Finance & Accounting

Build off the National Restaurant Association's Uniform System of Accounts for Restaurants (USAR, 8th edition) — it gives you the standard four COGS buckets (Food, Liquor, Beer, Wine) and the labor splits (FOH hourly, BOH hourly, salaried, payroll taxes, benefits) that every NYC operator and CPA expects to see. In QuickBooks Online, that means roughly 60-80 active accounts: 4 sales accounts, 4 COGS accounts, ~25 operating-expense accounts (rent, utilities, R&M, smallwares, linen, credit card fees, etc.), and the standard balance-sheet items. Do NOT lump beer and liquor into one account — they run very different margins (beer 22-25% pour cost vs liquor 18-20%) and you'll lose your ability to spot pour problems. Add a separate "Comps & Voids" contra-revenue account on day one, not as a discount expense, or your sales tax base will be wrong. Your bookkeeper should be able to hand you a P&L by the 15th of the following month using this structure.

For one location under $3M in revenue, QuickBooks Online Plus ($99/mo as of 2026) plus a POS integration (MarginEdge $329/mo, Shogo $40/mo, or a direct Toast/Square sync) covers 95% of operators and any NYC CPA can work in it. Restaurant365 starts around $469/mo per location and is overkill until you hit 2-3 units or want true daily sales journals, AP automation, and inventory in one platform — most operators move there at unit 3 or when monthly revenue clears $400K. Xero is fine but the NYC restaurant CPA bench that knows it well is thin compared to QBO, so you'll pay more in accounting fees to translate. Whatever you pick, make sure your POS pushes a daily sales summary (DSS) journal entry — not transaction-level imports, which will blow up your file. If you're running a hotel F&B outlet, your PMS (Opera, Mews) usually owns the GL and your restaurant accounting tool only feeds it a sub-ledger.

Run accrual for management reporting from day one — cash basis hides the timing of inventory, prepaid rent, and unpaid vendor bills, and you cannot make real margin decisions on it. The IRS lets restaurants under the $30M three-year average gross-receipts threshold (2025 Sec. 448(c) limit, indexed) elect either basis for tax filing, and most small NYC operators file taxes on cash to defer income but keep books on accrual internally. Your CPA can run the cash-to-accrual reconciliation at year-end — that's a normal $500-1,500 add-on. If you're doing tip pooling, vendor credits, or carrying a meaningful inventory (anything over $15-20K), cash basis will distort your prime cost by 3-5 points month to month and you'll chase ghosts. Make this decision before your first month-end close, because switching later is a Form 3115 filing.

In QuickBooks Online, use Classes for revenue centers within a single legal entity (bar / dining room / private events / retail) and Locations only if you have separate physical sites under the same EIN. A typical NYC restaurant-with-bar setup runs 3-4 classes — Dining, Bar, Private Events, and Catering / Off-Premise — which lets you produce a P&L by class showing where your gross profit actually comes from (often the bar carries 65-70% of contribution margin even if it's only 35% of sales). If you have a hotel F&B program with a rooftop, lobby bar, and restaurant, each is its own class, and shared overhead (GM salary, R&M, utilities) sits in an unallocated class that you allocate at month-end by revenue percentage. Don't over-engineer — more than 6-8 classes and your bookkeeper starts misposting and the data becomes worse than no segmentation at all. Tag every transaction at entry, not at month-end.

Book your opening balance sheet as of the day you take over the space: cash in the bank, opening food and beverage inventory at cost (count it the night before service starts), all prepaid items (first month's rent, security deposit, liquor license fees, insurance), CapEx assets at purchase price grouped by depreciation class, and any startup loans or owner contributions. NYS liquor license fees ($4,352 for an On-Premises Liquor in NYC for the two-year period) sit in a prepaid asset and amortize monthly. The security deposit (typically 3-6 months rent for an NYC restaurant lease) is an asset, not an expense — it stays on the balance sheet until you exit. If you took over an existing restaurant under an asset purchase, the purchase price allocation (Form 8594) drives goodwill, leasehold improvements, FF&E, and inventory values — get this right with your CPA before your first close, because it sets your depreciation schedule for the next 15 years.

Tips collected on credit cards and distributed to staff are pass-through — they hit a tip-payable liability account when collected and zero out when paid via payroll, never touching revenue or expense. The only piece that touches your P&L is the credit card processing fee on the tip portion, which you pay (around 2.6-3% of the tip amount on every Visa/MC swipe), and tip-share to the house under a valid pool. Service charges (auto-grats on parties of 6+, banquet service charges on contracts) are a different animal — those ARE revenue, ARE subject to NYS sales tax if they aren't separately stated and entirely distributed to employees per NYS DTF TSB-M-09(13)S, and ARE subject to payroll tax when paid out as wages. Most NYC operators get tripped up here — if you call a 20% banquet charge a "service charge" and don't meet the strict separately-stated-and-fully-distributed test, NYS will assess sales tax on the full amount on audit. Talk to your CPA before printing any contract that uses the word "gratuity" or "service charge."

Daily sales journal posted from POS the next morning, weekly bank and credit card reconciliation (every Monday for the prior week), bi-weekly AP run (Tuesdays and Fridays is the NYC norm), and a hard month-end close by the 15th of the following month. If you wait until month-end to reconcile, you will miss missing deposits and double-charged vendor invoices that are easy to catch within a 7-day window and nearly impossible to unwind 60 days later. Weekly inventory counts on liquor, monthly on food, quarterly on smallwares — anything less frequent and you cannot run a real prime cost. Your bookkeeper should be in the file 2-3 hours per week per location for a typical $2-4M operation, billing $35-60/hr for a remote bookkeeper or $65-100/hr for a NYC-based one with restaurant chops.

Daily summary only — never import line-by-line transactions into QuickBooks. A single Saturday night at a busy NYC restaurant can generate 400-700 tickets, and pulling all of those into QBO turns your file into mush within 90 days and breaks the bank rec. The right pattern is one Daily Sales Summary (DSS) journal entry per day with debits to undeposited funds (cash, Amex, Visa/MC) and credits to revenue, sales tax payable, tip liability, comps, and gift card liability. Tools like Shogo ($40-95/mo), MarginEdge, or Restaurant365 build this entry automatically from Toast, Square, Resy, or Aloha. Then your bank deposits clear out the undeposited funds account when they hit — which is also how you catch a missed deposit in 24 hours instead of 30 days.

Target the 15th business day of the following month for first-pass P&L delivery, the 20th for clean final numbers reviewed by the operator. Standard close checklist: bank and credit card recs done, all vendor bills posted and accrued, payroll fully booked including employer taxes and tip payouts, ending inventory counted and journaled, prepaid items amortized (insurance, license, rent), depreciation booked, intercompany or owner draw entries cleared, sales tax accrual reconciled to NYS filing, and tip liability balance verified to zero (or to whatever is genuinely held over). The single most-skipped step is inventory accrual — without it your COGS swings 5-10 points month to month and you can't trust prime cost. A clean close means you can hand the same package to your CPA at year-end with no reclasses; if you're doing more than 5-8 reclassifications in March for the prior year, your monthly close isn't tight.

Count physical inventory as close to the last day of the month as you can — most NYC operators count the morning of the 1st before lunch service, when the bar is empty and walk-ins are stocked. Value at most-recent invoice cost (FIFO), not retail or weighted average, unless your CPA has specifically elected another method. The journal entry is straightforward: ending inventory on the balance sheet equals your physical count, and the difference between beginning and ending inventory adjusts COGS for the month. So if you started with $18K of food, bought $42K, and ended with $20K, your food COGS for the month is $40K (18 + 42 − 20). If you skip the count and just expense purchases, your COGS will swing 4-8 points month over month based purely on stocking timing, and you'll chase pricing decisions that don't exist. Liquor counts should be weekly with cost-per-ounce by bottle for any bar doing meaningful volume.

Most NYC restaurants run bi-weekly payroll, which means 6 out of 12 months will have a pay period that straddles month-end, and the unpaid wages need to accrue or your labor line is wrong. The clean way: at month-end, calculate the days worked in the current month that haven't been paid yet (from POS clock-out reports), gross those wages up by your blended payroll tax rate (~10-11% for FICA + FUTA + NY SUI + NY MCTMT), and book a single journal: debit wages and employer taxes by category (FOH, BOH, salaried), credit accrued payroll. Reverse it on day 1 of the next month. If you don't accrue, your labor will look 7-9% low in some months and 7-9% high in others, and you cannot read prime cost trends. For 401(k) match, health insurance, and PTO accrual, build similar standing entries — they're worth the 30 minutes a month.

