Every opening I have run, and I have now been through more than a dozen across the Hana Group franchise footprint and later at Zareen's, has had the same fight in the same week. The buildout finishes late. The permits arrive late. The training week compresses. And the soft open, which is supposed to be the calm week where you prove the building works, becomes the week where you find out what you missed.
The soft open week is not a rehearsal. It is a controlled service. What you are testing is the entire operating system, from the moment the first prep cook clocks in to the moment the closing manager signs off, running under real load with real tickets and real inventory. If you treat it as a rehearsal, you get a rehearsal quality opening, and rehearsal quality openings cost you the first three months of reviews.
What follows is the hour-by-hour calendar I run now. It is not the calendar I ran the first time. The first time I gave the team one staff run and one friends-and-family night and called it good. That opening was rough. Every version since has been longer, tighter, and more boring, and every version since has opened cleaner.
The seven-day map
Before the hour-by-hour, the shape of the week. Seven days, five distinct phases, each with a specific job. If you compress a phase, know which one you are compressing and why.
Fig. 1 · The seven-day soft open week, phased.
Day minus 7 and minus 6: walkthrough, deep clean, stage the line
The first two days are not about cooking. They are about proving the building works.
Day minus 7, 8:00 AM: general contractor walkthrough
The GC and the area operations lead walk every square foot with the punchlist in hand. Every item gets a photo, a responsible party, and a date by which it will be resolved. Do not accept verbal promises. If it is not on the list with a name and a date, it does not exist. At Zareen's Palo Alto we found a hot water pressure issue at this walkthrough that would have shut down our dish pit on night one. Catch cost: 15 minutes. Miss cost: a night of comped tickets.
Day minus 7, 1:00 PM: deep clean, top to bottom
The buildout crew leaves drywall dust in places you will not think to look. Behind the hoods, on top of the walk-in, inside the ice bins. A four-person crew for six hours. If you cook on top of construction dust, you will taste it, and so will the health inspector.
Day minus 6, 8:00 AM: stage every station
The chef and the sous stage every station with the tools and mise it will hold during service. No cooking. Just proving that every station has what it needs, that every drawer opens without hitting another drawer, that the reach-ins are placed so line cooks are not crossing paths during a rush. This is boring work. This is also the work that determines whether your ticket time on night one is 14 minutes or 24 minutes.
Day minus 6, 2:00 PM: POS load and modifier test
The full menu goes into the POS. Every item. Every modifier. Every allergen tag. Then the front-of-house lead rings 20 fake tickets across the full menu while the chef watches the kitchen printer. Nine times out of ten this session finds three or four missing modifiers that the kitchen cannot execute against. Fix them now. Not on night one.
Day minus 5 and minus 4: menu load, station drills, first service run
Day minus 5, 9:00 AM: prep and pars
Full prep for a full-service day. The sous cooks against the pars the way they would on a live Tuesday. If the pars are wrong, you find out today, when the correction is a phone call to the produce vendor, not on Saturday night when the correction is 86ing a dish.
Day minus 5, 5:00 PM: station drills
Each station lead walks their team through the flow of a full ticket, from the moment it prints to the moment it hits the pass. No guests. Just the motion. The expo calls tickets aloud. The runner calls the routes aloud. Everyone hears the whole loop. This is where you catch that your expo cannot see the salad station from the pass, or that your runner path crosses the bar service lane at exactly the wrong angle.
Day minus 4, 5:00 PM: first staff service run, shortened menu
Staff cooks for staff. About 20 covers, seated in three waves, ordering off a menu cut down to two-thirds of the full offering. Timed tickets from the moment they print to the moment they hit the table. Voids and remakes tracked on a whiteboard by the expo. Debrief at 9:00 PM, at the bar, with a beer. Every station lead names one thing that worked and one thing that broke. The general manager takes notes. You listen more than you talk.
Day minus 3: full staff service run, full menu
This is the day that matters most. Full menu, full team, full flow. Around 40 to 50 staff covers, seated the way real guests would be seated, ordering across the full menu, paying nothing but paying attention.
