The Transition From Three Locations to Five

At Zareen's I lived this transition personally. The group had three Michelin-recognized Bay Area locations doing well, and the plan was to grow to five. Two openings, roughly 24 months, with the fourth stabilizing before the fifth was signed. On the surface it was two more units. In practice it was the transformation of a founder-run operation into a real multi-unit business, and the work of building that business took longer than the buildouts did.

Here is what happens between three and five, and what has to be built to make it stick.

Why three to five is the transition

At one and two units, the founder-operator runs everything. Every GM reports to them. Every hire is approved. Every menu change is decided in the room. This works because the founder can physically be in each unit multiple times a week and can hold the whole operation in their head.

At three units, this model starts to strain. The founder is in each unit maybe twice a week. Small drifts happen because the founder is not there to catch them. But the model still basically works.

At four units, the model breaks. The founder cannot physically be in each unit often enough to catch drift. GMs start making decisions the founder would have made differently, because the founder was not there when the decision came up. Standard begins to fragment.

At five units, the model is unsustainable. The founder either burns out trying to hold everything together, or a critical unit starts to drift while attention is elsewhere. The group's contribution margins begin to compress and nobody can identify a single cause because there are five simultaneous causes and only one person trying to see all of them.

The operations layer

The single most important build in the three-to-five transition is adding an operations layer between the GMs and the founder. Titles vary. Area director. Director of operations. Regional manager. What matters is the function: a single person who directly supervises the GMs on day-to-day operations, freeing the founder to focus on brand, capital, and strategy.

Where the person comes from:

• Promoted internally: Best case. Your strongest GM steps up to run the group. They know the brand, know the team, know the systems. Risk: you now need a new GM for the unit they left.
• Hired externally: Sometimes necessary. Look for someone with multi-unit operations experience in a similar concept. Risk: 6-12 months of onboarding before they are fully effective.
• Founder plays both roles temporarily: Sometimes required. Not sustainable. If the founder is still playing the operations layer role at month 18 after the fifth opens, something is wrong.

When to add the role: somewhere between the third and fourth opening. Adding it after the fifth is opening is too late. Adding it before the third is unnecessary overhead. The right window is when the founder can feel the strain but before the strain has caused visible drift.

The operating cadence has to become formal

At three units, the operating rhythm is often informal. The founder texts the GMs, drops by, has coffee meetings. This is fine for three. It does not scale.

Between three and five, the cadence has to become formal and repeatable. Weekly and monthly, with a standing agenda that runs whether the founder is in the room or not:

• Weekly P&L review with each GM: 45 minutes, same time every week, same agenda. Owned by the area director once that role exists.
• Monthly regional operating review: All GMs together, plus the area director and founder. Each GM presents their P&L and next month's plan in five minutes.
• Quarterly business review: Deeper dive on trends, initiatives, capital allocation. Founder, area director, culinary lead, catering lead.
• Annual planning: Full team, three-day retreat if possible. Sets the year's operating targets and capital plan.

None of this is exotic. All of it has to actually happen, on schedule, without exception. Cadence beats charisma. What holds a five-unit group together is the standing meeting, not the founder's personal attention.

Shared services need to exist

At three units, most administrative functions can be improvised. The founder does bookkeeping in the evenings. HR is handled ad hoc. Marketing is unit-level. This does not scale.

Between three and five, four shared services need to become real:

Accounting

Move from cash-basis to accrual accounting. Hire a bookkeeper or use a professional bookkeeping service. Move to a real chart of accounts. Get monthly financial statements produced within 15 days of month-end. If the founder is still doing the books at five units, the books are not accurate and the decisions being made from them are worse than the raw data suggests.

HR and payroll

At two to three units, a PEO or payroll service handles most of what you need. At five units, you need someone (part time or full time) who owns HR policy, employee handbook, onboarding, offboarding, benefits administration, and compliance across all units. HR issues at five units happen weekly. Someone needs to own them.

Marketing

Unit-level marketing does not scale into a coherent brand at five units. A part-time or fractional marketing lead becomes necessary. Manages the brand voice, the social calendar, the local marketing at each unit, and the email list across the group.

