← All posts

The second-generation operator's first 100 days

Spend the first 100 days after taking over a family business learning before changing anything. In days 1 to 30, map how money, stock and orders actually move, and who holds which part in their head. In days 31 to 60, count what the current way costs in hours and errors. In days 61 to 100, fix one thing that is small, visible and reversible. Leave payroll, long customer arrangements and the founder's key people alone until you can explain why they work.

Illustration: six tiles on a grid, a key, a group of people, a notebook and pen, a calendar, a document and a chat bubble, with a key highlighted in olive.

Why do the first 100 days matter so much for a family successor?

Because the research starts out against you, and the first months are when people decide whether it was right. A study of Danish firms by Bennedsen, Nielsen, Pérez-González and Wolfenzon found that when a family member takes over as CEO, operating profitability on assets falls by at least four percentage points around the transition. The gap was largest in fast-growing industries and relatively large firms.

A second study points at where the gap comes from. Bloom and Van Reenen surveyed management practices at 732 medium-sized manufacturers in the US, France, Germany and the UK. Better practices were strongly associated with productivity, profitability, growth and survival, and poor practices were more common where family-owned firms passed management control to the eldest son.

Read those carefully before taking them personally. Both are averages from particular countries and decades, and neither describes your company. What they do describe is a pattern: the successor inherits authority without inheriting the founder's knowledge, and the business pays for the difference while it is being closed. The second study measured management practice, which is learned, not inherited. That is the useful part.

So the first 100 days are not for proving you have ideas. Everyone already assumes you have ideas. They are for closing the knowledge gap faster than the average successor does, and for being seen to close it.

What should you leave alone at first?

Leave alone anything that works and that you cannot yet explain. In a business that has run for twenty or thirty years on memory, the arrangements that look strangest from outside are often the ones holding weight: the customer who gets sixty days instead of thirty, the clerk who re-checks every delivery receipt by hand, the cash that goes to the bank on Tuesdays and not Mondays. Each one exists because something went wrong once.

None of these is permanent. Some will turn out to be waste. The rule is about order: you change a thing after you know why it exists, not before. Until then, the things to leave untouched are these.

  • ·Payroll and its timing. People forgive a successor many things. A late or wrong pay run in the first quarter is not one of them.
  • ·Credit terms and pricing for long customers. These were often set in a conversation between the founder and the customer's founder. Changing them without that history reads as a breach, not a policy.
  • ·The person who reconciles. Every business of this size has someone who makes the numbers agree at month-end. Until you know how they do it, they are irreplaceable, and they know it.
  • ·Informal workarounds that cost time. The double-checking, the extra phone call, the notebook kept beside the system. Find out what each one prevents before you call it waste.
  • ·Reporting lines. Who people call when something goes wrong is the real organization chart. Redrawing the official one does not change who they call.

What should you map in the first 30 days?

Map three flows end to end: how an order becomes cash, how stock moves from supplier to customer, and where each step depends on one person's memory. Do it by walking the work, not by asking for a presentation. The version people describe in a meeting is the version they think you want; the version on the floor is the one that runs the business.

Thirty days is enough for a business of 80 to 500 people if you spend most of it outside your office. Write everything down as you go, in plain words, with the names of the people who do each step.

  1. Spend at least one full day with each function that touches money or stock: sales, warehouse or depot, purchasing, accounts.
  2. Follow one order and one item end to end, and write each step down with the name of the person who does it.
  3. Mark every place the same information is typed or written a second time.
  4. Mark every step that only one person can do.
  5. Ask the founder, in a sit-down, to explain the three or four exceptions everyone else mentioned and could not explain.

Follow the money from order to cash

Pick one ordinary order and follow it. Where does it arrive: a Viber message, a phone call, a sales rep's notebook, a purchase order by email? Who writes it down, where, and how many times is it typed again before an invoice exists? Who decides whether the customer is allowed more credit? When the payment arrives, who matches it to the invoice, and what happens when it does not match?

Every place the same information is typed a second time is a place it can change. Mark those. They are usually where the month-end arguments start.

Follow the stock from supplier to customer

Do the same for one item. Who orders it and how do they know when to? Who receives it and checks the count? Where is it recorded, in a bin card, a spreadsheet, a system, or not at all? How does it move between locations, and who records the transfer at each end? What happens to returns, samples and breakage?

