TL;DR
- The real decision: whether to extend succession work one to two levels below the executive team, into the roles whose absence stops work.
- When to skip it: the executive bench is already shaky, the last two senior moves were external, and there's no appetite for the deeper exercise yet.
- What has to be true: a willing line of business to name the critical roles, a place to write the answer down, and a calendar slot to revisit it before the next resignation.
- How the options split: map the work first then find the people, build a ready pool against named roles, or accept an external buy on every departure.
- Decision rule: if more than one role empties without a cover in the last eighteen months, the bottleneck is the role definition, not the talent.
- Expected outcome: fewer projects stalled by a single departure, and a much harder argument to make when someone asks why a successor was never named.
A regional finance manager in a mid-sized industrial business opens her laptop on a Tuesday morning in March, sees the resignation from her controls accountant, and counts what depends on her. The month-end close. A covenant report due to the bank in six weeks. A new ledger project that she is the only person who has actually touched. The executive succession plan on the shared drive names her boss, his boss, and three people above that. Her name is nowhere in it. But she has been the single point of failure for the audit committee's favourite number for two years, and the org chart never noticed. That's the gap this article is written to close. The real issue isn't whether the executive bench has cover. It's whether the work that actually stops when someone leaves has cover too.
When You Genuinely Do Not Need to Act Yet
There's a version of this problem where the answer really is "later". Pretending otherwise is the fastest way to lose the reader who is reading carefully. Four honest stages, from "fine" through "friction" through "risk" through "edge case", each with a concrete picture of the company it describes.
A professional services partnership of ninety people has named successors for its six equity partners, refreshed them every year, and the last mid-level departure was absorbed in under three weeks by a senior associate who was already doing the work. The bench is shallow at the partner level but the engine room is broad and the work is documented. Nothing about this firm is on fire. If the reader works here, the right move is to put this article down, finish the partner refresh, and come back when something breaks.
Best tools for HR Strategy
Friction looks different. A logistics operator with three hundred and fifty staff has a clean executive succession plan, but the last two people who resigned were a transport compliance lead and a depot operations manager, and both moves took over ten weeks to fill. Nobody raised it as a strategic problem because the executive team was unaffected. The work was affected. Customers were not, but the people underneath felt it, and a regional director has quietly started keeping a private list of who he would call if his depot manager left. That list is the friction.
Real risk shows up when the executive plan still looks immaculate but two layers down there's a role nobody outside one team can describe. A healthcare provider with eight hundred staff runs a flawless executive succession programme that the board sees every quarter. Two levels down sits a single compliance analyst whose departure would mean a regulatory submission misses a window. The CHRO knows. The CEO doesn't, because the work is so specialised that nobody at the top has ever had to think about who would do it. If that person resigns, the problem isn't "we lost an analyst". The problem is "we didn't know we had a single point of failure".
The edge case is the one this article keeps in mind for honesty. A business that genuinely can't afford the exercise because the executive bench is itself unstable, the last two senior moves were external because nobody inside was ready, and the leadership team is spending every succession conversation talking about the C-suite. Trying to extend below the top five in that situation is a polite way to do nothing. The right move there's to fix the top first, and revisit this piece when the executive plan stops being the only thing on the table.
Five Questions You Ask Yourself at 11pm
A CHRO who has read four articles on this and isn't satisfied asks sharper questions than the articles answered. Five of them, in the form they actually arrive.
"If the head of procurement left on Friday, how long until someone in this building could answer a supplier who phones on Monday?" The honest answer for many businesses is "longer than I would like to admit". The role is senior enough to appear on an org chart and not senior enough to appear on a succession plan, and the work it owns is dense enough that the answer is "weeks, not days". The reasoning behind asking it's that procurement failures rarely become visible at the top until a payment goes out wrong or a contract expires without renewal. By then the cost is real. If the answer is uncomfortable, the work that depends on that person has not been mapped. That's the gap to fix before any other question matters.
"Which project would I cancel if my head of FP&A resigned tomorrow?" The question is uncomfortable because the answer is usually a project the board cares about, not the one the CHRO would choose to lose. The reasoning is that the answer reveals which role holds something the executive team has not noticed it depends on. If the answer is uncomfortable, the role is more critical than its title.
