People decisions made in hallways quietly cost a company its best people. According to Kamyar Shah, the fix is to run people operations as a visible system. That means published pay ranges, defensible equity data, work sample hiring, opportunities posted before they are assigned, and roles reviewed against the work people actually do. Structure, not instinct, is what keeps talent.
A company rarely loses its best people to a competitor first. It loses them to its own system. Pay bands lag the work. Opportunities get handed out by proximity. The role a person performs outgrows the title on file. None of this looks like a decision on any given day, which is exactly why it goes unmanaged. Treat people operations as structure, not instinct.
The damage happens in the hallway. Raises follow who negotiated hardest rather than skill, scope, and tenure. Stretch assignments go to whoever stood near the manager. Resume screens reward whoever wrote to the keywords. Interns get graded on attendance while the standout goes unnoticed. Each case is the same anti-pattern, which is a consequential decision made in private, on instinct, with no rule behind it. Instinct is not a policy.
Fairness cannot be produced by asking managers to be fair. Exhortation is not a system. The calm move is to change what gets decided in the open versus behind a door. Diagnose where a private judgment is doing the work a published rule should do. Then move the decision into the light. Design the rule before the next vacancy forces an improvised one.
People operations resolves into a set of visible rules, each one a standard operating procedure that removes a hidden judgment. Shah installs these inside fractional COO engagements, because a company without documented people rules does not have a culture. It has a set of habits nobody chose. Five rules carry most of the weight, and each one replaces a hallway decision with a standard.
Published pay ranges begin with rationale, not numbers. People accept a range when they understand how it was built. The inputs are market data, internal compensation bands, and the factors that move someone within a band, meaning skill, scope, and tenure rather than who pushed hardest. Silence breeds worst case assumptions, so publish where each role sits and name what a range does not promise.
The step that decides whether a rollout succeeds happens before any number goes public. Every manager receives a one page rationale and a short script, so the hard questions get answered in the room rather than in the rumor mill. A range shows the path, not a guaranteed raise, and saying so plainly prevents the resentment that follows unmet expectations. Trust comes from consistency between what a company says and what it actually pays.
Pay equity rests on data a company can defend, not on good intentions. The instrument is a compensation audit that compares pay across role, tenure, and demographics, then closes the gaps that have no business justification. Pay transparency does the preventive work, because published bands remove the secret negotiation that quietly punishes people who do not push hard. Two people doing the same job land in the same band.
Free 20-Minute Operations Review
Dealing with a specific operational bottleneck? Kamyar Shah works with founders and CEOs to identify the root cause and build a fix.
Two habits keep equity from drifting back. Separate starting pay from salary history, because anchoring a new offer to what someone earned before simply carries old inequities forward. Then review the compensation data on a fixed schedule rather than when someone complains. Pay drifts as markets move and exceptions accumulate, and the only way to catch the drift is to look on purpose, every cycle. Equity is a maintained state, not a one time cleanup.
The resume screen predicts almost nothing, and it is getting worse. Candidates write to keywords, recruiters read for keywords, and generated resumes now erase whatever signal the format still carried. The stronger method is work sample evaluation, a small version of the real job watched under real conditions. An hour of that reveals more about capability than a week of paperwork.
The same logic sorts interns and early hires. Instead of grading attendance and task completion, hand each person a real problem the team is stuck on and give them a week. Watch who maps the problem before solving it, who asks sharper questions, and who follows through without being managed. The standout is rarely the one with the cleanest deliverable. It is the one who creates clarity for others under pressure, and a structured work sample surfaces that person fast.
Proximity bias rewards whoever is physically present. The people in the room collect the offhand praise and the stretch assignments that never reach a posting. Remote staff miss all of it through no fault of their own, and over time the gap compounds into unequal promotion and pay. The cause is not bad intent. It is that too many growth decisions get made informally, in a hallway a remote worker never walks through.
The correction is procedural, not motivational. Every growth opportunity gets documented and posted in a shared channel before it is assigned, which forces managers to state what the opportunity actually requires. Recognition follows the same rule and goes in writing where everyone reads it, not into the air of a conference room. Internal mobility becomes a visible pipeline rather than a private handoff. Make the pipeline visible, and the advantage of standing near the manager shrinks fast.
Roles outgrow their descriptions quietly, most often in operations. The person hired to run a twenty person process is, two years later, managing vendors and owning numbers that once sat with the founder. The pay band never moved, because no single day felt like a promotion. Automation accelerates the drift, because when routine analysis gets handled by software, the human work shifts up toward judgment and coordination, which is worth more than the old band assumes.
The bill usually arrives as an exit. Someone leaves, the company tries to backfill the written description, and no candidate at the old price can do what the last person quietly grew into. The real cost is not only replacement. It is that the most capable people notice the gap before the compensation system does, and they leave for an employer who priced the current job. Review roles against the work people actually perform at least twice a year, and start the leveling conversation when the work changes, not at the annual review.
A simple structure prevents most of this drift. Job architecture, the map of levels and the responsibilities each one carries, gives every role a defined band and a defined scope. When the work climbs past the scope, the map shows it, and the leveling conversation has a reference point instead of a debate. Span of control belongs on the same map, because a manager quietly absorbing twelve direct reports is another role that outgrew its description without anyone deciding it should.
None of these rules run themselves. They depend on managers who can explain a pay band, run a work sample, and post an opportunity instead of handing it to a favorite. That is why the first investment is manager training, not another policy document. A rule a manager cannot explain fails in the hallway, which is the exact place these systems exist to protect. Equip the managers, then hold them to the standard.
The math makes the case plain. Replacing a capable operations lead can cost well over half of annual salary once recruiting, lost productivity, and ramp time are counted. The institutional knowledge that walks out the door is harder still to price. A visible people system is cheaper than that outcome. It keeps the strong performer by showing a fair band, a real path, and recognition that does not depend on being in the room.
Cadence is what separates a system from a slogan. Pay bands get reviewed on a schedule, roles get reviewed twice a year, and opportunities get posted every time, without exception. A rule applied only when convenient is not a rule, and people learn quickly which is which. Consistency, repeated over cycles, is what compounds into trust, and trust is what keeps a team through the seasons when a competitor comes calling.
These rules are not bureaucracy. Each one protects a person from a careless decision about their livelihood, which is the practical meaning of servant leadership. Structure, applied to people, is empathy at scale, because it replaces favoritism with a standard everyone can see. Shared, visible rules align a team around the same reality and let managers lead without guessing. The company that documents its people decisions is choosing to protect its human capital on purpose.
One more principle underwrites all of it. A people system is only as strong as its weakest unwritten rule, so anything that affects pay, advancement, or belonging gets documented and applied the same way for everyone. Exceptions are logged with a reason, not made in silence. The company that writes its rules down, and follows them when it is inconvenient, is the one people trust enough to stay with.
The pattern holds across mid-market companies. The ones that make pay, hiring, and advancement visible keep their strongest people longer, because those people can see a fair path and a rising ceiling. The ones that leave it to instinct lose talent they never meant to lose, and pay the turnover cost twice, once to replace and once in lost knowledge. Read together, these rules are one principle: talent looks unreliable only when the system around it is. Build the system, refine it every cycle, and the people who could leave find a reason to stay. Shah connects these mechanics to the wider operating model in his business consulting work.
The same accountability structures extend outside the payroll to vendors, detailed in the guide on managing 3PLs and suppliers.