Cut off AP entry at the 10th of the following month — anything dated for the prior period (received-by date on the invoice, not invoice date) gets booked into the closing month. Your produce and seafood vendors (Baldor, Pierless, M&I) typically deliver 5-6x per week and bill in the same week, so by the 8th-10th you've seen 95% of last month's purchases. For the remaining stragglers, your bookkeeper should pull a list of recurring monthly vendors (linen, exterminator, music licensing, garbage) and accrue any that haven't billed using the prior-month average. The journal is debit COGS or operating expense, credit accrued expenses, and reverse on day 1 of the new month so when the actual invoice arrives it doesn't double-hit. Forgetting to accrue a $4K Baldor week makes your food COGS look 1.5 points better than reality, which is exactly the kind of fake good news that costs operators real money.

Every daily sales journal credits sales tax payable for the tax collected (8.875% in NYC = 4% NYS + 4.5% NYC + 0.375% MCTD). At month-end, the balance in sales tax payable should equal exactly what you owe NYS DTF for the period: total taxable sales × 8.875%, less the credit for the small NYS vendor collection allowance (5% of tax due, capped at $200/quarter for vendors who file quarterly). If your sales tax payable balance doesn't match what you'll file, something is wrong — usually a comp or void posted to revenue instead of contra-revenue, or a service charge that shouldn't be taxable getting taxed. Reconcile this every month, not at filing time, because NYS DTF interest on underpayments runs at 8% annually (2026 rate) and they'll find the error on audit going back three years. Bartender well drinks rung as no-charge for a regular still owe sales tax on the menu price unless properly comped under the documented house policy.

Gift cards sold are a liability, not revenue, until redeemed — debit cash, credit gift card liability. When redeemed, debit gift card liability, credit revenue and sales tax payable. NYS sales tax is collected at redemption, never at sale, so don't tax the card purchase. Track breakage (the 8-12% of cards that are never fully redeemed) per ASC 606 — most NYC restaurants recognize breakage proportionally as cards are used, or in full after 24 months of inactivity if your historical pattern supports it. NY General Business Law §396-i prohibits expiration dates and dormancy fees on gift cards under $5,000, so you can't write off a balance just because it's old — but you can recognize estimated breakage to revenue. If your gift card liability balance just keeps growing forever and never bleeds down, your bookkeeper isn't redeeming them properly against the POS gift card module.

Top-down USAR format: gross sales by category, less comps and discounts (contra-revenue), equals net sales. Then COGS by category (food, liquor, beer, wine) to get gross profit. Then labor — split into FOH hourly, BOH hourly, management salaries, payroll taxes, and benefits — to get prime cost (COGS + total labor). Then controllables (R&M, smallwares, linen, paper, cleaning, music, marketing). Then occupancy (rent, CAM, real estate tax, insurance, utilities). Finally non-controllables and depreciation to get operating profit (EBITDA). Every line should show dollars and percent of net sales side by side — that's the only way to read trends. NYC operators target prime cost at 60-65% of sales (anything above 68% is bleeding), occupancy at 8-12%, and EBITDA of 12-18% on a stable second-year-or-later restaurant. Bars run lower COGS (25-30%) and can absorb labor up to 32-35% and still hit 20%+ EBITDA.

Full-service NYC restaurants should target prime cost (COGS + total labor including taxes and benefits) at 60-65% of net sales. Quick-service or fast-casual can run lower at 55-60%. Cocktail-driven bars with light food can hit 50-55%. Above 68% you are losing money after rent and overhead in NYC's cost structure — there is no path to 15% EBITDA at 70% prime cost when occupancy is 10% and G&A is 8%. NYC pressure points: minimum wage at $16.50/hr citywide as of Jan 1 2026 with the tip credit eliminated for fast-food and pressure on full-service to follow, rents at $150-300/sq ft for prime ground-floor, and food inflation that ran 4-7% across most categories in 2025-26. The lever you control fastest is labor — schedule to actual sales, not to your wishful forecast.

Total occupancy (base rent + CAM + real estate tax pass-through + insurance + utilities) should land between 8-12% of net sales for a NYC full-service restaurant to leave room for a real margin. Below 8% you are either underutilizing the space or you have a legacy lease that the landlord will reset at renewal. Above 14% you are paying for the space but losing the business — the math doesn't work. Manhattan ground-floor rents in 2026 sit at $150-300/sq ft for prime hospitality corridors (West Village, LES, Flatiron) and $80-150/sq ft in Brooklyn (Williamsburg, Cobble Hill, Park Slope). Sanity check: a 2,500 sq ft space at $200/sq ft is $500K/yr base rent, which means you need $4.2-6.3M in net sales to keep occupancy in the healthy band — if your concept can't realistically hit that, walk away from the deal. Negotiate hard on percentage rent kickers above a breakpoint instead of fixed escalators in the first 5 years.

Compare three things every month: actual vs prior month, actual vs same month prior year, and actual vs your trailing-3-month average. A single bad month in a single line is usually weather, a holiday shift, or a one-time event — don't react. A two-month adverse trend in the same line is a real signal that needs investigation. Prime cost moving 1.5+ points off baseline is meaningful; 0.5 points is noise. The lines that lie the most are food COGS (because of inventory timing) and labor (because of pay-period cutoffs) — always check that both are accrued before reading them. The lines that lie the least are credit card processing fees (a clean percentage of card sales) and rent (fixed) — if those move significantly, you have a math error not an operating change. Build a 13-period rolling P&L view so you can see seasonality without the calendar-month distortion of February vs August.

Full-service NYC restaurants typically run 30-35% total labor (including taxes and benefits), split roughly 12-14% FOH hourly, 14-17% BOH hourly, 4-6% management salary. The shift in NYC since 2024 is BOH wages climbing faster than FOH because tipped FOH wages were already at the $11.00/hr tipped minimum (2026 NYC rate) plus tips, while BOH cooks went from $16/hr to $20-24/hr in two years to compete with delivery and construction work. If your BOH is above 18-20% of sales, either your menu is too labor-intensive for the price point, your prep yields are bad, or you are over-staffing a slow station. If your FOH is above 16% in a tipped house, you are not turning tables fast enough or you are over-staffing the floor. Track labor as percent of sales by daypart, not just by week — lunch is where most NYC operators bleed labor.

Healthy NYC full-service restaurants generate $40,000-80,000 in annual revenue per seat at the low end (neighborhood spots) and $100,000-150,000 at the high end (high-volume Flatiron/Soho/West Village concepts with strong bar programs). On a per-square-foot basis, $800-1,200/sq ft is solid and $1,500+ is exceptional — anything under $600/sq ft and the box is too big for the concept. Cocktail bars often hit higher revenue density ($1,500-2,500/sq ft) because liquor pours more dollars per square foot than food. Hotel F&B outlets are different — banquet/private events should generate $3-5K per available banquet seat per year; rooftop bars with seasonal closure can still hit $2-3K per occupied seat-month at peak. These are sanity-check ratios for a budget, not P&L targets — what you actually manage to is prime cost and four-wall EBITDA.

Target food cost at 28-32% of food sales for most NYC full-service concepts — anything consistently above 34% is leaking either through portioning, waste, or theft. Steakhouses and seafood-heavy menus run higher (32-36%) because protein is the dominant cost; pasta- and pizza-driven menus run lower (22-28%). The real number you manage to is theoretical vs actual food cost: build a recipe-costed menu in MarginEdge or Restaurant365, calculate what your food cost SHOULD be at the mix you actually sold, and compare to the inventory-based actual. A gap of more than 1.5-2.0 points between theoretical and actual is real money walking out — usually over-portioning, kitchen waste, or comps not getting captured. Run this monthly minimum, weekly if you're tight on cash. Inflation on key categories ran 4-7% in 2025-26, so if your menu prices haven't moved in 12 months, your food cost is silently climbing 1-2 points per year.

NYC anchor pour costs: liquor 18-22%, draft beer 22-25%, bottled/canned beer 25-28%, wine by the glass 28-32%, wine by the bottle 33-40%. The bar program drives the four-wall margin in most NYC restaurants — a 5-point pour cost slip on a $1.5M beverage program is $75K of pure profit gone. Cocktail-forward bars with house spec drinks should target 18-20% beverage cost; high-end wine programs running deep cellars may accept 38-42% on bottles to drive volume. Test pour every Saturday morning — count every open bottle to the tenth, reconcile to POS sales, and you'll know your variance within a week instead of finding a 4-point hole at month-end. The single biggest theft vector is bartender free pours and friend-and-family discounts, which never show up in COGS because there's no purchase variance — only a pour test catches it.

