Why "ATS optimization" is the wrong frame
The applicant tracking system is not the gatekeeper. It is the librarian. Treating it as the former produces the wrong document. Treating it as the latter inverts the entire workflow.
Read →
The Premise / Practice / Career Advancement
Career Advancement
A career is a strategic asset. The same diagnostic discipline used to test a market entry applies to it, starting with the evidence on how hiring decisions actually get made.
Capabilities
The work
Employers announce one standard and screen by another. Research by Joseph Fuller of Harvard Business School and Matt Sigelman of the Burning Glass Institute tracked companies that publicly removed degree requirements from their postings. For every hundred roles that dropped the requirement, fewer than four additional candidates without degrees were actually hired. Roughly forty-five percent of the firms that announced the change showed no meaningful difference in who they hired. The posting states an intention. The screen states a behavior. A search built on the posting is built on the wrong document.
Channel beats volume, and the evidence on that is now causal rather than folklore. LinkedIn ran randomized experiments on its own recommendation algorithm across more than twenty million of its members over five years, covering six hundred thousand new jobs. The results were published in Science by a team combining two LinkedIn researchers with academics at Harvard Business School, Stanford and MIT. Moderately weak connections produced more mobility than close ones. The relationship is not linear: past a point, weaker ties stop helping. Worth noting that the platform studied itself, which is a caveat rather than a disqualification. The practical read holds either way. The highest-yield action in a search is rarely one more application. It is a specific reconnection with the second ring of a network.
A selection process usually contains two kinds of criteria, and they behave differently. John Friedman of Brown University, with Raj Chetty and David Deming of Harvard, examined admission to highly selective colleges alongside later outcomes. Academic credentials predicted post-college success well. The three criteria that gave high-income applicants their advantage, meaning legacy preference, weight on non-academic credentials, and athletic recruitment, were uncorrelated or negatively correlated with those outcomes. So the process was not uniformly arbitrary. It mixed criteria that measured something with criteria that mostly reproduced existing advantage. Hiring runs the same way. The useful question in any given process is which of its filters are actually predictive and which are ceremony, because effort spent on the second kind returns nothing.
Where the durable premium sits is a separate question. The Oxford Martin School study of 702 occupations identified the work least susceptible to computerisation as the work requiring creative and social intelligence. Its headline estimate, that nearly half of United States employment sat in the high-risk category, has been contested since publication and later studies using task-level rather than occupation-level methods produced substantially lower figures. The task-level distinction has held up better than the number attached to it. Positioning therefore points a trajectory toward judgment, persuasion, and problems that are not yet well specified.
The posting is the one document in a hiring process that nobody is bound by. Most career advice optimizes for it anyway.
The four-phase method
Career history audit, narrative deconstruction, and target market segmentation. The deliverable is a defined addressable role universe, not a wishlist.
Resume and narrative architecture built against a screening threshold per role. Nothing ships until it would survive its own diligence.
A live job board with delivered applications, tracked outreach, and a documented audit trail. Every role logged. No duplicates unflagged.
Interview preparation against the firm's actual operating cadence, not generic drills. Negotiation strategy grounded in market comp data.
What makes this different
Every artifact is hand-built. Resumes pass an internal scrub for em-dash overuse, the marker phrases that signal generated work, and uniform bullet structures. These are the patterns recruiters who have read ten thousand applications recognize immediately.
Summaries are written in a first-person register that reads as confident and impact-led, not adjective-heavy. Bullets begin with the work, not the verb.
The standard
Every role is evaluated against a defined screening threshold before submission. The threshold is high enough to clear most enterprise screening. The rest is the human read.
The job board is the source of truth. Version-controlled, dated, reviewable. The client always knows exactly what has been done on their behalf.
Engagement model
Engagements are sold by deliverable, not by hour. Each tier has a defined timeline, a defined output, and a defined point at which both parties agree the work is complete.
A short, intensive review of positioning, market, and current materials. Deliverable: a written diagnostic memo and a recommended path forward.
End-to-end execution of the four-phase method. Live job board, delivered applications, interview preparation, and negotiation strategy.
For senior leaders managing a long-arc trajectory. Continuous market intelligence, narrative refresh, and on-call advisory.
All engagements scoped individually. A limited number are accepted per quarter.
Market intelligence
Most career advice treats the labor market as static, and most market commentary about it is assertion. What appears here is limited to what the evidence supports, which is why the list is short.
LinkedIn now averages roughly eleven thousand applications per minute, a rise of about forty-five percent in a single year, with generative tools driving much of the increase. Volume has stopped being a lever because everyone pulled it at once. The candidates clearing the funnel are applying to fewer roles with materials that could not plausibly have been generated.
LinkedIn platform figures, reported in Job seekers flood LinkedIn with 11,000 applications a minute, eWeek, 2025, which attributes the figures to reporting in The New York Times.
Resources
For candidates not yet ready for an engagement, a short list of resources published independently. These are the same building blocks used inside the practice, distilled into self-serve form.
Built against current Merit Hiring Plan guidance. Includes structure for KSAs, accomplishment statements, and the federal-format conventions that actually matter.
For senior strategy, operations, and product roles. Designed to read as a ship log rather than a deck. Impact-first. No AI tells. ATS-clean.
Knowledge, Skills, and Abilities essays. The part of the federal application most candidates either skip or do poorly. A field guide to writing them at the standard the panel scores against.
The pre-flight checklist for a federal application. Profile completeness, document hygiene, eligibility flags, and the sequencing that determines whether the application is even reviewed.
The Library
A small set of essays. Written for the candidate who would rather think the problem through than be told an answer.
The applicant tracking system is not the gatekeeper. It is the librarian. Treating it as the former produces the wrong document. Treating it as the latter inverts the entire workflow.
Read →Every resume opens with a "professional summary." Almost every one of them is wrong. A short field guide to the register that actually lands at the senior level.
Read →Most candidates read a JD as a list of requirements. A strategist reads it as a leaked planning document. The reframe changes the entire application.
Read →Related practices