"How many people in this company could step into my direct reports' direct reports' roles in under sixty days?" Most readers can't answer this question without going away and counting. That's the point. The reasoning is that depth below the leadership team is invisible until it's missing, and the question forces it into view. If the answer is "two", the bench is thinner than it looks.
"When did I last write down who could cover whom, and where is that document?" If the answer involves a memory rather than a file, there's no plan, there's a feeling. The reasoning is that plans which live in the head of one person don't survive the departure of that person. If the answer is "in my notebook", it doesn't exist.
"What did the last mid-level departure cost us, and was it on a spreadsheet anywhere?" If the cost was real and it doesn't appear in any leadership report, the organisation has decided it doesn't matter. That's the moment to argue that it does. The reasoning is that a problem with no visible cost gets no visible budget to fix.
Three Honest Categories the Approaches Split Into
Most of what is sold under the heading of succession planning falls into three honest categories. Each has a real use, a real failure mode, and a real example. None of them is a vendor product, because the choice is about method before it's about tools.
The first category is named successors for named roles. The organisation draws a list of its critical positions, and for each one names a person who could step in within a defined window, often ninety days. The strength is clarity. Everyone knows who is being developed for what, and the development conversation is honest. When it works, it works because the named person has a real runway to grow into the role, and because the line manager treats the development as a live obligation rather than a box-ticking exercise. When it fails, it fails in two specific ways. The first is that the named successor leaves and the document goes stale. The second is that the named successor is the wrong person, and the person who knows is too polite to say so. A CHRO in this situation reads the same document for three years running, knows the named successor is no longer ready, and never updates it. The failure mode isn't the absence of a plan. It's the plan that has stopped being believed.
The second category is a ready-now talent pool not tied to specific roles. The organisation identifies a cohort of people who could step up into broader categories of work within a defined window, without committing them to a specific destination. The strength is flexibility, because the pool can be deployed against a vacancy that has not been named. The weakness is that pools become invisible. The people in them aren't told, the line of business has not committed to develop them for any particular move, and the pool quietly empties every time a high performer in it gets promoted for a different reason. A business that runs this approach well has a development budget that's genuinely attached to the pool members, and a leadership team that revisits the membership every year. A business that runs it badly has a slide in a deck that no longer describes anyone's career.
The third category is to start from the work, not the seniority. The organisation maps which pieces of work would stop if a given person left, and asks, role by role, what would actually happen. The strength is that this is the only approach that surfaces the single points of failure two or three levels down that the other two miss entirely. The weakness is that the exercise is harder to defend at the executive table, because the roles it surfaces are often not senior. A CHRO who runs this exercise and produces a list of mid-level roles that need cover has to make an argument to the CEO that the CEO has not asked to hear. When it works, it works because the CEO has just watched a mid-level departure cost something, and the list is the response. When it fails, it fails because the executive team decides that mid-level cover is the line of business's problem and washes its hands of it.
Five Diagnostic Questions You Can Self-Assess Against
These aren't the questions to ask in a meeting. They're the questions to ask in a quiet hour, with a piece of paper, against your own organisation.
How would a new department head describe each of their reports' reports within five minutes, including who could cover whom? The way to answer it's to pick one department, give yourself fifteen minutes, and write it down without reference to any existing document. If the description requires more than five minutes or feels thin, the work has not been mapped at the level the question assumes. What it tells you is whether the critical-role map already exists, or whether it has to be built.
Which of the last four mid-level departures took longer than ninety days to fill, and what did the work do in the meantime? The way to answer it's to go back through the records of the last four non-executive departures and look for the gap between the resignation date and the date someone was actually doing the work, not the date someone was hired. If the gap is large and the work was held by temporary cover or by overworked peers, the cost is real and uncounted.
When a senior leader reviews their team, do they name cover for the critical mid-level roles out loud, or do they assume someone else has it? The way to answer it's to sit in on one talent review and listen for whether the conversation extends below the senior leadership team. If it stops at the layer that has direct exposure to the executive team, the layer below is invisible in the conversation and probably invisible in the plan.
