Succession planning for roles nobody has documented

Most succession plans are org-chart exercises. A name in a box, a successor in the next box, a review date on the calendar. The board sees continuity. HR sees a plan. Operations sees a name.

None of that describes how the work actually happens.

When the person in the box leaves, the successor inherits the title, the office, and the calendar invites. They do not inherit the twenty years of judgment that made the role work. That part was never in any file. It lived in one head. And now it is gone.

This is the gap succession planning almost never covers. And it is the gap that turns a planned transition into an eighteen-month operational drag. It is also the shape of key person risk in operations, a risk that shows up on nobody's dashboard until the person is already gone.

Succession plans cover titles, not operating knowledge

Walk into any large enterprise and ask to see the succession plan for a critical operations role. You will get a spreadsheet. Names, tenure, readiness ratings, development paths. Sometimes a nine-box grid. Sometimes a written development plan for the identified successor.

What you will not get is a description of what the role actually does.

Not the job description. Not the KRAs. The actual work, the decisions, the exceptions, the trusted sources, the escalation moments, the reasons behind judgment calls that never made it into any policy manual. The operating knowledge. The real process.

That part is not in the plan because nobody wrote it down. Not because HR was lazy. Because the person doing the role never had a reason to write it down. They just did the work. Roughly 80% of business processes are undocumented [Tallyfy]. In the knowledge-heavy roles that succession plans target, senior underwriters, plant planners, risk analysts, senior claims handlers, the undocumented share is usually higher, not lower.

So the succession plan is real, but it is incomplete. It plans who will sit in the chair. It does not plan for what the person in the chair actually knows.

What the replacement actually inherits

Here is what a well-prepared successor typically walks into on day one.

A title. They now hold the role.

A team. People who used to escalate to the departing expert now escalate to them.

A calendar. Recurring meetings, standing check-ins, the review cadence.

A policy binder or a wiki page. The official process. What the company says should happen.

Here is what they do not inherit.

Decision logic. Why one case gets approved and a nearly identical one does not. Why an exception was granted in March but refused in May.

Trusted sources. Which system field is fresh, which report is stale, which colleague to call before making a call.

Escalation instincts. When to stop and check with legal. When to just do it. Which "urgent" flags are real and which are not.

Relationships. The vendor contact who will actually pick up. The internal stakeholder who needs a heads-up before a decision goes through. The regulator who prefers a certain framing.

That entire second list is the operating knowledge, and it is the part that makes the role work. Research suggests that 42% of essential expertise lives only in employees' heads [Panopto]. In a knowledge-heavy succession, that share is almost everything the successor actually needs.

Absent all of that, the successor does not run the role. They run a title. They handle the easy cases well and get stuck on every case that requires judgment. That is where the operational drag starts.

The retirement cliff makes this urgent

For years, succession planning was a slow-moving discipline. Median tenure in the US private sector is 3.5 years [Bureau of Labor Statistics], so movement was constant, but experienced operators stayed in roles long enough that quiet, informal handoffs mostly worked. A new hire shadowed a veteran for a few months. Enough tacit knowledge transferred by osmosis.

That is no longer the shape of the workforce.

Deloitte and the Manufacturing Institute project 2.1 million US manufacturing jobs will go unfilled by 2030, largely driven by retirements. The same demographic pattern shows up in insurance, banking, utilities, healthcare, and skilled operations across the industrial economy. The most experienced operators, the ones holding the tacit judgment behind the work, are leaving in a compressed window.

Succession plans built for a slow-drip workforce do not survive a retirement cliff. Shadowing does not scale when a dozen experts retire in the same year and there is nobody senior enough to be shadowed. Wiki updates do not survive a departure schedule measured in months. And a title-only succession plan turns into a stack of successors who all inherit chairs without operating knowledge.

The retirement problem is not primarily a hiring problem. It is a knowledge transfer problem. And it is the one succession plans were never designed to solve.

Why "have them document it" does not work

The standard reflex is documentation. Give the departing expert three months and a mandate to write down everything they know. Build a runbook. Record a series of video walkthroughs. Populate the wiki.

I have watched dozens of enterprises try this. It does not work, and there are three reasons.

