the four shapes the role takes; and Expectations, specific outcomes rather than activities.

    The idea

    An undefined role is a trap disguised as freedom. When expectations stay vague, you fill the vacuum with whatever lands in your inbox, and your review becomes a debate about someone else's mental model. Writing scope, archetype, and outcome-based expectations turns ambiguity into an agreement. It also gives your manager something concrete to champion when promotion season comes.

    Try this: Draft your own one-paragraph job description using scope, archetype, and three expected outcomes.

11. 11Chapter 11

    Maria's visibility: staff-level work isn't enough by itself

    So Maria started writing design docs for cross-team projects and presenting them at architecture reviews. Within a year, she was promoted.

    The idea

    Impact that nobody sees doesn't count. Promotion committees judge the evidence in front of them, and quiet excellence reads as ordinary contribution. Visibility isn't bragging; it's making your reasoning inspectable through docs and reviews where decision-makers already look. Maria's story stings because the fix took a year, but the alternative is doing staff work indefinitely at senior pay.

    Try this: Submit one design doc to a cross-team architecture review this month.

12. 12Chapter 12

    Build the promotion packet: a story with numbers

    The structure is simple: a clear problem, your specific actions, and a measurable outcome. The shape of the evidence: 'Saved $X per year,' 'Reduced latency by Y%,' 'Onboarded Z teams.'

    The idea

    Promotion is a persuasion problem with a rubric. Committees read dozens of packets, and the ones that stick follow a story arc: stakes, action, result. Numbers do the heavy lifting because they survive translation between managers and levels. If you can't attach a figure to your impact, the fix isn't better prose; it's choosing work whose results can be counted.

    Try this: Start a running doc today logging every project's problem, action, and measurable result.

13. 13Chapter 13

    Writing is your most scalable tool

    Of the three, she calls writing the most scalable.

    The idea

    A staff engineer's reach is capped by their communication, and writing is the only tool that compounds. A doc can be read by fifty people across time zones and still carry your exact reasoning. Speaking persuades in the moment; relationships open doors; but writing builds shared understanding that outlives the conversation. Engineers who underinvest in prose cap their own influence.

    Try this: Convert one recent verbal decision into a short written doc and circulate it.

14. 14Chapter 14

    The six-page doc that beat the slide deck

    The discussion stayed focused and productive, and the strategy was approved with minimal debate.

    The idea

    The medium shapes the thinking. Slides reward punchlines; narrative rewards logic, because you can't hand-wave through six connected pages. Pre-reading flips the meeting from information transfer to actual deliberation, which is the only part worth having people in a room for. The extra hours writing buy back a meeting that would have sprawled for weeks.

    Try this: Replace your next strategy deck with a six-page doc sent 48 hours before the meeting.

15. 15Chapter 15

    The 50/50 rule: half your prep is practice

    Her prescription is the 50/50 rule: spend half your preparation time on the content and half practicing the delivery.

    The idea

    Engineers treat delivery as decoration, but a talk is a performance whether you like it or not. Practicing isn't vanity; it's load-testing your argument on real breath and real timing, where weak transitions show up immediately. Splitting prep evenly feels wasteful until you watch a brilliant idea die from a mumbled opening. The audience judges the whole package, not just the slides.

    Try this: For your next presentation, rehearse out loud for as many minutes as you spent building slides.

16. 16Chapter 16

    The warm path to a VP: coffee before the ask

    Instead, the engineer found a mutual connection, secured a warm introduction, and booked a 15-minute coffee chat before making any ask. The VP approved the project the next week.

    The idea

    Big asks travel better through warm channels. A cold request forces a stranger to evaluate both your idea and your credibility at once, with no context and no goodwill. A warm introduction borrows the connector's trust, and a short informal chat lets the VP meet the person before weighing the proposal. Fifteen minutes of coffee can be worth weeks of follow-up emails.

