Ask a room of restaurant owners what they’d do differently and almost nobody says they picked the wrong food. The regrets cluster somewhere else entirely: not enough cash, too many menu items, a hiring process invented on the fly, and a habit of reading the numbers once a month when the numbers were moving weekly.
That’s the useful thing about restaurateurs advice collected after the fact. It lands on a handful of structural decisions that set the ceiling on everything else, and most of them get made in the first ninety days, when the owner is too busy to think about them.
This is written for someone one or two years in, open and trading and discovering which of their assumptions were wrong. If you’re still earlier than that and the concept itself is the open question, our guide to restaurant concepts for delivery that actually work deals with that decision directly.
The Bottom Line
- Hold more cash than the plan says. Slow months arrive before the reputation does, and running out of working capital kills more restaurants than bad food.
- The failure rate is not 90%. Roughly 17% of restaurants close in year one, and around half reach five years, which is comparable to small businesses generally.
- Your menu is probably too long. Every item is another ingredient to hold, spoil and train on, and shortening the menu fixes food cost, waste and consistency at once.
- Turnover is the permanent condition, not a crisis. Restaurant turnover ran about 66% in 2025, so build hiring and training as a standing system rather than an emergency response.
- Read prime cost weekly. Food plus labour above 65% of revenue means the business doesn’t work, and a month is far too long to find out.
Boost customer satisfaction with just a few clicks
Most-Loved Features:
- On-demand drivers
- Real-time GPS tracking
- Delivery confirmation photos
- Over 50% of customers report a smoother delivery experience
Keep far more cash than the business plan says you need
The advice most repeated by people who nearly didn’t make it: secure enough operating cash to cover roughly twelve months of rent, payroll and supplies before you open, assuming sales come in slowly. Not because they will, necessarily, but because the restaurants that die in year two mostly die with reasonable revenue and no working capital.
The numbers behind this are less terrifying than the folklore. The widely repeated claim that 90% of restaurants fail in year one is false; measured first-year closure runs closer to 17%, with roughly 83% surviving year one and about half reaching five years (Owner, retrieved 2026-11-03). Survival odds are roughly those of American small businesses in general.
What that reframing changes is the diagnosis. You don’t need to survive a one-in-ten lottery. You need to survive a slow February, a broken walk-in, a health department re-inspection and a chef quitting in the same quarter. That is a cash question.
Three specific habits from operators who learned this the hard way:
- Separate the opening budget from the operating reserve. Money spent on an espresso machine during build-out is money not available in month seven.
- Assume the first three months are a training period you are paying for. Food cost will be high, waste will be high, ticket times will be slow. Budget for it instead of being surprised by it.
- Know your break-even covers per day, as a number, out loud. Owners who can’t say it are flying blind.
Your menu is almost certainly too long
This is the single most common regret, and the most repeated piece of advice in every collection of restaurateurs advice ever published: if your menu runs four pages, you’re trying to do too much. Do a few things extremely well instead of many things adequately.
The reason it matters is not aesthetic. Every additional item adds an ingredient to purchase and hold, a spoilage risk, a prep task, a station to staff, a line in training and a few seconds to every ticket it touches. Full-service restaurants generate more than 43% of all foodservice food surplus, and the menu is the lever with the most pull over that (ReFED, retrieved 2026-11-03).
Cutting the menu improves four things simultaneously: food cost falls because you hold less inventory, waste falls because less spoils, consistency rises because cooks repeat fewer dishes more often, and ticket times drop. Our guide to how to run a green restaurant without hurting margin treats menu length as a waste-reduction tool for exactly this reason, and the savings arrive in food cost before they ever show up in your dumpster.
A short menu also opens doors a long one closes. You cannot source directly from local growers across sixty dishes, but you can across twelve, which is how most restaurants that pull off local sourcing actually do it. The mechanics of that trade, including what it does to your receiving and price stability, are covered in how the farm-to-table movement is changing the industry.
The other half of this advice, equally common: don’t fall in love with your menu. Operators repeatedly describe the dish they refused to cut for sentimental reasons and the months it cost them.
Hiring and keeping people is the job
New owners budget for food and rent and treat staffing as something that will sort itself out. It doesn’t. Restaurant employee turnover ran roughly 66% in 2025 based on BLS data, and accommodation and food services reports annual separation rates well above 70% in most years (Ideal Traits, retrieved 2026-11-03).
That is the baseline condition of the industry, not a sign you’re doing it badly. The mistake is treating each departure as a surprise rather than running a continuous pipeline.
What experienced operators say they’d do from day one:
- Always call references. Mentioned constantly, skipped constantly.
- Train your core team for weeks before opening, not days. Several operators put this at two to three months for key staff. Soft openings exist so your first real service isn’t your first service.
- Write the training down. A documented opening checklist, closing sequence and prep list survives turnover; a trained individual doesn’t.
- Over-communicate schedules. Most avoidable quits trace back to scheduling, not pay.
