Loyalty, Membership & Retention

Wait 90-120 days unless you are a fast-casual concept doing 200+ covers a day where frequency is the entire economic story. Full-service NYC restaurants in the $60-120 check range get more lift in year one from a tight reservations + email database than from points, because guests do not visit you weekly. Use the first quarter to capture emails at the door (Sevenrooms or Resy autocaptures roughly 70-85% of cover names) and build a 3,000-5,000 person list before you spend a dime on a loyalty platform. Then layer on points or visit-based perks once you can see your real frequency curve in the POS data. Launching loyalty before you know your repeat rate is how operators end up paying $200-400 a month for a tool nobody redeems.

Points work best at concepts with a check under $35 and a target visit frequency of 2+ times per month — coffee, fast-casual, pizza-by-the-slice, ghost kitchens, neighborhood bars with food. Sweetgreen, Joe Coffee, &pizza, and Chopt all run points or visit-based programs and report 15-30% lift in monthly visit frequency from active members. At a $90+ AOV white-tablecloth restaurant, points feel cheap and dilute the brand — those operators should run a CRM-driven recognition program (preferred table, comped amuse, remembered allergies) instead. The breakeven math: if your loyalty cost is 3-5% of revenue from members, you need members to spend at least 20% more than non-members for the program to pay back, which only happens at high-frequency concepts.

Points programs ($1 = 1 point, redeem at 100 points) are the most flexible and the most generic — guests rarely remember the conversion ratio. Visit-based (buy 9, get the 10th free) is the simplest and works best at coffee, salad, and quick-serve concepts because the reward is concrete. Spend-tier (Silver/Gold/Platinum unlocks at $500/$1500/$3000 annual) is the right model for full-service restaurants, hotels, and high-AOV bars where status and recognition matter more than discounts. Pick visit-based if your check is under $25, points if you sell across multiple dayparts and want flexibility, and tiers if your average member is going to spend $1,000+ a year with you.

Industry standard is a 5% effective reward rate — $5 off every $100 spent — and that should be your ceiling unless you are a coffee concept with 70%+ contribution margin where 8-10% still pencils. Below 3% guests do not engage; above 6% you are funding visits that would have happened anyway. The math you must do: take your contribution margin (revenue minus food cost minus variable labor), multiply by your incremental visit lift, and confirm that exceeds the reward cost. Most operators forget that 60-70% of redeemers would have visited without the reward, so only the incremental 30-40% counts as new revenue. A safe target is 4% reward rate plus 1% in surprise-and-delight (free birthday item, anniversary upgrade) for the top 20% of members.

A reasonable target is $3-8 in CAC per enrolled member at a fast-casual NYC concept, and $8-20 at full-service where each member is worth more annual revenue. The math: if your average loyalty member spends $400/year with you and your contribution margin is 20%, that member is worth $80/year — pay no more than 25-30% of year-one value to acquire. Door capture (QR at table, prompt at POS) is the cheapest channel at $1-3 per enrollment because the guest is already onsite. Paid social to drive sign-ups runs $15-40 per enrollment in NYC and almost never pencils unless you are a multi-unit operator. Track enrollment-to-second-visit rate — anything under 40% means your reward is too weak or your onboarding email is missing.

Yes for an independent or soft-branded property; no for a flagged Marriott/Hilton/Hyatt where the master agreement requires Bonvoy/Honors/World of Hyatt to be the primary loyalty currency. Independents like The Greenwich, The Beekman, or Made Hotel can run a standalone F&B program through Sevenrooms or a dedicated platform like Bond and offer in-house perks (free welcome drink, comped corkage, priority restaurant booking) that reward both hotel guests and outside locals. The trick is to enroll outside-the-hotel diners separately so you build a neighborhood book of business, not just transient guests who will not return. Budget $500-1,500/month for the platform plus 10-15 hours of GM time monthly to actually run it.

Traditional points programs flop at bars because guests are intoxicated, transactions are split, and tabs close fast — the redemption friction kills it. What works at NYC bars is a $20-50/year bar membership card (think Death & Co's old Velvet Curtain, or 86 Co's Spirits Club) that unlocks a reserved seat, a house cocktail at cost, and a no-cover-charge perk on weekends. Bottle clubs at high-volume bars (a numbered bottle of bourbon kept on the back-bar with the member's name) drive 10-20 visits per member per year and lock in $800-2,000 of annual spend. The other format that works is birthday-month captures — get every birthday on file and send a free bottle of cava or shot card the week of the birthday, redeemable with 4+ guests.