Recipe-cost every menu item in MarginEdge or Restaurant365 — every ingredient at current invoice price, accurate yields (a 12 oz raw ribeye yields ~10 oz cooked, not 12), and every garnish counted. Multiply each item's recipe cost by units sold from the POS for the period, sum to get theoretical COGS, divide by net food sales to get theoretical food cost percentage. Compare to your actual (inventory-based) food cost: actual − theoretical = variance. A 1.0-2.0 point variance is normal kitchen reality (trim, drops, comps); above 2.5 points means you have a real problem — over-portioning, theft, undocumented comps, or stale recipe costs. Walk the line: the most common cause is the cook adding 1 oz of cheese when the recipe calls for 0.5 oz, repeated 200 times a week. Update recipe costs every 60-90 days because vendor prices move; stale recipe costs make theoretical look better than it is and hide real margin loss.

Price each item to hit your overall food cost target while accounting for popularity (the menu engineering grid: stars are high-margin/high-popularity, dogs are low-margin/low-popularity, plowhorses are low-margin/high-popularity, puzzles are high-margin/low-popularity). The mistake is pricing every item at a flat multiplier — instead, price plowhorses up (small bumps because guests will absorb them on items they already love) and puzzles down or rework them. NYC operators typically run a 3.0-3.5x markup on food (food cost 28-33%) and 4.5-5.5x on cocktails (pour cost 18-22%). Test with a small price increase on top sellers first ($1-2 on a $24 entree is invisible to the guest; $3 starts to register). Re-engineer the menu every 6 months minimum, immediately after any major vendor price change. If your menu hasn't been re-priced in 12 months in 2026, you've absorbed 4-7% of food inflation as pure margin loss.

Log waste daily in a paper or app-based waste sheet at the line — every drop, burn, miscook, and spoiled product gets weighed or counted with a reason code. Family meal (the staff meal) should be a separately tracked line in COGS, not buried — most NYC restaurants run family meal at 0.5-1.0% of food sales and use scrap, vegetable trim, or whole-protein pieces past prime presentation. Spoilage on a well-run kitchen runs 1-2% of food purchases; above 3% you have an ordering or rotation problem. Comps and voids on the POS need to capture the food cost of what was given away — most POS systems will report comp dollars at menu price, but the cost impact is the food cost percentage of that. If you don't track these, you'll see your actual food cost climb mysteriously and chase the wrong cause. Industry data shows NYC restaurants generate ~85,000 lbs of organic waste per outlet per year — track it, and you'll cut it 20-30%.

For any bar doing more than $400K in liquor sales annually, yes — perpetual inventory at the bottle level pays for itself in 60-90 days. Tools like BevSpot ($299-499/mo), BinWise, or Backbar let your bar manager scan invoices in, scan pours out via POS integration, and report variance per bottle weekly. The alternative — counting only at month-end — gives you a single big variance number but no idea which bottle, which bartender, or which shift is the source. Wine programs deep enough to need allocation tracking (anything over 200 SKUs) should be on perpetual too. The cost is $300-600/mo plus 4-6 hours/week of manager time; the return is typically 2-4 points of pour cost recapture, which on a $1M bar program is $20-40K/year. Skip perpetual only if your bar is under $300K annually or your concept is super-simple (5 cocktails, 6 beers, 4 wines).

Yes, and you should. Weekly prime cost is the single highest-leverage management report a NYC operator can run — you spot a labor or COGS problem 7-10 days into a 30-day window instead of finding it 45 days too late. Weekly prime cost requires: weekly inventory counts (food and liquor), weekly purchase totals (which your AP system already has), weekly POS sales, and weekly payroll (ADP/Gusto/Toast Payroll all report by week even on bi-weekly pay cycles). The math is identical to monthly — beginning inventory + purchases − ending inventory = COGS, plus labor, divided by sales. Tools like MarginEdge, R365, or even a clean Google Sheet pull this together in 30-45 minutes Tuesday morning for the prior week ending Sunday. Operators who run weekly prime cost beat operators who run monthly by an average of 2-4 points of EBITDA over a year — it's not the data, it's the response time.

NYC's three big broadline distributors (Sysco, US Foods, Performance Food Group) and regional specialists (Baldor, M&I, Pierless) all do vendor-rebate and growth-incentive programs once you're spending $8-15K/week with them. Standard ask: 1-3% rebate on annual purchases above a baseline, 60-day terms instead of 30 (or 30 instead of COD), price-locked contracts on top 25 SKUs (your A-items by spend), and quarterly business reviews with cost-creep transparency. The trap is the exclusivity language — never sign a single-vendor exclusive on broadline because it kills your competitive leverage on the next renewal. For beverage, NYC distributor consolidation (SGWS-Edrington in 2026, SGWS taking AB-InBev distribution Nov 2025) means more pricing pressure on operators, so push hard for case discounts, depletion allowances, and supplier-paid POS materials at every renewal. Track every rebate as a credit to COGS, not as other income, or you'll mis-state your actual food cost.

Delivery commissions (DoorDash, Uber Eats, Grubhub typically charge 15-30% of order value depending on plan) belong in a separate "Third-Party Delivery Commission" line that sits between COGS and labor as a contra-revenue or as a separate occupancy-of-channel cost — not in marketing, not in standard COGS. Why this matters: a $30 burger with a 30% delivery commission is a $9 channel cost on top of food cost, so the effective food cost on that order is closer to 60% than 30%, and you cannot read your real food cost performance if you mix the two. NYC capped third-party delivery commissions at 23% (15% delivery + 5% marketing + 3% credit card) per Local Law 88 of 2021, but enforcement is uneven and many operators have signed prior contracts at 30%. Run a separate P&L view for in-house orders vs delivery channels so you can see whether each channel is profitable on its own. Most operators discover delivery is their lowest-margin channel and use it for incremental volume only, not as a primary revenue source.

Comps and discounts should reduce gross sales as contra-revenue, not increase expenses — debit a "Comps & Discounts" contra-revenue account, credit the original sales account at gross. This matters for two reasons: (1) NYS sales tax is calculated on the comped/discounted price, not the menu price, so booking comps as expenses inflates your taxable sales base and you over-pay sales tax; (2) you cannot read your real top-line revenue if comps are buried in operating expenses. The food cost of comped items still lands in COGS naturally through inventory consumption — that's correct, that's a real cost. Track comps by reason code (manager comp, server error, marketing promo, VIP, employee meal) and by employee — if one server is comping 4% of their checks and the average is 1.5%, you have a control problem or a real performance problem to address. Above 2-3% total comps as a percent of gross sales, your guest experience is weak or your manager is over-using the comp button to smooth over service.

Combined NYC restaurant sales tax is 8.875% — 4% NYS, 4.5% NYC, and 0.375% Metropolitan Commuter Transportation District (MCTD). All restaurant food and beverage sold for on-premises consumption is taxable, regardless of whether it would be tax-exempt at a grocery (a bagel sliced on premises = taxable; the same bagel whole and to-go = exempt). Cover charges and minimum-purchase requirements are taxable. Take-out food is generally taxable when prepared/heated, but cold sandwiches/salads to-go are exempt unless served with utensils. Alcohol is always taxable on-premises. Service charges that are NOT separately stated and fully distributed to staff are taxable; tips that are voluntary, separately stated, and fully paid out are NOT. The tax shows on the guest check as a separate line item — never bundled into menu pricing for on-premise sales (NYS Tax Law §1132(a) requires separate statement). Check your POS tax rules at every menu re-engineer.

NYS DTF assigns filing frequency based on annual taxable sales: under $300K/year files annually (Form ST-101), $300K to $3M files quarterly (Form ST-100), and over $3M (or any vendor with $500K+ in any quarter) files monthly via PrompTax (PR-100 series). Most NYC restaurants land on quarterly within 6 months of opening because $300K in annual taxable sales is roughly $25K/month — a low bar. Quarterly returns are due the 20th of the month following quarter-end (Mar 20, Jun 20, Sep 20, Dec 20 for the standard fiscal year). Monthly PrompTax filers must remit electronically and there's a penalty for late ACH initiation. Once you cross thresholds, NYS will move you up automatically and notify by mail — read that notice, because missing a transition triggers automatic late-filing penalties even if you paid the right amount.

NYS DTF assesses a penalty of 10% of tax due for the first month late, plus 1% per additional month, capped at 30%. Interest accrues separately at the prevailing rate (8% annually for 2026, set quarterly). On a $25K monthly sales tax bill that's $2,500 in penalty within 30 days. More serious: NYS treats sales tax as a trust fund tax — money you collected from guests on behalf of the state — and personally pierces the corporate veil under Tax Law §1133(a) to assess the responsible person (owner, officer, anyone with check-signing authority). That means a missed $50K sales tax bill becomes your personal debt that survives bankruptcy of the LLC. NYS DTF will pad-lock a restaurant for sales tax delinquency under their warrant authority — visible padlock, credit destroyed, lease in default — within 60-90 days of repeated nonpayment. Pay sales tax before you pay yourself, every period, no exceptions.