Is there a single document that says, for each critical role, who is being developed to cover it, and when it was last updated? The way to answer it's to look. If the document doesn't exist, the question answers itself. If it exists and the date on it's older than twelve months, the document has stopped being a plan and become a record of a past decision.
If the executive team lost its chief of staff tomorrow, how many of the mid-level decisions would stall, and which ones? The way to answer it's to think about the last month of chief-of-staff work in your own organisation, and imagine it undone. The question is a probe for whether the chief of staff has become an unnamed critical role, which is the single most common shape of the problem this article is about.
Five Succession Models, Reviewed
A Named Successor for Each Critical Role
The most familiar shape. The organisation identifies its critical positions and names, for each one, the individual who is being prepared to step into it. The development work is then attached to that person, the line of business owns it, and the named successor is told they're in scope.
Why it earns a place is that it converts a vague worry about depth into a specific development obligation, with a name on it. The plan reads as a real plan. The development conversation is honest. The line of business can't pretend it doesn't know.
Where it falls short is the moment a named successor leaves, takes a sideways move, or simply stops being the right answer. The plan goes stale, and because the plan is in writing, it becomes a document that everyone can see is wrong but nobody updates. The second shortcoming is more subtle. A plan that names successors publicly is a plan that can disappoint publicly. The people who were named, and the people who were not, both react to the document, and the HR team ends up managing reactions to the plan instead of managing the work.
Ready-Now Talent Pools Not Tied to Specific Roles
The organisation identifies a cohort of high-potential people who could move into broader categories of work within a defined window. The pool is reviewed yearly. The members get development that's genuinely attached to the membership.
Why it earns a place is flexibility. Vacancies that haven't been predicted can be filled from the pool. Lines of business that lose someone unexpected can be served from a shared resource.
Where it falls short is that pools become lists. The people in them leave for other reasons, the membership goes quiet, and the pool becomes a slide in a deck that nobody believes. The development budget that was supposed to be attached to the pool members often isn't, because nobody owns the obligation specifically. The second shortcoming is that a pool not tied to a destination is harder for the line of business to engage with. A line manager asked to develop someone for "a broader category of work" hears that as "no specific work". A line manager asked to develop someone for a specific role hears it as a real conversation.
Critical-Role Mapping That Starts from the Work
The organisation sets aside seniority and asks, role by role, which work would actually stop if a given person left. The output is a map of critical roles that's not the org chart.
Why it earns a place is that it's the only model that surfaces the single points of failure two or three levels down that the other models miss. The compliance analyst. The controls accountant. The depot manager in a remote location. The people who hold something the executive team has not noticed it depends on.
Where it falls short is that the roles it surfaces are often not senior, and the executive team has to be persuaded that mid-level cover is its problem. The exercise is also slower, because it requires a real conversation with the line of business about what the work is, not a tidy list pulled from a system. The second shortcoming is that the critical-role map goes stale in a specific way. The work changes before the role changes, and the map describes last year's shape of the work, not this year's.
A Skills-Based Bench Where Cover Is Defined by Capability
The organisation defines the capabilities that matter for its critical roles and identifies, for each capability, the people in the business who hold it to a high level. Cover is defined as a combination of capabilities, not as a person.
Why it earns a place is that it survives the departure of a named successor in a way the named-successor model doesn't. The capabilities stay even if the people holding them change. The conversation about cover is a conversation about the work, not a conversation about who was passed over.
Where it falls short is that capabilities without context are a weak signal. A person who holds the capability to do a job isn't the same as a person who can do the job in this company, in this team, against this customer. The second shortcoming is that the exercise requires the line of business to articulate what the capabilities actually are, which is harder than naming a successor. Most lines of business would rather say "John is the person" than explain what the role actually requires.
Buying on Vacancy with No Internal Plan at All
The organisation accepts that every departure above a certain level will be filled externally, and runs no internal development work against the critical roles.
Why it earns a place is that it's honest in a way the other models sometimes aren't. If the business genuinely can't develop its own people for its own critical roles, an external buy is a real answer, and a confident one. It doesn't pretend to be a development programme that doesn't exist.