Stock is where a family business carries most of its cash, and it is usually the least documented flow. If the count and the record disagree, you do not need to know why yet. You need to know that they do, and by how much.

Find what lives in one head

For each step you mapped, ask one question: if this person were away for two weeks, who would do it, and how would they know how? The answers sort into three groups. Steps anyone can do. Steps one other person could do with help. Steps nobody else can do.

The third group is your real risk register. It usually includes the founder, which is the hardest conversation of the hundred days and the most valuable one to have early, while the founder is still around to answer.

How do you find out what the current way actually costs?

Count it, in hours and in errors, before you form a view of what to fix. Three counts matter most: hours spent re-typing information that already exists somewhere, hours spent reconciling records that should agree, and the number of steps that depend on one person. Multiply the hours by the loaded wage of the people doing them and you have a cost per year that nobody can argue with.

I think counting first is the single most useful habit a successor can bring, because it changes the conversation from opinion to arithmetic. The founder ran the business on judgment and it worked. You cannot win an argument against twenty years of judgment with a newer opinion. You can win it with a number the founder's own people helped you count.

Keep the method simple. For two weeks, ask the people doing each repeated task to tally how long it takes, on paper if that is what they use. Do not estimate for them. Estimates from the top are always too low, because the person estimating does not do the work. The hidden hours audit sets out the full method if you want it.

  • ·Re-keying hours: each time information is typed again, how long it takes, how often, by whom.
  • ·Reconciliation hours: time spent making two records agree, at month-end and in between.
  • ·Error count: wrong deliveries, wrong invoices, stock adjustments, credit notes. A tally per week is enough.
  • ·Single points of failure: steps from your 30-day map that only one person can do.
  • ·Waiting time: how long an order, approval or payment sits before anyone acts on it.

A worked example with your own numbers

Take one task from your map, say the daily re-typing of orders from messages into the invoicing system. Write down the minutes it takes each day, multiply by the working days in your year, divide by 60 for hours, and multiply by the hourly cost of the person doing it, including benefits and the 13th month where that applies. Then add the cost of the errors it causes: each wrong invoice, multiplied by the time it takes to correct it.

Do that for the five biggest items on your list and add them up. Whatever figure comes out is yours, measured, and it is the baseline every later decision will be judged against. Without it, nobody will ever know whether anything you changed worked.

What is the first thing you should actually fix?

Fix one thing that is small, visible to the people who do the work, and reversible if you are wrong. The best first fix usually removes a re-key or a recurring argument, and makes someone's week easier in a way they notice. It should take weeks, not months, and cost little or nothing. The point is less the fix itself than showing that change under you means less friction, not more.

Good first fixes tend to look boring. A controlled list of customer names so the same customer is no longer spelled four ways. One named owner for the stock count at each location. A written procedure for the three exceptions only the founder could explain. One message format for orders, so they stop arriving in six different shapes.

My position on order here is fixed, and it is the spine of how we work: fix and document the process before automating it. Automating a broken process makes it faster, not better, and now you have the broken process and a system to maintain. Systems and automation are a sequence, not a choice between them.

  • ·Small: one process, one location, one team.
  • ·Visible: the people doing the work feel the difference within a fortnight.
  • ·Reversible: if it is wrong, you can put the old way back by Friday.
  • ·Counted: you measured the before in days 31 to 60, so you can show the after.

How do you win over people who worked for your father?

Ask before you decide, credit them out loud when their knowledge shapes a change, and let them keep authority over what they already run well. Long-serving staff are usually not resisting change. They are protecting arrangements they built and that worked, and they are watching to see whether you understand why they worked. The 30-day map is how you show that you do.

There is evidence that trust and delegation travel together. In a survey of almost 4,000 firms in the US, Europe and Asia, Bloom, Sadun and Van Reenen found that trust and the rule of law facilitate decentralization by improving cooperation, and that decentralization is complementary with information technology. That study measured trust across regions and countries, not inside one family business, so treat it as a direction rather than a rule. The direction is clear enough: a business where people trust each other can hand decisions down, and a business that can hand decisions down gets more from its systems.