Yes — NYS gives a vendor collection credit of 5% of sales tax due, capped at $200 per quarter ($800/year), for filing and paying on time. It's automatic on the return; you just check the box. It's small money for a $2-3M restaurant ($800/year on a $200K sales tax bill is 0.4%) but free, so take it. The credit does NOT apply if you file late, even by one day, and does NOT apply to PrompTax (monthly) filers. The vendor collection credit applies separately to NYS, NYC, and MCTD portions, but in practice the cap eats most of it for any meaningful operator. Worth knowing about; not worth a moment of management attention beyond making sure your filer takes it.

NYS DTF audits restaurants on a regular cycle (3-5 years between audits is normal for established operators) and on triggers: anonymous tips, mismatched 1099-K from credit card processors vs reported sales, sudden drops in reported sales, employee complaints, or industry sweeps (NYS ran one across NYC nightlife in 2022-23). They will pull POS data, bank deposits, vendor purchases, and your sales tax returns for 3 years (extends to 6 if they find substantial under-reporting, unlimited if fraud). The two killer findings: (1) cash sales not deposited and not reported, which they reconstruct from purchase markup analysis (if you bought $500K of liquor at a 20% pour cost, they'll assess on $2.5M in sales whether you reported it or not); (2) misclassified service charges. Keep daily POS Z-tapes for 6 years minimum, all bank deposits matched to POS by date, and a separately stated tip vs service charge methodology you can document. If you get an audit notice, hire a sales tax specialist CPA (NYC has 20-30 firms that specialize) before responding — DIY usually goes badly.

Catering sales delivered within NYC are taxed at the destination rate (8.875% if delivered in NYC; different if you cater into Westchester, Long Island, or NJ — destination-based for NYS, origin-based for NJ). The catering invoice typically itemizes: food and beverage (taxable), labor/staffing (taxable when bundled with food service per NYS DTF TSB-M-09(13)S), rentals (taxable), service charge (taxable unless meeting the strict tip test), and gratuity (not taxable if voluntary, separately stated, fully distributed). Catering at a customer's premises on Long Island gets Suffolk's rate (8.625%); in Westchester it's 8.375%. Out-of-state catering (NJ, CT) is a different state's tax filing — you may need to register as an out-of-state vendor in NJ if you cross their economic nexus threshold ($100K or 200 transactions annually). Most NYC caterers under $1M off-premise revenue stay in NYS and file NYS only; once you regularly cater into NJ, talk to your CPA about NJ registration to avoid the use-tax assessment.

Form ST-120, the Resale Certificate, lets you buy items you'll resell to customers without paying sales tax to the vendor — you collect tax from your guest at the point of sale instead. For a restaurant, that means liquor, beer, wine, food, paper goods that go to the guest (napkins, takeout containers), and packaging. You give your vendor a completed ST-120 once and they keep it on file. You CANNOT use it for items you consume internally — kitchen smallwares, cleaning supplies, R&M parts, office supplies — those are use-tax taxable to you at 8.875% even if you bought them from an out-of-state vendor who didn't charge tax (you self-assess on the sales tax return). The most common audit finding is an operator using ST-120 to buy bar smallwares (jiggers, strainers, glassware) — those are consumed internally, sales tax is owed. Keep your ST-120s on file for 6 years and update annually.

For most NYC restaurants the choice is between Toast Payroll, ADP RUN, Gusto, and Paychex Flex — all four handle NYC's tipped wage rules, NY State withholding, NYC local tax, MCTMT, NY SUI, and tip credit calculations. Toast Payroll ($69/mo + ~$9/employee) integrates natively with Toast POS so tips and hours flow without re-keying — that alone is worth it for operators on Toast. Gusto ($40/mo + $6/employee) is the cleanest interface and best for operators who want self-service for staff but doesn't have native POS integration outside of a few platforms. ADP RUN ($79/mo + $4-12/employee depending on tier) is the workhorse most NYC CPAs prefer because the reports and tax filings are bulletproof. Paychex Flex is similar to ADP. Avoid running payroll in-house manually — NYC's combination of tip credit, spread-of-hours pay (NY Labor Law §142-2.4), uniform allowance, and ~14 different tax filings per quarter makes DIY a guaranteed compliance disaster.

As of January 1 2026, NYC tipped food service workers (servers, bartenders, bussers) have a cash wage minimum of $11.00/hr with a tip credit of $5.50/hr to reach the $16.50/hr full minimum wage. Service employees outside food service have a $13.75 cash wage with a $2.75 tip credit. The tip credit only applies if the employee actually earns enough tips to bring total compensation to at least $16.50/hr in every pay period — if they don't, you owe the difference (called "makeup pay" or "shortfall pay"). Required: the 80/20 rule effectively no longer applies federally after 2024 court rulings, but NY still requires that tipped work be the primary duty. Provide written notice of tip credit at hire (NY Labor Law §195(1)) on a NYS DOL-prescribed form, in the employee's primary language, signed and dated. Failing to give that notice voids the entire tip credit retroactively — a $5.50/hr × 40 hrs × 52 wks × every server lawsuit, which is exactly how a $20K mistake becomes a $200K class action.

Spread of hours is a NY Labor Law (12 NYCRR §146-1.6) requirement that any non-exempt employee whose workday spans more than 10 hours from start to finish is owed an extra 1 hour at the full minimum wage ($16.50 in NYC for 2026), in addition to all hours actually worked. The 10-hour spread includes all unpaid breaks and split-shift gaps — so a server working 11am-3pm lunch and then 5pm-10pm dinner has a spread of 11 hours and is owed 1 extra hour at $16.50 even though they only physically worked 9 hours. This is independent of overtime; it stacks on top. NYC restaurants get burned on this constantly because most POS time clocks track hours worked but not spread, and most payroll systems require manual spread-of-hours entry. Audit your time records weekly — Toast, 7shifts, and HotSchedules all have spread-of-hours flags now. A NYS DOL audit finding on spread-of-hours easily reaches $30-80K in back wages plus liquidated damages on a 30-employee restaurant with 3 years of bad practice.

Employer payroll taxes for a NYC restaurant typically run 10-12% of gross wages: FICA (Social Security 6.2% on first $176,100 in 2026) + Medicare 1.45% (with no cap, plus 0.9% additional for the employee on wages over $200K) = 7.65% federal; FUTA 0.6% on first $7,000; NY SUI rates vary by experience (new employers ~4.025% in 2026, established can be as low as 0.6% or as high as 9.825%) on first $13,000; NY Re-employment Service Fund 0.075%; NYC MCTMT 0.34% on quarterly payroll over $312,500. Plus workers' comp (~$2-4 per $100 of restaurant payroll, NCCI class 9079), NY State Disability ($0.60/wk per employee max), Paid Family Leave ($0.388% of wages capped at $354.53/year per employee for 2026), and short-term disability. Health insurance, 401(k) match, and other benefits sit on top. Budget all-in payroll loaded cost at 13-16% above gross wages for a typical NYC operation.

Almost never. NYS DOL applies a strict ABC-style test (NY Labor Law and the Vega v Postmates line of cases) plus the IRS 20-factor common law test, and front-of-house, back-of-house, dishwashers, hosts, and even most bartenders fail virtually every prong — they work your hours, with your equipment, under your direction, on your premises. Misclassifying a W-2 as a 1099 is one of the highest-frequency NYS DOL audit findings on hospitality and triggers back wages, penalties, unpaid SUI, unpaid workers' comp (which is a separate criminal exposure under WCL §52), and personal liability for officers. The narrow categories where 1099 may legitimately apply: a hired DJ for a one-off event, a consulting chef who's developing a new menu and not running service, a freelance bookkeeper or designer, a private trainer or piano teacher coming in once a week with their own students. If they wear your uniform, take direction from your manager, or work a recurring schedule, they are a W-2 employee — period. The cost of getting this wrong is multiples of the cost of doing it right.

Yes if you're a "large food and beverage establishment" — defined as a tipping business with 10 or more employees on a typical business day during the prior calendar year (a common NYC threshold to hit fast). Form 8027 is filed annually by Feb 28 (paper) or Mar 31 (electronic) and reports gross receipts, charged tips, charged sales, and reported tips. If reported tips fall below 8% of gross receipts, you must allocate the shortfall among tipped employees and report the allocation on each W-2 in box 8. Most operators avoid the allocation by participating in TRDA (Tip Rate Determination Agreement) or attributed-tip programs, and by enforcing complete tip reporting through POS-driven cash tip sign-off at end of shift. The IRS audit angle: if your reported tips on 8027 are 6% but the IRS thinks NYC reality is 18-22%, they'll assess employer FICA on the difference. Use Toast/ADP/Gusto's automated tip tracking and require staff to electronically declare cash tips at clock-out — that's your audit defense.