Where it falls short is the cost, in the form of leadership that doesn't know the business as well, transitions that take longer than an internal move, and a culture that quietly tells its best people that there's no path through. The second shortcoming is that the model collapses when the external market is also thin, which it's for many of the specialist roles this article is actually about.
The Decision Table
| Situation | Scale | Setup | Primary Pain | Recommended Starting Point |
|---|---|---|---|---|
| Executive bench solid, last mid-level move stalled | Mid-size, single country | Named exec successors, no work-based map | A project stalled when one person left | Critical-role mapping that starts from the work |
| Executive bench shaky, mid-level deep | Mid-size, multi-site | Exec plan in progress, broad mid-level | No time to fix both at once | Fix the top, defer this exercise |
| Multiple mid-level moves in eighteen months, no cover held | Mid-size, single HQ | No work-based map, no named cover | Repeated transitions | Named successor model for the critical roles only |
| Specialist role just emptied, no internal candidate | Any size | Talent pool exists but is generalist | Single role with no obvious answer | Skills-based bench against the capability mix of that role |
| High-potential cohort exists, line of business disengaged | Mid-size or large | Pool exists, no destinations named | Pool is not being used | Connect the pool to named critical roles |
| Business cannot name the critical roles at all | Any size | No map | Cannot start any other work | Critical-role mapping exercise first |
| All-internal development culture, but no documentation | Mid-size, single country | Plan lives in heads | Plan does not survive a departure | Write the named-successor plan down |
| Recent external hire because no internal candidate | Mid-size | Exec plan solid, mid-level thin | Internal candidates feel blocked | Skills-based bench against critical-role capabilities |
Deciding Which Roles Are Actually Critical
The single most common mistake in extending succession below the top five is to confuse seniority with criticality. They overlap, sometimes, but they're not the same thing. The test below is the difference between a role that earns cover and a role that earns a salary.
| Test | What a critical role looks like | What a merely senior role looks like | How to check |
|---|---|---|---|
| If the role is empty for sixty days, does the work stop? | A specific project misses a deadline, a regulatory submission is late, a customer escalates | A meeting gets rescheduled, a decision gets queued, an initiative slows down | Ask the line manager what would happen, and write the answer down verbatim |
| Is the work in the role documented? | A few things are, and the rest is not | The work is documented, or could be | Look for the standard operating procedure; if none exists, ask how a new hire would learn the role |
| Are there two other people in the company who could do parts of it? | One, possibly, and they are also busy | Several, and at least one is ready | Name the two people out loud; if either name is unclear, the role is more critical than its title |
| Would an external hire take more than ninety days to be effective? | Yes, because the role holds context that only the incumbent has | No, because the work is generalist or transferable | Ask the line of business how long the last external hire took to be useful in a comparable role |
| Does the role's output reach a senior audience? | A board paper, a regulator, a major customer | An internal team, a process owner | Trace the last month of the role's output upwards and see how far it travels |
The mistake to avoid is treating every senior role as critical. A senior role that's also well-documented, broadly understood, and has two ready internal candidates isn't the role this article is about. The role this article is about is the one where the answer to several of these is "no", and the executive team has never been asked.
Telling Someone They Are on the List, or Not
The disclosure question is the one this article doesn't pretend to answer for the reader. The shape of the decision is this. There's a list. Some people are on it. Some people aren't. The act of telling people, or not telling them, has consequences in both directions.
If you tell someone they're a named successor for a critical role, the consequences are mostly positive for that person. They feel seen. They know the development conversation is real. They're likely to engage with it more seriously. The cost is that they may also become less patient with their current role, may interpret any setback as a signal that they're no longer on the list, and may leave the moment an external offer arrives, on the assumption that they were not going to get the role anyway.
If you don't tell someone they're on the list, the cost is the development conversation becomes harder. The line manager has to deliver development work without explaining why. The person being developed has to take on stretch assignments without knowing they're stretch. The system works, but it works quietly, and quietly is fragile.
