Outgrace
OutgraceThe Leadership Magazine, home
LeadershipHow-to guide

How to set approval limits between a founder-chair and a new chief executive

Write down the founder-chair's approvals, sort them against the new chief executive's, and have the board adopt the line. The sources we checked offer no code or study to lean on.

A founder who hands the chief executive title to someone else often keeps the chair, and with it the habit of approving things. The new chief executive usually learns where the job's authority ends in the first capital decision that needs the founder's sign-off, and nobody has written down whether it does. This guide shows how to draw that line on paper before the moment arrives, so that the founder, the new chief executive and the board work from the same page. One caution at the outset. We searched for a board-governance code, and for research on founder-chair and chief executive approval splits, and found neither in the sources we checked. The method that follows is editorial judgement, and we say so where it matters.

Before you start

Four things should be true before anyone drafts a line. The first is a shared purpose. The schedule exists to stop decisions falling between two people. It is not there to shrink the founder's role or to give the chief executive a list of grievances. If either party reads it as a demotion, the document will be argued over line by line, and the argument will be about status rather than about the company.

  • The current delegation of authority, in writing, even if it is out of date.
  • The approval thresholds already in use for spending, hiring, contracts and financing.
  • The founder's standing approvals, including the ones made out of habit rather than by rule.
  • A board resolution, or a plan to pass one, that makes the final schedule the board's rather than either individual's.

The fourth item is the one boards most often skip. Without a resolution the schedule is a memo, and a memo loses to the next urgent phone call. On the founder's side, the National Bureau of Economic Research says in a Bulletin item that entrepreneurship training may help founders judge when to stop investing time and money in a venture. Knowing when to stop is a skill, and it applies to approvals as much as to money. Ask the founder plainly which decisions they would find hardest to stop making. The answer tells you where the schedule will be tested first, and it is usually a decision the founder cares about personally.

  1. Step 1

    Count the founder's approvals

    For the last twelve months, list every decision the founder approved outside the formal board process. Cover payments, hires, customer terms, product launches and financing talks. Build the list from payment records, minutes and email, not from memory. What usually goes wrong is that the founder's most frequent approvals are the ones least remembered, so the first list is always too short.

  2. Step 2

    Sort each decision into one of three columns

    Mark each item as owned by the chief executive, as needing the chair's approval, or as reserved to the board. A sensible first test is reversibility: if a decision cannot be undone inside a quarter, it belongs higher up. What usually goes wrong is that the columns fill unevenly, with almost everything in the chief executive's column, because the founder has stopped regarding the rest as theirs to approve.

  3. Step 3

    Set money limits per decision and per year

    Write a threshold for each category of spending, commitment and contract, in the company's own currency, and put a cumulative annual limit beside it. The annual figure matters because it stops a series of small approvals adding up to a large commitment nobody reviewed. What usually goes wrong is that limits are set per invoice, so a large contract is quietly split into smaller ones that each fall under the line.

  4. Step 4

    Write the chair's limits on the same page

    The founder's own authority should be printed on the schedule, with its own limits and a short list of decisions the chair may not take alone. A schedule that binds only the chief executive reads as a rule imposed on someone else, and it will not last past the first disagreement. Writing the chair's limits is also the clearest sign that the board, not either individual, owns the document.

  5. Step 5

    Name the route for disputes

    Say who decides when the founder and the chief executive disagree, how quickly the question must be answered, and whether the board hears it before or after the decision is made. Name a director who can be called in on the day. What usually goes wrong is that the schedule is silent on conflict, so each disagreement is settled by whoever happens to be in the room.

  6. Step 6

    Run one live decision through it

    Choose a real decision due in the next month, ideally one that sits close to a line, and walk it through the schedule with both people present. Note where the schedule gave a clear answer and where it did not. Amend the text on the spot where the answer was unclear. What usually goes wrong is that this test is skipped because everyone assumes the document is obvious, and the first real dispute arrives without warning.

  7. Step 7

    Set a review date and triggers

    Put the schedule on the board's annual calendar, and attach a short list of events that force an early review. Useful triggers include a financing round, a new market, a new product line, or a change in the chief executive's team. What usually goes wrong is that the schedule is never reviewed, and by its second year it describes a company that no longer exists.

Where these schedules go wrong

In our experience, two errors recur, although we have no survey to measure how often. The first is a schedule that leaves the founder with everything carrying a long-term consequence. The chief executive then runs daily operations, and nothing that matters to investors is hers to decide. The second error is the reverse. The chief executive gets the capital decisions while the founder keeps the customers, the key hires and the product. The chief executive then answers for results she cannot shape, and the founder can still overrule her in public.

Investors will read the schedule too. Condon, writing in the Washington Law Review in 2020, argues that institutional investors can influence firm-level managerial decisions for the benefit of a broader portfolio. A company with institutional shareholders should expect them to ask who approves what. The founder-chair should be able to answer from a page rather than from memory. A written line also protects the founder, because it shows which decisions were theirs to keep and which were delegated on purpose.

The third error is the exception that becomes a habit. The founder approves one contract out of hours, the next one follows, and within a year the schedule describes a practice nobody follows. Keep a short log of every exception, with the date and the reason given. If the same exception appears twice, change the schedule rather than tolerate the gap. A recorded exception is a signal to the board. An unrecorded one becomes a precedent without anyone deciding it.

Do we need outside advice before we adopt this?

Yes, for the parts that touch the company's articles, any shareholder agreement and the employment terms of both people. A company lawyer should check the schedule against those documents before the board adopts it. This guide does not interpret those documents. A governance code in your jurisdiction may set requirements that we have not checked, and you should ask your adviser to confirm them.

What if the founder will not write their own limits?

Treat the refusal as information and take it to the board. The founder can decline to draft their own limits, but the board can still set them. A schedule with only one side written is not finished, and the board should say so in its minutes. Make the decision in the room, not in a corridor afterwards.

Should the schedule be shared outside the board?

That depends on the company, its shareholder documents and the people who ask for it. Decide the answer before a dispute, not during one. Whatever the decision, the directors and the two people named in the schedule should have the same version, and the version should carry a date.

How will we know whether it is working?

Count the decisions that reach the board in the first two quarters, and read the exceptions log. If the chief executive escalates routine matters, the limits are set too low. If the founder still signs off routine matters, the schedule is not being used. Either finding is a reason to amend the schedule, not to abandon it.

This week, find the current delegation of authority, if one exists, and read it against last month's approvals. Then ask the founder, in private, which three decisions they would find hardest to stop making. Put the schedule on the next board agenda, with a date for a first draft and a named director to hold the pen. Before any vote, ask a company lawyer to check the draft against the articles and any shareholder agreement. Start with the list of approvals. The line itself can wait a day.

Newsletter

The Briefing

One email a week: the leaders we profiled, the ideas worth stealing, and nothing else.

Weekly, free, one click to leave. Read our privacy policy.