In practice, the successor's job is to make delegation safe. That means writing down what each person decides on their own, what they decide and report, and what they bring to you. It means saying the same thing to the founder's people as to your own hires. And it means not discovering, in your first quarter, that you are checking everyone's work because you do not yet know who to trust.

  • ·Interview each long-serving manager alone, with one question: what would you change if it were your decision?
  • ·When one of their answers shapes a fix, say so in front of the team.
  • ·Do not bring in a new hire to run something an existing person already runs well.
  • ·Agree with the founder, privately and early, which decisions are now yours and which are still theirs, and tell everyone the same version.

Should you buy new software in your first 100 days?

Almost never. Buying software in the first 100 days means choosing a system before you know which problem it is for. You will not yet know which processes are sound and only badly recorded, which are genuinely broken, and which are the one thing the business does differently. A vendor demo will happily answer that question for you, and the answer will be its product.

I think most software evaluations in this size band start about two years early and get pointed at the wrong problem. A successor under pressure to modernize is the most likely person to start one. Resist it until the map and the count are done. If the count shows that several locations keep separate copies of the same record and argue every month-end, that is a structural problem and buying is probably right. If it shows that the process is fine and only the record broke, fixing is cheaper, faster and reversible. Build, buy, or fix what you have sets out how to tell the difference.

If you inherit a half-finished app or an abandoned system, do not restart it in the first 100 days either. Add it to the map: what was it meant to fix, who used it, and why did they stop. That last answer is usually the most useful thing you will learn about the business all quarter.

What does the 100-day plan look like on one page?

Four phases, each with a deliverable you can show the founder and the team. The plan below is a sequence, not a calendar to hit to the day. If the map takes six weeks instead of four because the business is bigger or the founder is harder to pin down, let it take six. Skipping the map to stay on schedule is the one shortcut that costs more than it saves.

  • ·Can you name every step in order-to-cash, and who does it?
  • ·Do you know which steps only one person can do?
  • ·Do you have a cost per year for the current way, counted by the people who do the work?
  • ·Has one fix landed that the team noticed and did not resent?
  • ·Could the founder read your one-page list and agree with the diagnosis, even if not every remedy?
  1. Days 1 to 10: listen. Meet every manager alone. Ask what works, what breaks, and what they would change. Change nothing. Deliverable: a list of names, roles and the one problem each person raised.
  2. Days 11 to 30: map. Follow one order and one item end to end. Mark re-keys, reconciliations and one-person steps. Sit down with the founder on the exceptions. Deliverable: the three flows written in plain words, with names.
  3. Days 31 to 60: count. Have people tally their own time on repeated tasks for two weeks. Count errors. Multiply by loaded wages. Deliverable: a cost per year for the five biggest items, which is your baseline.
  4. Days 61 to 90: fix one thing. Pick the item that is small, visible and reversible. Write the new procedure down before changing anything. Make the change with the people who do the work. Deliverable: the fix, and the before-and-after count.
  5. Days 91 to 100: decide the next year. With the map and the count in hand, sort every remaining problem into buy, build or fix. Deliverable: a one-page list, in order, that you show the founder and the managers who helped.

Common questions

Is 100 days the right length of time?

It is a planning convention, not a rule. What matters is the order: listen, map, count, fix one thing, then decide. A business with several locations or a founder who is hard to schedule may need longer for the map. Stretching the map is fine; skipping it is not.

What if the founder is still in the business?

That is an advantage for the map, because the person who knows the exceptions is still there to explain them. Agree early and privately which decisions are now yours, tell the team the same version, and use the founder's time on the questions only they can answer.

What if something is clearly broken on day one?

Fix it if it is urgent and small: a payroll error, a safety issue, a customer about to leave. Then go back to the plan. An emergency fix is not the same as a change to how the business runs, and the team can tell the difference.

Do family successors really perform worse than outside CEOs?

On average, in the studies available, yes. A Danish study found operating profitability on assets fell by at least four percentage points around family CEO transitions, and a four-country survey of manufacturers linked eldest-son succession with poorer management practices. Both are averages. The second measured management practice, which can be learned.

Sources

If this is the problem you're sitting with, show us the task and we'll audit one workflow for free.

Show us the task

One email, now and then

Where the hours actually go

Every new post, and nothing else. No pitch, no schedule to keep up with, one click to leave.