The harder problem is the people who find out they were not on the list. If a CHRO runs a succession exercise and the line of business names John for the critical role and not Priya, and Priya later sees the plan, Priya will draw conclusions. The conclusions won't be generous. The way to manage this is to be honest with Priya about what she is being developed for, separately from what John is being developed for, and to make sure the development plan for Priya is real. The temptation to keep everyone happy by telling everyone they're on the list is the temptation to produce a plan that means nothing. The plan that means nothing is worse than no plan, because the document exists and the next CHRO inherits it.
What to Put in Writing
The artefacts this decision has to leave behind are the difference between a plan that survives a departure and a plan that doesn't. Each one has an owner, a moment when it's written, and a specific failure it prevents.
| Artefact | Who owns it | When it is written | What it prevents |
|---|---|---|---|
| Critical-role map | CHRO, with line of business sign-off | After the work-mapping exercise, before any successor naming | A plan that targets the wrong roles |
| Named-successor record per role | Line of business head, with HR partner review | When a successor is identified, refreshed annually | A stale plan that names someone who has left |
| Development plan per named successor | Line manager, with HR partner | Within thirty days of being named on the plan | A succession plan with no development behind it |
| Disclosure policy | CHRO | Once, before the first conversation | Ad-hoc disclosure that damages trust |
| Cadence and review schedule | HR operations | At plan launch | A plan that goes twelve months without review |
| Decision log for non-obvious calls | HR business partner for the area | At the moment of decision | A plan whose reasoning nobody can reconstruct a year later |
| Refresh record showing what changed | HR operations | After each annual review | A plan whose drift nobody noticed |
| Departure-cover record from each actual move | Line of business head | After each departure, before the post-mortem closes | The same gap recurring in the next role |
Questions to Ask Before You Commit
These are the questions to put to a provider, an adviser, or your own team before this decision is locked. They're written to be copied into an email, used in a meeting, or held against your own plan. Each one names what a bad answer sounds like.
On the work first. What does your method do to map the critical roles before any successor is named? A bad answer is "we use the existing org chart and the senior leadership list". A plan that has not separated seniority from criticality is a plan that has not started.
On the line of business. Which conversations happen with the line of business, and at what level? A bad answer is "we do it all from HR". A line of business that doesn't own the plan is a line of business that won't develop the successor.
On the development behind it. What development activity is attached to a named successor, and who is accountable for it? A bad answer is "we put them on a leadership programme". A generic programme isn't a development plan for a specific role.
On the disclosure question. How does your approach handle telling people they're, or aren't, on the plan? A bad answer is "we leave that to the line manager". Disclosure done ad hoc is disclosure done badly.
On the refresh. How often is the plan revisited, and what triggers an out-of-cycle refresh? A bad answer is "annually, in the talent review". A plan that only refreshes once a year is a plan that has gone stale by month seven.
On the documentation. What artefacts does the plan produce, and where do they live? A bad answer is "in the system". A plan whose documents can't be found is a plan that doesn't exist.
On the predecessor question. What does your approach say about the people who aren't named? A bad answer is silence. A plan that doesn't address the people who were not chosen is a plan that will damage trust the moment it's seen.
On the post-departure learning. What does your approach do with what is learned after a vacancy is actually filled? A bad answer is "we close the requisition". A plan that doesn't learn from departures is a plan that will repeat them.
What Getting This Wrong Costs
The cost that lands on an invoice is the easy one. It's the recruitment fee, the agency spend, the signing bonus, the lost quarter of productivity from the new hire ramping up. The cost that doesn't land on an invoice is the one that's harder to argue against, and is also the one that compounds. So this section names it.
The second-order cost is the project that gets cancelled or delayed, and the customer or regulator who notices. A controls accountant who leaves at the wrong moment isn't a recruitment problem. It's a covenant report that misses a window, or a year-end close that the audit committee sees with the wrong number in it. The cost of that's not in the recruitment budget. It's in the credibility of the finance function, and credibility is a budget that doesn't get topped up by Finance. It's drawn down once, and rebuilt slowly.
The second second-order cost is the people who watch it happen. A high performer in the team below the controls accountant, watching the close stall for a quarter, watching the team work late, watching the audit committee receive an explanation, draws a conclusion about what this organisation is. The conclusion isn't "they have a vacancy". The conclusion is "they don't know what they depend on". The high performer who draws that conclusion is the high performer who answers the next external call more quickly.