The metrics you are watching, in this order:
- Average ticket time from fire to pass. Target under 18 minutes. If you are at 25, you have a station bottleneck and you need to find it before day zero.
- Void and remake rate. Target under 5 percent tonight. It will drop from here. If it is at 15, something is fundamentally wrong with either the recipes or the POS build.
- Dish pit turnaround. Are clean plates hitting the line before they are called for? If not, the mid-service push will collapse on you.
- Closing time from last ticket to lights out. The closing checklist has to be run to completion tonight, not skipped because the team is tired. If you skip it now, you will skip it on night one.
The point of the staff service run is not to prove you can cook. It is to prove that ten people can cook the same ticket the same way, twice in a row, without talking to each other.
Day minus 2 and minus 1: friends and family, then vendors and neighbors
Day minus 2, 5:30 PM and 7:30 PM: friends and family
Two seatings of 25 to 30 covers each. Comped food. Cash bar. Printed feedback cards at every seat with four questions: What did you order, how was the timing, what would you send back, what would you bring a friend for? The general manager works the floor. The chef works the pass. I sit at the bar and watch, and I do not intervene unless something is on fire.
Debrief that same night at 10:30 PM. Not the next morning. The lessons decay by breakfast. Every station lead reports out. Every feedback card gets read aloud. The next day's shift meeting starts with the top three fixes.
Day minus 1, 5:30 PM and 7:30 PM: vendors and neighbors
Same format, different crowd. Invite your produce vendor, your linen vendor, your linen driver, the shop owners on your block, the property manager, the leasing agent, and any local operators who have been generous with advice. This is not a marketing night. This is a goodwill night, and goodwill compounds in a way marketing dollars never do. At the Zareen's Redwood City open, our produce vendor caught a plating issue that our whole team had missed for two nights. He mentioned it once, we fixed it that night, and it stayed fixed.
Day zero and beyond: capped paid service
Day zero, 11:00 AM: pre-open all-hands
Fifteen minutes. General manager runs it. Three points. Here is what worked at friends-and-family. Here is what we are still fixing. Here is the cover cap for tonight and here is what we do if we hit it. Then the team goes to setup and does not think about the meeting again.
Day zero, 5:00 PM to close: first paying guests, hard cover cap
The cover cap for night one is 60 percent of your target full-service volume. Non-negotiable. When you hit the cap you close the reservation book and you politely turn away walk-ins. This will feel like leaving money on the table. It is. You are trading that money for a clean review body and a team that does not get buried on night one.
Fig. 2 · Cover cap ramp, first eight nights of paid service.
Nights two and three stay at 60 percent. Night four goes to 80 percent if and only if two conditions clear on night three. Ticket time held under 18 minutes and voids ran under 3 percent. If either misses, hold the cap another night. Nights five through seven step up in 10-point increments until you hit full volume. By the end of the first paid week the team should have run a full-volume service twice, clean.
Where soft opens usually break
Across the openings I have been through, the failure modes cluster. Four places account for probably 80 percent of the pain, and all four are cheap to fix during the staff run and expensive to fix after.
- The POS is missing modifiers the kitchen needs. Every time. Test the full menu, ring the full menu, catch the misses.
- The dish pit cannot keep up with the mid-service push. Almost every time. Watch the dish-return rack at minute 45 of the staff run. If it is stacked, you have a problem.
- Expo has no clean handoff protocol between kitchen and floor. Runners freelance. Tickets sit. Food goes out cold. Write the protocol, drill it in the station walkthrough.
- The closing checklist has never actually been run to completion. The first time it is run is not the night of the open. Do it on the staff service run and again on friends-and-family.
The point
A soft open week is not the week you find out if the restaurant is good. It is the week you find out if the operating system holds. The food will be good if the team is good. The system holds only if you have proven it holds, under load, with real tickets, and with the closing checklist run to completion, before a single paying guest sits down at your host stand.
Every opening I have run cleanly ran on this seven-day calendar. Every opening I have run poorly compressed one of the phases and paid for it in the first month of reviews. There is no glamour in a boring, well-paced soft open week. There is a lot of glamour in a soft open week that saves you a year of dug-out reputation.
Cadence beats charisma. Especially on opening week.