How to Do Succession Planning for the Roles You Will Actually Need.

Most guides to succession planning give the same answer: list the key roles, find people who could fill them, train those people. It is sound advice for a world where the job stands still.

The job is not standing still. A successor chosen against today's job description can be fully ready for a role that no longer exists by the time the seat is empty.

Many plans also start too late. In Korn Ferry's 2026 survey of 250 board directors and chief executives, half the boards said succession planning started too late in their last CEO transition (Korn Ferry, July 2026).

This guide covers the usual steps, but it starts where most plans skip: with what each critical role will actually need in three years.

What succession planning is, and what it is not.

Succession planning is the ongoing work of identifying the roles a business depends on and developing people who can step into them, so that a departure becomes a handover rather than a crisis.

It is easy to confuse with two neighbors:

• Replacement planning answers who covers the seat tomorrow if someone leaves today. It is one part of succession planning, not a substitute for it.
• Workforce planning answers how many people, with which skills, the business will need. The difference from succession planning is mostly scale: one plans the whole headcount, the other a handful of named roles.
• Succession planning answers who leads next, and whether they will be ready when it matters.

One note on scope: this guide is about leadership succession inside an organization. Handing ownership of a company to the next generation adds legal, tax and family questions that deserve their own advice.

Why succession plans fail.

Most organizations know they should do this, and most are not getting the result. In i4cp's 2026 study of 759 business and HR leaders, only 40% said their organization was effective at building sufficient bench strength for critical roles (i4cp, September 2026).

Those that were effective were nearly seven times as likely to report successful succession outcomes. Yet few tie the plan closely to where the business is going: only 18% said succession is highly integrated with enterprise business planning.

How business and HR leaders rate their succession planning

Effective at building bench strength for critical roles 40%

Highly integrated with enterprise business planning 18%

Strongly integrated with strategic workforce planning 14%

Ask why succession planning fails and the same five causes come up:

• The plan is built for the role as it is. Success criteria get copied from the current job description, so the bench is ready for yesterday's job.
• The current leader never lets go. A successor cannot get ready for decisions they are never allowed to make.
• Readiness is judged on impressions. Ratings and a nine-box grid measure how people are seen, not what they have decided.
• Nobody is paid to develop a successor. So, under pressure, nobody does.
• The plan is written once and filed. By the time it is needed, half of it is out of date.

When nobody inside is ready, the seat gets filled from outside by default. Gartner research on executive development, published in June 2026, puts the share of succession gaps filled internally at only 53%.

An outside hire is sometimes the right call. It should be a choice, not the only option left.

How to do succession planning in 7 steps.

This is the succession planning process I would run with a CHRO today, while the HR function itself is being redesigned around agents. The order matters: steps 1 and 2 decide whether everything after them is worth doing.

How to do succession planning: the seven-step cycle

01

Pick the critical roles

What breaks if the seat is empty for 90 days

02

Describe each role three years out

The judgment calls that will stay with people

03

Look wider than the obvious deputy

At least one candidate from outside the team

04

Judge readiness on decisions

Evidence from real calls, not ratings

05

Develop through stretch decisions

Acting roles and decisions handed down

06

Make the current leader part of the plan

Hand over real decisions, and be paid for doing it

07

Name emergency cover, then review

Twice a year, with the leaders who own each role

When strategy or AI changes the work, go back to step 02 and describe the role again.

01Pick the roles the business cannot run without

Not every role needs a succession plan, and a plan that tries to cover everyone gets reviewed by no one. For each role you are considering, ask three questions:

• What breaks if this seat is empty for 90 days?
• How hard would it be to hire for from outside, and how long would it take?
• How much is the role itself changing?

A role that scores high on the first two is critical. A role that also scores high on the third is critical and at risk, because the person you would promote today may not fit it in three years.

For most mid-size organizations the list is the top two layers of leadership plus a handful of specialist roles, often fewer than twenty seats.

02Describe the role as it will be in three years

This is the step most plans skip, and the one that matters most now. Before anyone is named, write down what the role will need, not what it needs today.

In Rewired Leadership I call the method the Judgment Audit. Take the role's main outputs and sort them into three groups: work that AI and automation will execute, work people will still do but AI will speed up, and the judgment calls that have to stay with a person.

The success criteria for a successor come from the third group: the trade-offs, the ethical calls and the decisions made with incomplete information, which are the calls that stay with people as AI takes on more decisions.

If you do not yet know which parts of a role AI will absorb, that is the first question to answer. Naming a successor before you have answered it means choosing someone for a job that is already changing shape.