The FICA tip credit (IRC §45B) lets you claim a federal income tax credit equal to the employer's share of FICA (7.65%) on the portion of employee tips that exceed the federal minimum wage of $5.15/hr (a frozen 1996 number used only for this calculation, NOT the current minimum wage). For a busy NYC restaurant with $400K in reported tips, that's roughly $25-30K in annual federal credit — real money. The credit is claimed on Form 8846 with your business return. You don't have to choose between the credit and deducting the FICA expense — you can take the credit and only the deducted FICA portion is added back to income. Most CPAs catch this; if yours doesn't, switch. You can amend prior years going back to the standard amendment window (3 years from filing) to recapture missed credits — worth doing if you've been on the wrong CPA.

Yes — NY Workers' Compensation Law §10 requires every employer with one or more employees to carry workers' comp, no exceptions for small operators or part-timers. Restaurant payroll falls under NCCI class code 9079 (Restaurant - All Other Employees) and 9082 (Restaurant - All Other Service Employees), with NY rates running roughly $2.00-4.50 per $100 of payroll depending on experience modifier and carrier in 2026. On a $1M payroll that's $20-45K/year. Purchase through a NYS-approved carrier (NY State Insurance Fund is the public option and often cheapest for new operators with no experience rating; private carriers like The Hartford, Travelers, Cerity offer better service once you have 2+ years of clean claims). Failure to carry is a criminal offense under WCL §52 with fines up to $2,000 per 10-day period of noncompliance, plus personal liability for any injury during the gap. Your liquor license, your lease, and your business loan all require active proof of WC coverage — let it lapse and your whole operation freezes.

For most independent NYC restaurants, the practical choices are Chase Business Complete, Bank of America Business Advantage, Citizens Bank, Signature Bank's successor (Flagstar after the 2023 takeover), or one of the NYC community banks like Amalgamated, M&T, or Apple Bank. Chase has the densest branch network in NYC for cash deposits (a real consideration for a cash-heavy bar) and the cleanest QuickBooks integration. Mercury and Relay are online-only options that work well for newer operators with low cash volume but won't take in-person cash deposits — disqualifying for most bars. Avoid the temptation to bank with a single relationship for every account; most NYC operators run an operating account at one big bank for daily flow plus a separate sweep/reserve account at a community bank that pays meaningful interest (4-5% as of early 2026). Your liquor license requires a NY State business bank account in the licensed entity name — out-of-state online banks don't qualify for SLA purposes.

Minimum four accounts for a single-location NYC restaurant: (1) operating account where credit card deposits and cash deposits land and AP runs out; (2) payroll account funded weekly by transfer from operating; (3) sales tax / payroll tax reserve where you sweep tax-due amounts daily or weekly so the cash is segregated and untouchable; (4) capital reserve / emergency account holding 60-90 days of operating expenses. The tax reserve is the one most operators skip and the one that ends them — when you collect $50K of sales tax monthly, that money is not yours, and the only way to make sure you don't accidentally spend it is to physically move it out of the operating account every Monday. Adding a fifth account for tip pool / tip-out distribution can simplify reconciliation if you do tip pooling. Multi-unit operators add an entity-level holding account that sweeps from each unit and fronts capital back. Online sweep accounts at community banks pay 4-5% in 2026; that's $4-5K/year on a $100K reserve — pure free money.

Daily armored car service (Loomis, Garda, Brinks) for any restaurant with more than $2-3K/day in cash — pickup costs $90-180/week depending on frequency and is cheap insurance against the manager-walking-to-Chase-with-a-deposit-bag scenario. The control flow: cashier drops counted bank into manager safe at end of shift with witness sign-off; manager closes out POS cash tender to drop sheet; armored car logs sealed bag with serial number; bank credits within 1-2 business days. Reconcile the bank deposit to the POS Z-tape daily, not weekly — if a $1,200 deposit is short by $400, you want to know within 24 hours. NYC restaurants in cash-heavy neighborhoods (Lower East Side, parts of Brooklyn, immigrant-community areas) should also negotiate cash-handling fees with their bank — Chase charges $0.0030 per dollar deposited above the free cash limit ($7,500/mo on Business Complete), so a bar depositing $30K cash/month pays $67 in fees. Switch to a cash-friendly bank like Amalgamated or community options if your fees climb above $200/month.

All-in effective rate for a NYC restaurant should be 2.4-2.9% on a blended basis (mix of Visa/MC ~2.3%, Amex ~3.0-3.5%, debit ~1.5%, online cards higher). If you're north of 3.0% you're being over-charged or you have a high mix of premium rewards cards. The big POS processors (Toast, Square, Stripe, Clover) bundle processing into the platform at flat-rate (Toast at 2.49% + $0.15 per transaction is typical for in-person card-present 2026 pricing) — clean and predictable but you pay a premium over interchange-plus pricing. Interchange-plus pricing through a traditional processor (Heartland, Worldpay, TSYS) saves 0.2-0.5% but requires you to negotiate and audit statements monthly because hidden fees pile up. Lever to pull: surcharging up to 4% on credit card transactions is legal in NY since Visa's 2023 settlement (not on debit), with proper signage at entry and POS — passing this on can save a $2M restaurant $40-60K/year, but test guest reaction before deploying. Re-shop processors every 24 months; rates have come down meaningfully and your loyalty earns nothing.

Build a rolling 13-week cash flow forecast updated every Monday — it's the single most important financial tool a NYC operator runs. Format: weeks across the top, in-flows by source (credit card deposits net of fees by 2-day lag, cash deposits, gift card sales, banquet deposits) and out-flows by category (payroll on the right week including tax remittance, AP run dates, rent on the 1st, sales tax on the 20th, license renewals, insurance, debt service) summing to weekly net cash change and a running cash balance. Most operators discover that their cash low point isn't where they thought it was — for a typical full-service restaurant it's the week of the 1st (rent) plus the week of the 20th (sales tax) in low-revenue months (Jan, Feb, late Aug). If your 13-week minimum cash balance dips below 4 weeks of operating expenses, you need to either delay AP, draw on a line of credit, or hold a vendor payment — and you want 3-4 weeks notice, not 3-4 days. Tools: Float, LivePlan, or a clean Google Sheet linked to your P&L update.

Merchant processors hold a reserve (typically 5-10% of monthly volume, or a flat $5-25K) on high-risk or high-chargeback merchants — restaurants are not usually flagged as high-risk, but ticketed events, deposits collected far in advance (catering, weddings booked 12 months out), or any merchant with chargebacks above 1% will see a reserve imposed. The reserve sits in a non-interest-bearing account at the processor for 90-180 days. If your business is healthy you can request reserve release after 6-12 months of clean processing history (chargeback rate under 0.5%, no fraud flags) — write a formal letter to your processor's risk team with your trailing 12 months of statements. Switching processors does NOT release an old reserve; the prior processor holds it for the full reserve period after your last transaction. Plan for this in your cash flow forecast — a $50K reserve held by Stripe for 6 months is $50K of cash you can't touch. Negotiate the reserve down or eliminated as part of a renewal or processor switch — it's almost always negotiable for a multi-year operator.

Use a $2,500 per-item de minimis safe harbor (IRS Reg §1.263(a)-1(f), available without an Applicable Financial Statement) — anything below $2,500 gets expensed, anything above gets capitalized as a fixed asset and depreciated. Standard restaurant capitalizable categories: leasehold improvements (build-out, plumbing, electrical work attached to the structure), kitchen equipment ($2,500+ per piece — a $5,000 ice machine yes, a $400 immersion blender no), FF&E (tables, chairs, dishwashers if $2,500+ each), POS hardware including stations, computer equipment, security systems, and capitalized software (over $2,500 with multi-year useful life). What NOT to capitalize even if expensive: smallwares (plates, glassware, flatware — these are expensed under the IRS smallwares method), uniforms, food and beverage inventory, opening promotion expenses, and any single repair (replacing a compressor in an existing walk-in is repair, not capital, even at $4K). Document the capitalization policy in your accounting policies file once and apply consistently — IRS audits flag inconsistent capitalization first.

Standard MACRS lives that apply to restaurant assets: kitchen equipment, refrigeration, and most FF&E = 5-year property; furniture and fixtures (tables, chairs, banquettes) = 7-year property; computers and POS hardware = 5-year property; leasehold improvements = 15-year property under Qualified Improvement Property (QIP) rules post-CARES Act. Buildings (if you own) = 39-year nonresidential. Section 179 expensing lets you immediately expense up to $1,220,000 of qualified property in 2025 (indexed annually, ~$1.25M for 2026), phasing out above $3.05M of total purchases. Bonus depreciation has phased down — 60% in 2024, 40% in 2025, 20% in 2026, 0% in 2027 unless Congress restores it. The combination of Section 179 + remaining bonus depreciation means a typical $400-800K NYC restaurant build-out can be substantially expensed in year 1 if structured properly. This is a CPA conversation before you write the first check, not after — the depreciation strategy interacts with passive loss rules, basis tracking, and any LLC partner allocations.

