The pattern almost every successful family and founder-led business hits — and a working diagnosis of how to redesign decision rights without losing control of the business.
The founder approves every purchase over a threshold. The founder signs off every discount above a certain margin. The founder is copied on every appointment above a certain level. The founder resolves every escalation the operations team cannot close. On paper the delegation exists. In practice, when a decision genuinely needs to be made — a customer wants a variation, a supplier changes terms, a large invoice arrives — everyone waits for the founder. The senior team is capable. They have chosen not to decide, because the last three times they decided without asking, the founder had a view they should have known about. Now they ask. The founder is the approval workflow, and the business runs at the founder's speed.
It happens because the business grew and the operating model did not. In the early years the founder had to be in every decision because the founder was the only person who understood the business well enough. That habit produced the business's success. When the business hit 200, 500, 1,000 employees, the habit did not scale. The senior team learned that the fastest path to a correct decision was to ask, not to decide. The founder learned that the fastest path to a correct outcome was to be asked, not to teach. Both parties are optimising for accuracy over autonomy — and the aggregate is a business that cannot move faster than one person. PwC's Family Business Survey and INSEAD's Family Business Network have documented the pattern across markets; the specific version varies but the shape is the same.
The obvious cost is speed. Every decision waits in a queue behind the founder's day. The second cost is opportunity — decisions that should have been made never are, because they never quite made it to the top of the queue. The third cost is talent — senior people who were hired to run the business discover they were hired to prepare decisions for someone else to make, and they leave. The fourth cost is succession risk — the business is entirely dependent on the presence of one person, and the leadership pipeline has never had the practice it needs. The fifth cost is a founder who cannot take a two-week holiday without leaving the business unable to close.
Every workflow-automation platform, approval-management tool and process-management system routes approvals — routes them to a person. If that person is the founder for anything meaningful, the software is a faster path to the same bottleneck. Tools do not fix decision rights. Tools implement decision rights. If the decision-rights design keeps sending every consequential approval to the founder, no amount of workflow software will change the founder's day.
The founder's judgement. On the decisions that genuinely require it — capital allocation, senior appointments, exceptional risk — the founder's judgement is what has produced the business's success. Do not attempt to systematise that judgement. Also do not replace the senior team; the point of the redesign is to give them decisions to make, not to add another layer above them. The redesign is about which decisions belong at which level, not about removing anyone.
The definition of which decisions go where. The right level of a decision is the lowest level at which it can be made competently, with the information and authority the decider needs to make it well. Most family and founder-led businesses have never named this explicitly. Once named, the pattern becomes clear: perhaps four categories of consequential decision belong with the founder, perhaps twelve belong with the senior team, and perhaps forty belong with the operational leaders. Reengineer the operating model so those decisions actually go to the right level — with the information, authority and accountability the decider needs.
The information the decider needs at each level. A senior team member cannot make a pricing decision competently without visibility of margin, capacity and customer history. An operational leader cannot make an inventory decision without visibility of the demand pipeline and current stock. Most decision-rights failures in mid-market family businesses are actually information-rights failures — the decider does not have what they need to decide, so the decision goes up. Integrate the systems so the information sits with the person doing the deciding.
The routine approvals. Purchase requests within budget and within policy. Discount approvals within pre-agreed bands. Standard employment actions. Standard vendor onboarding. Somewhere between 40 and 70 per cent of the approvals that currently reach the founder in a mid-market business fall into this category. Automating them removes the founder from the queue without removing the founder's authority — the policy that governs the automation is set by the founder, and exceptions still escalate.
AI helps at the exception layer. The routine decisions get automated by rule-based workflow. The consequential decisions still go to a human. The interesting middle — decisions that look routine but have unusual signals attached — is where AI can flag *this one needs a person* without a human having to look at everything to find the ones that matter. Used this way, AI extends the operating model rather than replacing decision-makers.
The redesign does not need to happen everywhere at once. Start with one function — often finance approvals or procurement — because they have the highest volume and the clearest policies. Redesign the decision rights, automate the routine, expose the exception path, and let the founder experience the change on a low-risk surface. When the pattern proves itself, extend it to the next function. A common twelve-month path looks like: months 1–2 name the decision categories; months 3–5 redesign one function; months 6–9 automate the routine layer and expose exceptions; months 10–12 extend to the second function. The founder is in the loop throughout — but progressively on the decisions that actually require the founder's judgement.
This work is more organisational than technical. What is needed is senior operating experience — someone who has redesigned decision rights in a family or founder-led business before, and who understands the political weather that comes with the redesign — plus enough technology capability to configure the workflow tools that already exist. It does not require a large internal IT organisation. It requires a small, senior team that can hold the redesign steady while the leadership team adapts.
A Jettifi engagement of this class begins with a decision-rights diagnostic. Which decisions currently reach the founder? Which of them require the founder's judgement? Which do not? What information would a senior operator need to make each one competently? Senior Jettifi core leads the diagnostic and the redesign; specialists configure the workflow and integration layer that supports the redesigned decision rights. The engagement is deliberately sensitive — the founder's habits are the business's success, and the redesign has to respect that. The engagement ends when the founder is back to making the decisions that require the founder's judgement, and the business is moving at the speed the market requires.
Discuss decision-rights in a family or founder-led business
Talk to Jettifi →For the underlying working paper, read the Digital Reengineering pillar or the executive edition of The Guide.