03Look wider than the obvious deputy

The obvious successor is usually the current leader's deputy, and often the current leader's copy. That is a risk when the role is changing, because the person most like today's leader is the person best fitted to today's job.

Boards feel the same pull. Half the directors in Korn Ferry's 2026 survey said they played it too safe in their last CEO succession decision.

Run the talent review across adjacent teams, not just down the reporting line. Describe roles by the capabilities they need rather than by title, so that someone in finance who has run a hard supplier negotiation shows up as a candidate for an operations role.

Ask each leader to name at least one candidate from outside their own team. A short list with one or two unexpected names is healthier than three people who all sit in the same meeting.

04Judge readiness on decisions, not ratings

Readiness is where most plans turn vague. Define the levels, and define the evidence that earns each one.

The evidence column is the point, because strong performance in the current job is a weak guide on its own. Potential is hard to predict and much easier to observe once someone has had to make real decisions in real time.

In the i4cp study, the organizations best at building bench strength were almost twice as likely to evaluate candidates through stretch assignments. A good plan creates those moments on purpose.

05Develop successors through stretch decisions

Courses and reading lists build knowledge. Readiness comes from making the role's decisions while someone experienced is watching.

Give each candidate a development plan built around three or four stretch decisions over the next year: an acting assignment while the incumbent is away, ownership of a decision that normally sits a level up, a cross-functional project with real consequences.

A bigger role also demands unlearning. The habits that made someone excellent in their current job, such as always having the answer and doing the work themselves, are often the ones that hold them back in the next.

That is why developing leaders through real decisions works better than a course catalog. For the one or two people closest to a top role, coaching a successor into a bigger role shortens the distance between understanding the job and doing it.

Whether that help is coaching or mentoring depends on whether anyone has actually done the job the successor is stepping into.

06Make the current leader part of the plan

This is the step no template includes, and it is often why plans stall. A successor can only practice the decisions the current leader lets go of.

In Rewired Leadership I call it the indispensability trap. Leaders who solve every hard problem themselves train the organization to stop learning and wait for them.

They build a company that is structurally dependent on their presence.

One senior executive put it to me plainly: "I didn't realize how much my availability was training people to depend on me."

The fix has two parts. First, the incumbent has to extend trust before it is fully earned and hand real decision rights to their candidates, not just tasks. Second, make it pay.

Part of a senior leader's bonus should depend on how well their people step up: how many decisions they delegate successfully, and how many of their team move into bigger roles.

07Name emergency cover, then review twice a year

Every critical role needs an emergency answer as well as a long-term one: who holds the seat for 90 days if the incumbent leaves tomorrow. Often it is not the long-term successor, and that is fine.

Write it down, along with a short handover list for each role: the key relationships, the open decisions and the things only this person knows.

Then review the plan twice a year, with the leaders who own each role in the room. Move candidates between readiness levels on evidence, drop roles that are no longer critical, and reopen step 2 whenever strategy or technology changes what a role needs. A succession plan that is never reviewed is a document, not a plan.

A simple succession planning template.

One table per critical role is enough. The example column shows the level of detail to aim for; it is illustrative, not drawn from a real company.

Who owns succession planning.

Succession planning in leadership works when ownership is explicit:

• The board owns CEO succession, and should review it at least once a year whether or not a change is expected.
• The CEO owns succession for the senior team.
• Every leader owns successors for the roles that report to them, and is measured on it.
• The CHRO owns the system: the review calendar, the readiness definitions and the succession criteria written for the roles you will need.
• HR business partners run the reviews and keep the evidence honest.

This is also where leadership development and succession planning meet. When they are run by different teams with different lists, people get developed for roles nobody has planned, and planned for roles nobody is preparing them for.

Succession planning best practices.

A short list beats a long one:

• Develop before you select. Give several people stretch decisions before naming one successor.
• Tell people they are being developed, without promising the job. Secrecy drives good candidates away; promises turn a plan into an entitlement.
• Keep the plan small enough to review. Twenty seats reviewed properly beat two hundred never reviewed at all.
• Review twice a year, and whenever the strategy changes.
• Measure outcomes, not paperwork.

The measures worth tracking:

• The share of critical roles with at least one ready-now successor.
• How often leadership roles are filled from inside, and how those appointments work out.
• How many candidates moved up a readiness level this year.
• Talent mobility, which Rewired Leadership defines as the share of people who move into a significantly different area of work within 18 months. A high number means people are growing into new work rather than waiting in line.

Frequently asked questions.

The short version.

A successor is only ready for the role you described. So describe the role you will need, judge candidates on decisions they have actually made, and make the person in the seat today part of the plan. Review it twice a year, and a departure becomes a handover instead of a scramble.