Yes for any build-out over $500K, almost certainly yes over $1M. A cost segregation study (typically $5-15K from a NYC engineering firm like KBKG, CSSI, or Cherry Bekaert's restaurant practice) reclassifies portions of leasehold improvements out of 15-year QIP into 5-year (kitchen equipment, dedicated decorative lighting) and 7-year (furniture, casework) buckets, accelerating depreciation. On a $1.5M NYC restaurant build-out, a typical study moves $300-600K from 15-year to 5/7-year property, generating $150-250K of additional first-year depreciation deductions (with Section 179 + bonus). At a 35-40% combined federal+NY+NYC effective rate, that's $50-100K of cash tax savings in year one — well above the study cost. The study has to be done by qualified engineers, documented contemporaneously with the build, and survives IRS audit on its substance. Don't bolt one on years later from your CPA's worksheet — get it done as part of construction closeout while the contractor pay apps are fresh.

IRS Reg §1.263(a)-3 (the tangible property regs) sets the BAR test — Betterment, Adaptation, or Restoration — for whether something is an improvement. Replacing a broken faucet with a similar faucet is a repair (expense). Renovating a bathroom with new fixtures, tile, and plumbing is an improvement (capitalize and depreciate over 15 years). Replacing the compressor in an existing walk-in is repair (expense). Adding a new walk-in is improvement (capitalize over 5-year). Re-painting the dining room is repair; gut-renovating it with new finishes and built-ins is improvement. The routine maintenance safe harbor (Reg §1.263(a)-3(i)) lets you expense recurring activities that keep property in normal operating condition — HVAC tune-ups, hood cleanings, exterminator visits, pressure washing, light bulb replacements. NYC operators get this wrong by capitalizing too much (afraid of the IRS) when they should be expensing — the IRS audit angle is more often the opposite: expensing a clear improvement. When in doubt, ask your CPA before booking; the answer is binary and worth getting right.

When you abandon a leasehold improvement (close the restaurant, move out, leave the build-out behind for the landlord), you take a loss equal to the remaining undepreciated basis on Form 4797 in the year of abandonment under IRC §165. On a $800K build-out depreciated for 4 years out of 15, you'd have ~$640K remaining basis to write off as an ordinary loss. This is why CapEx tracking matters even after the build is done — operators who never separated their build-out costs by component can't take the full abandonment loss because they can't substantiate basis. Make sure your fixed asset ledger lists every CapEx item by category (LHI, FF&E, equipment), purchase date, original cost, and accumulated depreciation. If you sell the business as an asset sale, the leasehold improvements get a separate basis allocation on Form 8594 — different tax treatment than abandonment. Structure exit transactions with your CPA and a hospitality M&A attorney, not after the fact.

Buy outright when you have the cash and the equipment is core to the operation (combi oven, walk-in, range, bar refrigeration) — total cost is lower over 5-10 years and you own a depreciable asset. Lease (true operating lease) when the equipment becomes obsolete fast (POS hardware, espresso machines, dish machines on service contracts) or when you want to preserve cash for opening reserves. Equipment financing (a capital lease or equipment loan from CIT, Wells Fargo Equipment Finance, or specialty lenders like Restaurant Equippers) typically runs 8-12% APR over 36-60 months in 2026 and lets you preserve working capital while still getting full Section 179 expensing in year 1 (you depreciate as if you owned it because economically you do). For a brand-new NYC restaurant, financing 50-60% of the kitchen equipment package and using cash for the bar build-out and FF&E is the typical capital-light play — you front $300-400K of cash and finance the other $400-500K, paying ~$8-12K/month in equipment debt service from operations. Get the lease vs loan distinction right before signing — true leases don't get Section 179.

A TI allowance is taxable income unless properly structured — IRC §110 provides a safe harbor that allows a short-term lessee (lease term 15 years or less) to receive cash or improvements from a landlord for retail space without recognizing income, as long as the cash is used for qualified long-term real property and the lease specifically identifies it as such. For accounting purposes (GAAP and books), the TI allowance reduces the basis of the leasehold improvements you're capitalizing — so if your build is $1.5M and the landlord paid $500K of TI, your depreciable basis is $1.0M, and your rent expense (which should otherwise be reduced by the allowance amortized over the lease term) reflects the full cash rent paid. Modern NYC retail TI allowances run $50-150/sq ft on a 10-year lease for a hospitality buildout — a 3,000 sq ft restaurant might see $200-400K. Structure with your real estate attorney AND your CPA before lease signing; the §110 safe harbor language must be in the lease itself or you lose the income exclusion.

A clean fixed asset register is a single spreadsheet or accounting subledger with one row per capitalized asset and the following columns: asset ID/tag number, description, location/department, vendor, invoice number, in-service date, original cost, IRS depreciation class (5/7/15/39), book life, salvage value, depreciation method (MACRS half-year convention is standard), accumulated depreciation as of period end, net book value, and disposal date if removed. QuickBooks Online's built-in fixed asset module is weak; most NYC restaurant CPAs maintain it in Excel or a tool like Sage Fixed Assets ($1,500-3,000/yr) and post the monthly depreciation entry from there. Reconcile to GL every quarter. The two times a clean register pays off: (1) at year-end when the CPA needs to compute Section 179 + bonus depreciation eligibility on current-year additions, and (2) at sale/abandonment when you need to substantiate remaining basis. Operators without a clean register either over-pay tax or take aggressive write-offs that don't survive audit. Build the register on day one and update at every CapEx purchase, not at year-end.

IRS Rev. Proc. 2002-12 specifically allows restaurants to expense smallwares as supplies under the smallwares method, regardless of unit cost or useful life — plates, glassware, flatware, knives, pots and pans, sheet pans, hotel pans, bar tools, and similar items used in food and beverage operations. The election applies to opening packages and ongoing replacements. This is a major operator-friendly provision: a $40K opening smallware package gets fully expensed in year 1 instead of capitalized over 5 years. The election is automatic for new businesses and is made by simply expensing smallwares from day one (no separate IRS form). Existing operators switching methods need a Form 3115 filing — but most are already on the right method. Don't confuse smallwares with FF&E: a $400 immersion blender = smallware (expense); a $5,000 stand mixer = equipment (capitalize). Ask your CPA to confirm method election in your accounting policy memo; this is one of the cleanest tax wins available to a NYC operator.

Open a working capital line of credit in your second year of operation, after you have 12-18 months of clean financials to show — getting one before you have a track record means SBA-only or merchant cash advance, both painful. Size: 2-3 months of operating expenses is the standard target ($300-500K for a typical $2-3M NYC restaurant). The line covers seasonal swings (NYC restaurants typically see Jan-Feb and late Aug at 30-40% below summer/holiday peaks), unexpected equipment failures (a dead walk-in compressor is $8-15K), and bridges sales tax / payroll tax due dates when a slow week hits. Pricing in 2026: SBA Express line at Prime + 4.75% (around 12.25%); a conventional bank line at Prime + 2-4% (10-12%) once you have 2-3 years of financials and DSCR above 1.25x; a merchant cash advance at effective APR 50-100% (avoid unless terminal). Draw and repay regularly so the line stays active — banks pull unused lines after 12-24 months of zero draws.

SBA 7(a) loans up to $5M and SBA 504 loans (for owner-occupied real estate and major equipment up to $5.5M) are the workhorse capital sources for NYC independent restaurants — almost every multi-unit operator under 10 locations has at least one SBA-backed loan. Pros: 10-25 year amortization vs 3-5 year on conventional, smaller down payment (10% vs 25-30%), and approval for borrowers without massive collateral. Cons: personal guarantee from any owner with 20%+ equity (no exceptions), full collateralization including blanket UCC on business assets and often a second mortgage on personal real estate, processing time of 60-120 days, and packaging fees of $5-15K to a SBA-preferred packager. Top NYC SBA lenders for restaurants in 2026: Live Oak Bank, Newtek, Ridgewood Savings, M&T Bank, and Cross River Bank. Rates currently Prime + 2.5-4.75% (10.5-13%) on 7(a). Don't shop SBA rate alone — shop the relationship and the lender's restaurant experience, because covenant flexibility on a bad quarter matters more than 50bps of rate.