- Promote from inside where you can. The person who already knows your systems is worth more than the better résumé who doesn’t.
The paperwork mistakes that cost the most
Compliance problems are cheap to prevent and extremely expensive to discover. Operators describe failed inspections, forced closures and insurance gaps that cost weeks of revenue, and almost all of them trace back to a habit nobody set up in month one.
The highest-regret items:
- Treating inspections as events. A kitchen that is ready on any random Tuesday never scrambles. One that cleans for a visit always gets caught eventually. Our four-step routine for staying compliant with restaurant health inspections is the version of this to set up before you need it.
- Labelling and dating casually. The most frequently cited violations in US inspection data are temperature and storage issues, and both are paperwork habits before they are equipment problems.
- Letting licences and certifications lapse. Food handler cards, liquor licence renewals, grease trap servicing, fire suppression inspections. Put every one on a calendar with a reminder.
- Under-insuring. Business interruption cover is the one people wish they’d read properly.
Read prime cost every week, not every month
Prime cost, meaning cost of goods sold plus labour, is the number that tells you whether the business works. Industry guidance puts the healthy band at 55% to 65% of revenue, and labour alone now runs a median of about 36.5% of sales at full-service restaurants (Eagle Rock CFO, retrieved 2026-11-03). Average net margins sit around 3% to 5%.
With a 4% margin, a two-point drift in prime cost is half your profit. That’s why monthly reporting is too slow: by the time a monthly P&L tells you food cost crept up, you’ve run four weeks of it.
The weekly discipline that operators wish they’d started earlier:
- Count inventory on the same day each week, even roughly.
- Pull labour as a percentage of that week’s sales, by day.
- Look at the five highest-volume dishes and recalculate their plate cost every month.
- Write the number on a whiteboard where managers see it.
None of this needs expensive software. It needs a recurring hour nobody is allowed to cancel.
Choose location on traffic data, not on rent
Cheap rent that nobody walks past is the most expensive lease in the business. Operators who regret their location almost always describe choosing it on price, then spending years buying the foot traffic the site didn’t come with.
What to check before signing: actual counts at the hours you’ll trade, not daily averages. Where the parking is, and whether it’s yours after 6pm. What’s on the block that generates evening traffic. Whether a courier can stop legally outside. Whether the kitchen’s existing infrastructure (gas line, hood, grease trap, three-phase power) matches what your concept needs, because retrofitting any of those costs more than a year of the rent you saved.
Train before you open, then pace yourself
The last cluster of regrets is about timing. Owners open too early because the rent clock has started, then run on adrenaline for eighteen months and burn out the team along with themselves.
The advice, from people who did it the hard way: push the opening date rather than open unready, run friends-and-family services until ticket times are real, and build one day a week when you are not in the building at all. A restaurant that only works when the owner is present is not a business yet, and the owners who understood that early are the ones still trading.
What operators say they would change
| What they did | What they wish they’d done | Why it mattered |
|---|---|---|
| Opened with a 60-item menu | Opened with 15 and added slowly | Food cost, waste and ticket times all improved at once |
| Budgeted three months of cash | Budgeted twelve | Survived the slow season instead of refinancing in it |
| Hired to fill a gap | Kept a standing pipeline | Turnover near 66% makes hiring continuous by default |
| Reviewed the P&L monthly | Counted and read prime cost weekly | A two-point drift is half the margin at a 4% net |
| Chose the cheaper lease | Paid for proven foot traffic | Marketing could never fully replace the location |
| Cleaned up before inspections | Ran inspection-ready every day | No closures, no re-inspection fees, no scramble |
Frequently asked questions
What is the most common piece of advice from experienced restaurateurs?
Shorten the menu. It appears in nearly every collection of restaurant owner advice, and it is the one change that improves food cost, waste, consistency and speed of service at the same time. The second most common is to hold far more operating cash than the business plan assumed.
Do 90% of restaurants really fail in the first year?
No. That figure is folklore. Measured first-year closure is closer to 17%, with around 83% of restaurants surviving year one and roughly half reaching five years, which puts them close to American small businesses as a whole. The useful takeaway is that failure is usually slow and cash-driven rather than sudden.
How much cash should a new restaurant have in reserve?
The commonly repeated benchmark is twelve months of fixed operating costs (rent, payroll and supplies) held separately from the build-out budget, on the assumption that sales ramp slowly. Less than six months is where operators report feeling exposed.
What do first-time restaurant owners underestimate most?
Staffing. Finding, training and keeping competent people is described as harder and more time-consuming than anything to do with food, and it stays that way. Owners who built a documented training system and a continuous hiring pipeline in year one spend year three on the business instead of on recruitment.
The pattern underneath all of it
Read enough of this advice and the same shape appears. The regrets are almost never about ambition or talent; they’re about running the business on instinct when a weekly number would have told you sooner. Shorter menu, bigger reserve, written systems, numbers read weekly, compliance handled as routine. None of it is exciting, and all of it is what the operators who are still open say they wish somebody had made them do in month one.