The third second-order cost is the next time the same role empties. If the gap that the controls accountant left is closed by an external hire, and the new hire takes six months to be useful, and the gap isn't named in any document, then the same gap will appear again the next time the role turns over. The cost compounds across cycles, and the cycle is shorter than the planning horizon. If a critical role empties twice in five years and both moves take more than ninety days, the question worth asking isn't "how do we fill this role faster" but "how did we let this be the shape of this role twice".
When You Are Ready to Go Further
If the diagnostic questions above have produced more answers than the reader was comfortable with, the next move is to see how other organisations in similar positions have approached the same problem. HROpsLab publishes independent comparison work across HR operations, payroll, and workforce planning, and the case study library is the part of the site where the longest-running decisions are written up. The work is independent. Nothing on the site is paid for by a vendor, and no vendor has editorial input. The team doesn't sell software, doesn't sell payroll services, and doesn't sell advisory engagements. The comparison work exists so that a reader can come to a decision with the shape of the trade-offs already in view.
For readers who want a second opinion on their own position before committing to a specific approach, the site carries a short advisory note from the editorial team on what to take to a provider and what to ask of them, written specifically for readers who have read four articles on this and would like the fifth to be a checklist rather than a definition.
Frequently Asked Questions
What does succession planning cover below the executive team?
Below the top five, succession planning covers the roles whose absence would actually stop work in the business, which is usually a much shorter list than the senior leadership team. The exercise starts from the work that would stall if a given person left, and works upward to the role. It isn't the same as mapping every senior role in the business. The output is a list of roles that earn development investment because the work that depends on them can't afford a long vacancy.
How do I identify a critical role?
A role is critical when the work in it wouldn't survive a sixty-day vacancy without a named cover, when most of the work is held in the head of the incumbent rather than in a procedure, and when an external replacement would take longer than ninety days to be effective. The check is to ask the line manager what would happen if the role were empty, and write the answer down. If the answer names a missed deadline, a regulator, or a major customer, the role is critical regardless of its seniority.
Should I tell someone they are on the succession plan?
Telling someone they're a named successor is usually the right call, because the development conversation is more honest when the person knows what they're being developed for. The cost is that the person may read every subsequent decision as a signal about whether they're still on the list, and may leave if they conclude they're not. The right balance is to be honest with the named successor, and to be honest with the people who aren't named about what they're being developed for separately, so the plan doesn't become the only signal anyone receives.
How many successors does a critical role need?
For most critical roles, one ready successor is the realistic answer, with a second person identified who could grow into the role over a longer window. The mistake is to name three people as successors and develop none of them. A plan that names three and develops one is a plan that has done the work. A plan that names three and develops three is a plan that's not credible to the line of business.
What do I do when there is no internal candidate for a critical role?
The first move is to be honest with the line of business that the role is being filled externally, and to name what that means for the transition. The second move is to identify the capabilities the role requires and start developing a bench against those capabilities, so that the next vacancy has a different answer. The wrong move is to pretend the external hire is a long-term answer and not develop the bench behind it.
How often should the plan be refreshed?
The plan should be revisited annually, and out of cycle whenever a critical role turns over, whenever a named successor leaves, and whenever the shape of the work in a critical role changes materially. A plan that's only refreshed once a year is a plan that's wrong for the second half of every year. The refresh should be short, written down, and visible.
Is succession planning the same as development planning?
No. Development planning is broader, covers everyone, and is attached to current role performance. Succession planning is narrower, covers a small set of critical roles, and is attached to the question of who could step into a specific destination. A development plan can exist without a succession plan. A succession plan without development behind it's a list of names, not a plan.
How do I handle a named successor who leaves?
The named successor record is updated immediately, the line of business is asked whether to name a replacement or treat the role as externally filled, and the development investment that was attached to the departing successor is reviewed to see whether any of it can be redirected. The post-departure review should also ask whether the role's criticality has changed because the person leaving has changed the shape of the work. If it has, the critical-role map needs updating too.
Make better HR decisions, faster, with independent comparison work from HROpsLab.