Almost never on the first 5-10 years of operating, and never on SBA. Personal guarantees from every owner with 20%+ equity are standard on conventional bank loans, equipment financing, real estate leases (NYC commercial leases routinely require 6-12 month personal guarantees, sometimes "good guy" clauses), and any merchant credit line. The only paths to limited or no personal guarantee: (1) operator with substantial track record (5+ profitable units, 7+ years) negotiating with a bank that knows them; (2) franchise operator with the franchisor providing partial guarantee; (3) PE-backed group with sponsor support letter. Even these usually carry a "limited" personal guarantee — capped at a fixed amount or burning off after 3 years of clean covenant compliance. Negotiate the burn-off and the cap, not the existence. Most NYC operators try to negotiate "good guy" lease guarantees that limit landlord recourse to unpaid rent through the date of vacancy and surrender — this is the most achievable carve-out and typically saves 6-12 months of personal exposure on a bad exit.

Merchant cash advances (MCAs) — the daily-debit financing offered by OnDeck, Kapitalus, Forward Financing, Yellowstone, and dozens of others — carry effective APRs of 50-150% and trap operators in renewal cycles that are mathematically impossible to escape without a major capital event. Structure: lender advances $100K against 12 months of credit card sales, you repay $135K via a daily ACH of $540 from your operating account regardless of revenue. On a slow February week, that $540/day takes 10-15% of your gross — and the math gets worse if you stack a second MCA on top. NYC restaurant lawyers describe MCAs as "credit card debt for restaurants except worse," and the NY Attorney General sued five major MCA providers in 2023 over usury and harassment. Avoid by: building a real working capital line in year 2, keeping 60-90 days of operating cash in reserve, and maintaining a 2nd-tier lender relationship (Square Capital, Toast Capital, American Express Business Line) for emergency liquidity at 1.10-1.25 factor (still expensive but not predatory). If you're already in an MCA stack, talk to a restructuring attorney before signing the next renewal — there are settle-down strategies that work.

Conventional and SBA restaurant loans typically carry 3-5 financial covenants tested quarterly: (1) Debt Service Coverage Ratio (DSCR) of 1.20-1.35x — EBITDA divided by total annual debt service, your most-watched number; (2) tangible net worth minimum (often the SBA loan amount or a fixed dollar floor); (3) maximum total debt-to-EBITDA of 3.5-5.0x; (4) maximum CapEx in any year without lender consent ($100-250K typical); (5) minimum cash on hand at quarter-end ($50-150K). Operating covenants: maintain primary depository accounts at the lender bank, deliver quarterly financials within 45 days, deliver annual CPA-reviewed financials within 120 days, no change of ownership without consent, maintain insurance with lender as additional insured. Trip a covenant and you're in technical default — the lender can call the loan, but in practice they issue a waiver letter for a fee ($1-5K) on a one-off and renegotiate covenants if it's chronic. Be the operator who calls the lender BEFORE the covenant trips with the explanation; lenders punish surprises 10x more than they punish bad numbers.

Refinance when at least one of three conditions is true: (1) prevailing rates are 100+ bps below your current loan rate (with rates having moved meaningfully in 2024-26, many 2022-vintage SBA loans at 11-13% can refinance to 9-10.5%); (2) you've grown into better terms — moved from SBA to conventional, or from a community bank to a regional with better rates; (3) you need to release a personal guarantee, replace a covenant set, or extend amortization to free up cash flow. Refinancing costs: typically 1-2% of loan balance in fees plus prepayment penalty on the old loan (SBA 7(a) prepayment penalty is 5/3/1% in years 1-3 on loans 15+ years; conventional bank loans vary). Run the math: a $1M loan at 12% refinanced to 10% saves $20K/year in interest, payback period on $20K of refi costs is 12 months. Don't refinance just to extend the term and lower payment if the underlying business needs work — that's putting cosmetics on a broken income statement. Talk to 3-4 lenders, get term sheets, and let your existing lender match rather than refinance if relationship matters.

Standard restaurant loan package for SBA or conventional in 2026: (1) 3 years of business tax returns plus YTD P&L and balance sheet through prior month-end; (2) 3 years of personal tax returns from each 20%+ owner; (3) personal financial statement (SBA Form 413) from each guarantor showing assets, liabilities, real estate, and contingent liabilities; (4) 12-month projection (P&L and cash flow) with assumptions; (5) executive summary / business plan including ownership structure, management bios, market analysis; (6) copy of fully executed lease and any amendments; (7) copies of all licenses (liquor, food service permit, certificate of occupancy); (8) accounts receivable and accounts payable aging if applicable; (9) debt schedule listing every existing loan and lease; (10) AKAs/d/b/a's, organizational chart, articles of organization, operating agreement. Total package is usually 80-150 pages. Use a SBA-preferred packager ($5-15K fee) for SBA loans — they know the lender's underwriter and clear closings 30-45 days faster. Underwriting at most lenders takes 30-60 days from complete package; allow 90-120 days from first conversation to funded.

Both, in different roles. A bookkeeper does the monthly close, AP, payroll coordination, sales tax filings, and management reports — typically $1,500-4,000/month for a single-location NYC restaurant depending on volume and complexity. A CPA does the annual tax return, tax planning conversations, depreciation strategy, year-end adjustments, audit defense if you get one, and entity-level questions (S-corp election, partnership allocations, 1031 exchanges if you own real estate) — typically $2,500-7,000/year for a single-location restaurant on the tax return side. Don't hire a CPA to do bookkeeping at $200/hr when a $50/hr bookkeeper can do the same work; don't hire a bookkeeper to file your tax return because they can't sign the federal return as a paid preparer (must be EA or CPA) and won't represent you on audit. Ideal setup: a restaurant-specialist bookkeeping firm handles month-to-month, a CPA reviews their work quarterly and files the annual return. Both should know NYC restaurant tax issues cold — generic CPAs make expensive mistakes on tip credit, hospitality wage order, and NYS sales tax.

Ask three questions before signing an engagement letter. (1) How many restaurant clients do you have, and what's the revenue range? You want a CPA with 10-30 active hospitality clients in your size band ($500K-$5M for an independent), not a generalist with 1-2 restaurants. (2) Walk me through how you handle the FICA tip credit, the smallwares method, hospitality wage order spread-of-hours, and NYS service charge taxability. If they fumble any of these, walk. (3) Have you defended a NYS DTF sales tax audit on a restaurant in the last 3 years? If yes, ask how it went. Top NYC restaurant CPA firms in 2026 include PKF O'Connor Davies, Citrin Cooperman (large), Janover, Marks Paneth (now part of CBIZ), Anchin (private clubs and high-end hospitality), and a deep bench of mid-size firms (Mazars, BDO, Withum). For small operators, look at boutiques like Schiavo & Co, Berdon (now Citrin), or solo CPAs who came out of those firms. Reference checks with 2-3 current clients are non-negotiable. Industry membership (NYC Hospitality Alliance, NYSRA) signals real bench.

Single location under $5M revenue: outsource everything (bookkeeping firm + CPA). Two to five locations or $5-15M revenue: hire a director of finance / controller in-house at $90-130K/yr in NYC, keep the CPA outsourced for tax, keep some bookkeeping outsourced for transaction-level work. Five+ locations or $15M+ revenue: full in-house finance team — controller, AP/AR clerk, payroll specialist, plus a fractional CFO (Fractional CFO market in NYC: $4-12K/month for 1-2 days/week of CFO-level work) or full-time CFO at $180-275K. The trigger to bring finance in-house is when you have so many transactions and decisions per week that the outsourced firm can't keep up with response time, not when you cross a revenue threshold per se. Signs you need to bring it in: outsourced bookkeeper is delivering month-end past the 25th; AP errors are costing real money; you can't get a P&L on demand; or your bank/lender wants weekly cash flow updates and you don't have time to build them. Hire fractional CFO before full-time — most NYC operators don't need 40 hrs/week of CFO until they're $20M+.

Scope of work spelled out specifically — not "accounting services" but "monthly close including bank rec, AP entry up to 80 invoices/mo, payroll coordination, sales tax filing for NYS Form ST-100 quarterly, monthly P&L delivery by the 15th business day." Anything outside scope is a separately billed project (you don't want surprise $3K invoices for "answering questions about the loan application"). Fee structure: monthly fixed fee (preferred for predictability) or hourly with a not-to-exceed cap. Termination clause with 30 days notice and turnover obligations (they hand you the QuickBooks file, all source documents, and a clean trial balance — get this in writing or you'll fight to extract data). Confidentiality and data security language (especially important if they'll have access to your bank logins). Audit defense — does it include responding to a NYS DTF or IRS notice? Most engagement letters carve out audit defense as separate hourly. Get the letter signed before any work starts and renew annually with updated scope.

