Most board meetings run the same way. The CEO lays out strategy. The CFO walks through the financial lens. Operating leaders explain how they will execute against it. Then, with five minutes left and half the room checking their phones, HR gets its slide: engagement is down ten percent, retention risk is up, and nobody in the room connects it back to anything that was just said.
Jackson Lynch, a four-time chief human resources officer and founder of Talent Sherpa, doesn’t think that’s a scheduling problem. He thinks it’s a diagnosis problem. Citing the same math W. Edwards Deming built modern quality management on, he puts a hard number on it: “94 percent of all of your execution problems are system problems, not people problems.” Most CEOs still file talent issues under people problems anyway, then wonder why the fix never holds.
Culture is decision residue.
Engagement Scores Are a Symptom
Ask most HR leaders why they’re investing in a new program and the answer usually stops at the program itself: we’re launching a leadership initiative, we’re improving engagement. Lynch treats that as an unfinished sentence.
Every initiative needs a second half: I’m doing this so that turnover in our most pivotal roles drops, so that time to full productivity shrinks, so that the number on the P&L moves. If a leader can’t finish that sentence within a step or two, they’re managing a cost, not a return.
I’ve watched this play out firsthand. An HR director once walked into a board meeting with real numbers instead of a mood, pulled straight from that morning’s payroll: roughly 16 percent of annual salary to replace an hourly worker, 21 percent for a professional role, well over 150 percent for senior, skilled positions. The CEO pushed back like he didn’t trust his own numbers. She held her ground, because the numbers were right.
Lynch’s read on that story surprised me. He didn’t fault the data. He faulted the frame: starting from retention still treats the problem as an HR issue with a business consequence, instead of a business issue with a talent root cause.
He compares it to a decision paper mills make without blinking, keeping a spare part on hand for a machine that takes a year to replace, so a breakdown costs two weeks instead of twelve months. Companies rarely extend that same logic to people. Lynch puts it in six words: “We have pivotal jobs without benches.” Most executives never diagnose why that gap is more dangerous than an idle machine.
Culture Is Decision Residue
Culture surveys get treated like a thermometer. Lynch treats them like a symptom chart instead. His definition is blunt: “Culture is decision residue.” It’s not the values poster in the break room. It’s the pattern left behind by the last hundred decisions a company made, whether anyone meant to send a signal or not.
He recommends a simple, uncomfortable exercise: pull your last hundred real decisions and hold them against whatever is written on the wall about who you want to be. Most leaders skip this because they already suspect what they’ll find.
Lynch has seen the gap firsthand: “I say I want collaborative people. And then I promote the non-collaborative jerk that has really, really good performance, except for all the dead bodies sitting to the left and right.”
The lesson isn’t subtle, but it’s easy to avoid. People don’t organize their effort around a mission statement. They organize it around what gets someone promoted, protected, or fired. Say collaboration matters all you want. If the person who steamrolls a team gets the corner office anyway, everyone has already learned the real rule, and no town hall language will unteach it.
Hire for the Ability to Connect Disconnected Things
Lynch has asked one interview question for thirty years, and it still throws people off. He asks what someone does for fun, then the harder follow-up: how does that make you better at the job you’re applying for? Most candidates can’t bridge the gap. A handful can, and those are the ones he hires.
He remembers one candidate, a computer scientist named Amy, who said she liked composing music. Pressed on why that mattered, she explained that writing music forced her to think about how parts build on each other in a system, and she used that same instinct when she wrote code. Lynch’s reaction was immediate: “Holy crap, Batman, you’re hired.” Thirty years later, it’s still the model for what he screens for.
That’s why he pushes back on hiring for years of experience alone. I’ve had almost the same conversation with people who tell me they’ve got thirteen years of experience in one narrow lane. What they usually have is one year of experience, repeated thirteen times. Pattern recognition built inside a single, stable environment breaks the moment that environment changes, and the only real defense is a demonstrated ability to move an idea from one domain into a completely different one.
Trade Accountability for Reliability
At times, leadership teams default to one question after something breaks: who’s accountable? Lynch thinks that question makes organizations worse, not better. He puts it plainly: “Accountability is backward-looking, and it is blame-oriented.” It hunts for a person to pin a failure on instead of asking what let the failure happen in the first place.
His alternative is reliability: what has to be true for this outcome to happen the same way every time, and what design gap let it slip this time. That question doesn’t produce a scapegoat. It produces a fix.
He borrows a tool from Cold War military planning: put a cross-functional group in a room, assume a plan has already failed twelve months from now, and ask why before it happens. Done well, it turns an uncomfortable truth into something a team can say out loud without anyone losing their job over it, which is the only way anyone flags a real problem before it becomes a crisis.
Skip that step and the meeting still happens eventually. It just happens after the damage is done, gets called an accountability review, and lands on whoever was standing closest when the system finally broke.
How to Tell the CEO What Nobody Else Will
Every senior HR leader eventually spots a problem above their pay grade: two executives who’ve stopped speaking, a board member who’s lost confidence in the CEO, a pattern everyone half sees but nobody wants to name, because naming it wrong ends careers. Lynch has a rule for that exact moment.
The rule is about positioning, not courage. He puts it this way: “It’s gotta be third-party removed. It needs to be formed as an observation, not an accusation.” An accusation forces the other person to defend themselves. An observation just asks them to notice something too. Push it at the wrong moment, or as an accusation instead, and you become part of the dysfunction you were trying to name.
His mechanism is almost boring: a five-minute block inside his one-on-ones with the CEO, reserved for things he’s watching but not yet asking anyone to act on. In one company, that meant flagging that the CFO had gone quiet in strategy meetings and started managing email instead, three sessions running. He framed it the way he coaches everyone else to: “That’s not a problem to solve. That’s an observation to deliver.” By the time that pattern needed a real decision, it wasn’t the first time the CEO had heard about it.
The Courage to Say What Everyone Already Knows
Succession plans are where this all comes together, and where it usually falls apart quietly. Most companies keep a binder with names next to every critical role, everyone nods when it comes up, and nobody opens it again until the person in that role leaves. Lynch calls that what it is: a well-intentioned fiction, because there’s no forcing function attached to it. A list of ten names nobody has actually vetted works out the same as a list of zero.
The fix isn’t a better binder. It’s a different question. Instead of asking who they’d call if this person left, Lynch asks whether the best available people are already sitting in the roles that matter most, right now, no hypothetical attached. That question has teeth, the same way naming a real turnover number has teeth instead of a vibe.
That’s the thread running through everything Lynch does: replace the comfortable version of a problem with the true one, even when the true one is harder to say out loud in a room that would rather not hear it. He calls it plainly: “It’s by having the courage to name what everyone kind of knows, but no one says out loud.”
That’s the job, not the binder, not the slide, not the survey.