    Try this: Before your next big ask, find one mutual connection and request an introduction.

17. 17Chapter 17

    RAPID: know who actually decides

    A staff engineer often plays the Recommend role, crafting the proposal and shaping the options.

    The idea

    Decisions fail less from bad options than from unclear ownership. RAPID separates the work of recommending, the obligation of agreeing, the labor of performing, the value of input, and the authority of deciding, so nobody assumes someone else will pull the trigger. For a staff engineer, mapping these roles before the meeting is the difference between influence and frustration. You can recommend all day; if you don't know the decider, you're shouting into a void.

    Try this: Map the RAPID roles for your team's next pending decision and confirm who Decides.

18. 18Chapter 18

    Say no to the big customer, keep the customer anyway

    The engineer said no, explaining it would add unsustainable maintenance overhead for one user's benefit. The customer stayed, and the team's roadmap stayed intact.

    The idea

    A no doesn't have to end the conversation; it can redirect it. Building one-off features for big customers feels like revenue protection but quietly taxes every future sprint with maintenance you can't shed. The stronger play is protecting the roadmap while solving the customer's actual problem with what already exists. Saying no to the feature and yes to the outcome is the whole trick.

    Try this: For the next one-off request, draft a workaround using existing capabilities before agreeing to build.

19. 19Chapter 19

    Disagree and commit is a superpower

    'Disagree and commit is a superpower.' Once a decision is made, even if you argued against it, your job is to support it fully. Undermining a decision after the fact destroys trust and slows execution.

    The idea

    Organizations can only move as fast as their losers. If every settled decision stays open for relitigation, meetings multiply and execution crawls, because nobody knows which arguments are actually closed. Committing after disagreement isn't hypocrisy; it's the price of being trusted with influence over future calls. The superpower is being safe to disagree with, because people know you'll still ship the outcome.

    Try this: Pick one decision you lost, state your disagreement once, then execute it without relitigating.

20. 20Chapter 20

    Maker time: guard two hours or lose them all

    Her countermeasure is blocking out Maker Time: uninterrupted stretches of at least 2 hours for deep work.

    The idea

    Deep work doesn't survive fragmentation. Two hours is the minimum unit because real problems need warm-up time before the thinking starts, and a meeting 40 minutes in resets the clock to zero. Staff engineers who let calendars fill reactively become full-time meeting attendees who code on weekends. The block isn't a luxury; it's where any technical judgment worth having gets formed.

    Try this: Block two 2-hour maker sessions in next week's calendar and decline anything that lands on them.

21. 21Chapter 21

    Burnout, sabbatical, and the 7 PM line

    The outcome Reily reports is the one that matters: sustained productivity without burnout.

    The idea

    Sustained output beats heroic output, and the math only works with boundaries. A year of overwork doesn't just hurt the person; it degrades the judgment and patience the role runs on. The sabbatical fixes the acute damage, but the 7 PM line and protected weekends are what prevent the relapse. Recovery without changed habits is just a delay before the next collapse.

    Try this: Set a hard evening stop time this week and honor it every night.

22. 22Chapter 22

    Your legacy is the people, not the code

    The ultimate goal of a staff engineer, she reminds readers, is to leave the organization better than you found it. 'Your legacy is not the code you wrote, but the people you grew and the systems you made more resilient.'

    The idea

    Staff-level careers are measured in what persists after you move on. Every commit will eventually be deleted, but the engineers you mentored carry your judgment into teams you'll never touch, and resilient systems absorb shocks for years. This reframes daily choices: the extra hour reviewing someone's design outlasts the extra hour of your own code. Legacy is an accumulation of small investments in people and durability.

    Try this: Name one engineer you'll actively invest in this quarter and book the first mentoring session.

Senior engineers weighing the leap to staff, and new staff engineers wondering why the job feels nothing like the title promised. Also valuable for managers defining staff roles and promotion committees deciding what staff-level impact looks like.

FAQ