The expert does not know what they know. Most of the tacit judgment behind a role is not accessible by introspection. Ask a senior underwriter to write down their decision logic and you get the official version, because the real version only surfaces when a real case hits the desk. The expert cannot document what they cannot see.

Documents describe the official process. Even when the expert tries hard, what ends up on the page is what the company says should happen, not what actually happens. The workarounds, the "I check this field first because the other one is usually wrong," the "I always send this before 3pm on Tuesday", those rarely make it in. Documentation captures the surface. The real process lives underneath.

Documents are not machine-readable in the way an AI or a structured onboarding path can use. A binder of prose is a static artifact. It cannot be queried, cross-referenced against a live system, or wired into a workflow. The successor gets a document to read. That is not the same as inheriting the working logic.

The right output is not a document. It is a structured, validated capture of the operating knowledge itself, usable both by humans in onboarding and by the systems the role runs on top of. This is the discipline behind capturing tribal knowledge before it walks out the door: treat the tacit judgment as an asset the company owns, not a story the expert tells on their way out.

What succession planning looks like when you plan for knowledge

Once the reframe is done, the plan changes shape.

Identify the roles where the risk is real. Not every role needs deep knowledge transfer. But for the small set of roles where a departure creates an operational gap, senior underwriters, planners, risk analysts, senior claims handlers, key operators, succession has to include a knowledge component, not just a name.

Start capturing while the expert is still there. This is the part almost every plan gets wrong on timing. The right moment is not the notice period. Two weeks is not enough to extract twenty years of judgment. The right moment is now, years before the departure, while the expert is doing the work every day and the exceptions are surfacing naturally. The expert everyone calls is the person to start with.

Use structured extraction, not documentation. Sit with the expert in a conversation that is designed to surface the real work: why this case, not that one. What happens when the process breaks. Which sources they trust. Where they stop and ask. This is expertise mining, a discipline for capturing what documentation cannot reach. The output is not a transcript. It is a machine-readable expert profile, structured so a successor or a governed agent can use it directly.

Validate against real cases. A single expert's account is a starting point, not the truth. Compare what one expert says against what another does. Test the captured logic against historical cases. Correct the drift.

Wire it into onboarding. The successor no longer inherits only a title and a binder. They inherit a structured expert profile that shows the decision logic, the exceptions, and the escalation reasoning behind the role. Onboarding time drops because the successor is learning the real work, not reverse-engineering it. This is what changes in onboarding knowledge-heavy roles once the operating knowledge is captured.

Wire it into the systems. If the role runs on top of SAP, a policy admin system, or a planning tool, the captured expertise can also drive a governed agent operation next to the successor, handling the routine cases with human-in-the-loop approval and letting the successor focus on the ones that need judgment. This is where succession stops being a one-time handoff and becomes a continuous operating capacity that survives the next departure and the one after that.

The successor should not be starting from zero

Every succession is a handoff. The question is what gets handed off.

In most enterprises today, what gets handed off is the title. The successor starts from zero on operating knowledge, and the company absorbs eighteen months of drag while they rebuild it, imperfectly, from scratch, against real cases in production.

We have watched this pattern break inside real companies. At a large multi-brand insurance group, the senior renewal expertise sat with a small team. Once that expertise was mined and turned into a structured profile, the role stopped being one departure away from disruption. The operating knowledge belonged to the company, not to one person's memory. At a supplements manufacturer, senior HR judgment on candidate screening was captured the same way. At a maritime chartering firm, senior matching judgment was captured before the departure risk showed up on anyone's dashboard. In each case, the succession plan finally had something to hand over besides a job title.

The board keeps asking whether the company is prepared for the next transition. The honest answer, in most cases, is that the plan covers who sits in the chair and not what the chair actually does. Fixing that is not an HR-tech project. It is a decision to treat the tacit judgment behind critical roles as an asset that the company owns, captures, and hands over, instead of an asset that walks out the door every time someone retires.

Succession that plans only for titles is not succession planning. It is chair assignment. The plan has to reach the knowledge, or the plan does not work.

Start with the roles where the risk is real. Capture the operating knowledge while the expert is still doing the work. Hand the successor something worth inheriting.

Mine the expertise. Then operate differently.