Technology

At three units the POS system, scheduling system, and reporting system can be sometimes-integrated. At five, the tech stack has to actually integrate. Toast plus 7shifts plus a Power BI-style reporting layer, or the equivalent. Data flows across units into one dashboard. Otherwise the area director is comparing apples to oranges every week.

Capital gets more sophisticated

The capital story changes between three and five. A three-unit group can often fund the fourth from cash flow. The fifth usually requires debt or outside capital. This is when SBA loans, bank debt, or family office partnerships enter the story.

The founder's relationship with money changes too. At three units, cash management is personal. At five units, it has to become institutional. Rolling 13-week cash flow forecasts. Bank line of credit. Real reserve targets. Monthly balance sheet review. This is the moment the group becomes financially institutional, not just operationally.

The Zareen's story

At Zareen's, the transition from three to five looked roughly like this. When I came in as fractional head of operations, the group was three units, all Michelin-recognized, all with distinct operating characters that had drifted despite serving the same brand. The immediate work was the turnaround of the three, which I have written about elsewhere. That work stabilized the group and recovered roughly $4.9M in operating profit.

The four-unit and five-unit openings happened in the 18 months following the turnaround stabilization. Between them we:

• Added an area director role and hired into it from a senior GM's promotion
• Installed weekly one-on-ones, monthly regional operating reviews, and quarterly deep dives
• Moved accounting from cash to accrual and hired a full-time bookkeeper
• Formalized the training crew model and the promotion path
• Built out a real corporate catering operation with dedicated account management for Stanford, Google, Apple, Meta, LinkedIn, Salesforce, Cisco, Adobe, and Nvidia
• Upgraded the tech stack to give the area director real cross-unit visibility

The fourth opened on plan. The fifth opened four months later than the pitch deck said it would, because we prioritized the organizational build over the opening date. That trade paid back within a year. Both new units were profitable at contribution level by month six. Group revenue grew from $18M to $30M across the two years. Employee count went from about 130 to 215. Enterprise catering, which had barely existed at three units, became a distinct revenue channel with its own leader and its own targets.

The physical openings are the visible part of the transition. The organizational build is the invisible part, and it is the part that makes the difference between five units that work and five units that drag each other down.

What to build in order

If I were doing this again, the sequence I would run:

1. Months 1-6: Formalize the operating cadence at the three existing units. Weekly P&L review. Monthly regional operating review. The rhythm has to work at three before it can work at five.
2. Months 6-9: Identify the area director. Internal promotion if possible, external hire if necessary. Begin their transition into the role.
3. Months 9-12: Stand up accounting and HR shared services. Upgrade the tech stack for cross-unit visibility.
4. Months 12-15: Open the fourth unit. Area director is running weekly management. Founder is available but not primary.
5. Months 15-24: Stabilize the fourth, plan the fifth. Ensure the operating rhythm is holding across all four units.
6. Months 24-30: Open the fifth unit. By this point the organizational build is complete enough to absorb it without breaking.

What breaks if you skip the build

Groups that open the fourth and fifth units without the organizational build have a consistent failure pattern. The new units stabilize. Nine months later, one of the older units has quietly drifted into the same distress that the founder used to catch in real time. The GM of that unit resigns because they have not gotten leadership attention. By month 15, the group is a mix of a few strong units and a few struggling ones, and the founder is spending most of their time trying to save the struggling ones instead of building the business.

I have seen this exactly this shape multiple times across different operators. The pattern is not about talent or effort. It is about the fact that solo attention does not scale, and the operations layer plus formal cadence is the specific tool that solves for it. Skip the tool, get the pattern.

The point

Three to five is the transition where a restaurant group either becomes a real business or gets stuck. The physical openings are visible and manageable. The organizational build is invisible and much harder. Add the operations layer between the fourth and fifth opening. Formalize the operating cadence. Build shared services. Upgrade the tech stack. Do it in the year before you open the fifth, not after.

Done right, five units is where the group has real leverage. Contribution margins improve. Bench depth grows. Catering scales. The founder gets their time back to build the next phase. Done wrong, five units is where a good group starts to unravel. The difference is whether the operating system scaled with the unit count. Cadence beats charisma. Build the machine.