Senior technology leaders learn a specific version of the executive one-on-one. The engineer who used to push back on architecture agrees quickly. The technical lead who used to bring problems is bringing status updates. Proposals that used to attract three counter-ideas from the same senior contributor come back "it looks fine."
That is a leading indicator, and by the time the enterprise notices what it is looking at, the person has often already decided.
What the silence signals
The behavior arrives before the resignation letter. Meetings that used to be conversations turn into presentations. Design reviews that used to spark disagreement pass with nods. Information the executive used to hear from that engineer surfaces from other channels now, later, with less nuance.
This is easy to mistake for passive-aggression or laziness. Usually it is neither. It is the observable output of a senior person who has decided the enterprise no longer earns the extra work. They may or may not have decided to leave. They have decided to stop investing beyond the job description. The gap between what they used to bring and what they are bringing now is where the enterprise-risk sits.
What the silence costs the enterprise
The economics on this are unforgiving. A senior engineer with tenure carries knowledge that does not sit in a document: which vendor promises turned out reliable, which junior engineers can be trusted with which kinds of ambiguity, why the last three architecture decisions went the way they did, which customer commitments assume their availability. None of that transfers cleanly.
Replacement cost goes past the recruiter fee and the sign-on bonus. Add the ramp-up period during which productivity sits materially below the departed person's, the ripple effect on peers who read the departure as a signal about the organization's direction, and the decision-quality tax of losing the person who used to catch mistakes before they shipped. The SHRM baseline for knowledge workers puts replacement cost at fifty to two hundred percent of annual compensation. For a senior specialized IC, once institutional knowledge loss is factored in, the total sits at the upper end of that range, and in vendor-critical or regulated-system roles it can climb toward two-and-a-half times. That is only if the successor is a competent replacement, which is not always available in the current labor market.
Exposure grows when the departed engineer was carrying institutional context about vendor relationships, security posture, or regulated systems. Their absence takes their productivity and, quietly, the reliability of the systems they held together.
Why "check in more" does not work
Once the executive notices the pattern, the reflex response is usually the wrong one: schedule more one-on-ones, add a directed conversation, ask how things are going.
The formalized check-in produces the formalized answer. The person who has decided to disengage is not going to spend limited discretionary energy unburdening themselves in a scheduled meeting labeled career check-in. They will say things are fine. They will describe their current sprint. They will not describe the reason the enterprise stopped earning the extra work.
The information the executive needs sits outside the agenda, in the texture of the conversations that happen unprompted. A better question than "how are you doing" is "what have you stopped doing that you used to do, and what changed."
What actually works
Better signal comes from three moves that most calendar-based check-ins skip.
The first is watching second-order signals. The engineer's outputs are one signal; the engineer's interactions with peers are another. Ask the peer group's manager whether the engineer's involvement in cross-team decisions has changed. Ask junior engineers whether their questions are being answered the way they used to be. Read the pull-request comments. Read the design docs. Silence in those places is complementary to what happens in one-on-ones, and often more honest, because the person's discretionary energy shapes it rather than a scheduled meeting. The SPACE framework (Satisfaction, Performance, Activity, Communication, Efficiency) names exactly what an executive should be watching. Activity and communication data is diagnostic in a way one-on-one content is not.
The second is getting comfortable with uncomfortable answers. The executive who visibly cannot handle the answer the reason I stopped pushing back is that pushback did not change anything is the executive who will not get that answer again. Signal follows willingness to hear. If the last three times the engineer surfaced a hard problem the executive triaged it away, the engineer has learned that the enterprise does not compensate for the extra work.
The third is treating attention as an operating discipline. Senior technology executives running organizations at scale cannot rely on the informal signals that worked at smaller size. Attention has to be systematized: a named cadence, a defined set of signals to watch, and time set aside to actually read them. Dashboards do not do this work; a recurring, uninterrupted read of the organization's texture does. The enterprise applies this discipline to security posture. Talent posture deserves the same.
This signal is not always about one person. When the same disengagement shows up across a cohort, a division, or a level in the organization, the frame shifts from management problem to culture problem, and the intervention has to move with it. (See: Culture Is the Retention Strategy. The Program Isn't.)
The signal also runs one layer up. Gallup's 2025 workplace research finds manager morale is the single strongest leading indicator of team morale. If the executive is reading their senior ICs' silence with rigor, they should be reading their VPs' silence with the same rigor. Manager disengagement cascades, and the signal at that level moves before anything shows up in the team.
What the executive owns
The talent-risk signal sits above HR and above the line manager. Both functions run process, and process does not read texture. The executive owns the attention layer of the organization: the layer where the pattern is visible early enough to intervene, where the intervention is a series of small course corrections instead of a retention counter-offer, and where the enterprise avoids the cost of learning about a departure from an email in the inbox.
The senior engineer who used to bring problems brings status updates now. Proposals that used to attract three counter-ideas come back "it looks fine." That is the signal. The executive who reads it and does nothing has decided, on behalf of the enterprise, to absorb the cost of what happens next.
Adam Cooper is a Marine Corps veteran who leads global technology operations across maritime, transportation, hospitality, and industrial environments. He writes about enterprise IT governance, distributed operations at scale, and the executive dynamics of senior technology leadership. Connect on LinkedIn or Send Email.