Most independent NYC restaurants run as a multi-member LLC taxed as a partnership (passes through income to owners' personal returns, easy to allocate special allocations to investors, no double tax). Single-owner restaurants run as a single-member LLC with S-corp election (pass-through plus payroll-tax savings on owner distributions above a reasonable salary). Avoid C-corp unless you're raising serious institutional capital or planning a strategic exit to a corporate buyer where Section 1202 (qualified small business stock) applies. The decision drives: how investors are paid (preferred return + waterfall in LLC vs dividends in C-corp), how you take money out (guaranteed payments + distributions in partnership vs salary + dividends in S-corp/C-corp), how losses pass through (immediately useful in LLC/partnership/S-corp; trapped in C-corp), and how the entity sells (asset sale vs stock sale tax math is wildly different). NYC operators planning to scale to multi-unit usually create a holding company LLC with operating LLCs underneath, one per location — your CPA and a hospitality M&A attorney structure this together before the second unit opens.

Run through this checklist by mid-November. (1) Review YTD P&L vs projection — if you're trending high-income, accelerate equipment purchases (Section 179 + 20% bonus dep in 2026) before year-end; if low-income, defer purchases to next year. (2) Pay year-end bonuses by Dec 31 to deduct in current year (must be paid by 3/15 of next year for accrual basis, but cleaner to pay in current year). (3) Take physical inventory on Dec 31 morning. (4) Process all open AP into the current year — don't let January-dated invoices for December services slip into next year. (5) Confirm all retirement plan contributions (Solo 401(k), SEP-IRA, SIMPLE) are funded by deadlines (most by tax filing deadline including extensions). (6) Reconcile sales tax payable balance to the December filing. (7) Issue 1099-NEC to contractors paid $600+ during the year (must be filed with IRS by Jan 31). (8) W-2 forms generated from payroll provider (deadline Jan 31). (9) Schedule a year-end planning call with your CPA to discuss owner distributions vs payroll mix for the next year. (10) Pull a tax projection for state estimated payment due Jan 15.

If you're an LLC or S-corp owner with restaurant income flowing to your personal return, you owe quarterly estimated taxes to IRS (Form 1040-ES) and NYS (Form IT-2105) on April 15, June 15, September 15, and January 15. Safe harbor: pay either 110% of prior year's total tax (if AGI was over $150K) or 90% of current year's tax to avoid underpayment penalty. Underpayment penalty rate for 2026 is 8% federally and 8% NYS — not catastrophic but it stings on a $30K shortfall. NY State also has the PTET (Pass-Through Entity Tax) election under Tax Law §860, where the LLC pays NY tax at the entity level (currently 6.85-10.9% bracketed) and the owner gets a federal deduction for it that bypasses the $10K SALT cap — this is a 2-3% effective federal tax savings on every dollar of NY income, and almost every NYC restaurant owner over $50K of income should be electing PTET annually. The PTET election is made by March 15 each year for the current year — easy to miss and not retroactive. Talk to your CPA in February.

S-corp owner-operators must take a "reasonable salary" (W-2 wages with full payroll taxes) for services rendered before taking any distributions — IRS audits hit S-corp owners hard for under-paying salary to dodge FICA. For a NYC restaurant owner-operator who works as the GM, reasonable salary is typically $90-150K depending on revenue and hours; for an absentee owner who only oversees, it can be much lower ($30-60K) since they're not providing significant services. The IRS reasonable comp test looks at: comparable wages for similar roles in the area (NYC GM range), hours worked, business size and complexity, training, and what you'd pay an outside hire. Distributions above salary avoid FICA (saves 15.3% on Social Security + Medicare up to the wage base) — that's the main benefit of S-corp election. Get this wrong and the IRS reclassifies distributions as wages, hits you with back FICA + interest + penalties. Document compensation rationale annually with comparable salary data; your CPA should run this analysis at year-end. Multi-unit operators usually move salary upward as they take on more management responsibility and as concept revenue grows.

Top patterns in order of frequency: (1) bartender no-rings — pouring drinks for cash that never hit the POS, pocketing 100% — typically caught by a bar pour test that shows 4-8% liquor variance and POS sales/seat audits showing too few drinks per hour for visible bar volume; (2) void/comp abuse — voiding a cash sale after the guest pays so the sale doesn't show in POS — caught by tracking voids by employee and reason; (3) over-rings to skim cash — entering $40 on the POS for a $35 cash sale and pocketing the difference — caught by analyzing cash variance per shift; (4) friend discounts and unauthorized comps — caught by comp percentage tracking by employee; (5) refund fraud — running a fake refund to a personal card — caught by manager-required refund authorization with second sign-off. Industry benchmark: untracked theft costs the average restaurant 3-5% of revenue (per US National Restaurant Association data). The single best control is a cash variance tolerance of $5/shift with anything over investigated within 24 hours, and POS access restricted by role with manager-only voids and comps. Cameras over the bar (point-of-sale view) cost $1,500-3K and pay for themselves in 60 days.

Hard controls in the POS: (1) only managers (PIN-protected) can authorize voids and comps over $5; (2) every void/comp requires a written reason code from a fixed list (manager comp, server error, kitchen error, marketing promo, VIP, employee meal); (3) daily comp/void report emailed automatically to the operator each morning showing every void/comp from prior shift with employee, amount, and reason; (4) total comps/voids as percent of gross sales tracked weekly with a 2-3% ceiling; (5) any single employee with comps above 3% of their checks gets reviewed weekly. Soft controls: train every manager that a comp without a documented reason is a fireable offense; rotate manager assignments so the same person isn't always closing out the bar; spot-check 5-10 comps per week against actual table activity. The single biggest red flag is an employee whose comp dollars cluster on cash-tendered checks — that's not service recovery, that's theft. Operators who skip comp/void reporting lose 1-2 points of revenue annually to internal abuse.

Cash drawer counted at end of shift by cashier, signed off, and dropped into a manager safe with a witness — never carried home, never "just deposited tomorrow morning." Every drop bag is logged with serial number, shift, employee, and amount. Manager combines drops into a single deposit, seals into a tamper-evident bag, and hands to armored car (Loomis, Garda, Brinks at $90-180/wk). Reconcile the bank deposit to POS Z-tape to drop log within 24 hours of credit posting — if any of those three numbers don't match, investigate immediately. Common fraud: manager pockets the variance between Z-tape and drop while telling the cashier they were short and the bookkeeper they were over. The control is daily 3-way reconciliation with a second person reviewing weekly. NYS DTF audit angle on undeposited cash: any cash collected and not deposited becomes presumed unreported revenue, and they reconstruct via purchase markup at audit — a $200K hole becomes a $230K assessed sales tax bill plus penalties. Never let undeposited cash sit more than 48 hours; a safe full of cash is an invitation to theft and a paper trail problem with the state.

The most common vendor fraud pattern in NYC restaurants is the kickback — chef or purchasing manager directs business to a vendor who inflates invoices and kicks back 5-15% personally to the manager. Caught by: (1) pulling 12 months of invoices from the same vendor and looking for unit price drift relative to market (Baldor's published prices are a benchmark even if you don't use them); (2) requiring 3 quotes for any vendor over $5K/year and rotating spec audits; (3) separating ordering, receiving, and AP approval — three different people, never the same person; (4) random monthly invoice audits where the operator personally calls 5 random vendors to verify recent invoices and pricing. The other pattern is fictitious vendor — someone in your org sets up a fake vendor (often a slight variation on a real vendor name) and submits fake invoices that get paid into a controlled account. Caught by: requiring W-9 on file for every vendor before first payment, verifying address and EIN against IRS TIN match, and requiring physical receipt of goods documented before AP processes payment. NYC operators who skip these basics typically lose 1-3% of food cost to vendor fraud over 3-5 years.

Six early warning signs that should trigger immediate investigation. (1) Pour cost or food cost climbing 1.5+ points over trailing 3-month average with no menu price change — theft, waste, or under-portioning of sales. (2) Cash sales as a percent of total dropping faster than NYC's secular shift to cards (which is real, ~1-2 pts/year) — could be staff diverting cash. (3) Voids or comps above 3% of gross sales — control failure. (4) AP aging stretching past 60 days with no operating-cash explanation — you're behind on cash even though P&L looks fine, which usually means a balance sheet leak (uncollected catering, growing tip liability, unrecorded loan). (5) Sales tax payable balance growing larger than what you'll file — you've collected tax you're not remitting, which is the path to a NYS warrant. (6) Bank account balance not matching the cash flow forecast within $5K — something is moving without your knowledge. Run all six checks every Monday morning; the operator who catches a fraud pattern in week 2 saves 50x what the operator who catches it in month 6 saves. The biggest financial mistakes in NYC restaurants come from looking at the P&L without looking at the balance sheet — review both, every month, side by side.