Toast Loyalty runs $25-75 per location per month as an add-on if you are already on Toast POS, which is dramatically cheaper than the $200-500/month standalone competitors. The catch: it only works if your POS is Toast, and the feature set is intentionally simple — points or visit-based, basic email, no real campaign automation, no advanced segmentation. For a single-location NYC restaurant doing under $3M revenue on Toast, it is the right answer almost every time because the integration is free and the data flows back to your POS reports. Step up to Punchh or Thanx only when you have 3+ locations, $5M+ revenue, or you need real lifecycle email automation that Toast cannot do.

Punchh (owned by PAR since the $500M April 2021 acquisition) is built for multi-unit fast-casual and QSR concepts with 5+ locations and serious mobile-app ambitions — think the Sweetgreen, Cava, &pizza tier. Pricing starts around $200/location/month and climbs to $500+ once you add the white-label mobile app, advanced segmentation, and POS integration fees. It is overkill for a single-location operator and underpowered for hotels or full-service. Where it shines: pre-built A/B campaign experiments, geo-fenced push notifications, and integrations into roughly 20+ POS systems including Brink, Aloha, and Square. If you are not building a custom-branded mobile app, you do not need Punchh.

Thanx is the right choice when you want loyalty plus first-party online ordering plus campaign automation in one stack, and you are willing to migrate ordering off DoorDash/Uber direct and onto Thanx-powered web ordering. Pricing is $300-600/location/month all-in, comparable to Punchh but with a stronger emphasis on direct-channel revenue capture and credit-card-linked offers (no app required — guests just register their card). Thanx works particularly well for fast-casual with 3-15 NYC locations where the operator wants to break dependence on third-party delivery. The credit-card-linked model also works for full-service concepts where guests resist installing yet another app. The main weakness is that Thanx does not own the POS layer, so integration depth varies by which POS you run.

Yes, if you are already on Square POS, Square Loyalty at $45/location/month for the first location ($25 for additional) is the cheapest viable program in market and integrates with zero engineering. The feature set is intentionally narrow — visit-based or points, basic email/SMS, no segmentation, no real lifecycle automation — but for a single-location coffee shop, juice bar, or bakery doing under $1.5M revenue, that is plenty. Enrollment happens at checkout via phone number, which gets you 60-80% capture rates without the friction of an app download. The platform also handles SMS sending compliantly, which is one less vendor to manage. Outgrow it once you hit 3+ locations or want to send segmented campaigns.

For full-service NYC restaurants — the kind on Sevenrooms or Resy OS — the built-in CRM with auto-tags, visit-count tracking, lifetime-spend fields, and email automation is enough for 90% of what an operator needs. Sevenrooms (Bain Capital acquired in November 2023, then sold to private equity) prices around $300-600/month per location depending on tier and includes the marketing automation, guest profiles, and reservation-data linkage that small loyalty platforms cannot match. You would only add a dedicated points/visits platform on top if you have a fast-casual sister concept where points genuinely drive frequency. For a tasting-menu spot or steakhouse, the recognition-tier model in Sevenrooms (VIP/Regular/First-Time tags) is more valuable than any points program.

Demand a same-POS demo where you ring up a real check, apply a real reward, and watch the data flow back into the POS report — anything less and you will discover integration gaps after you have signed a year-long contract. The four integration tiers from worst to best: (1) reward applied as manual discount at POS, no return-data flow; (2) item-level discount auto-applied via integration but no member ID in POS report; (3) full bidirectional sync where POS check feeds member profile and rewards apply automatically; (4) full sync plus loyalty status visible to the server on the check screen. Toast Loyalty is tier 4 because it owns both layers; Punchh and Thanx are tier 3 with most major POS systems; small standalone tools are usually tier 1-2. Tier 1 kills the program — servers forget to apply, guests blame the brand.

Plan 6-10 weeks from contract signature to live program for any platform deeper than Square or Toast bundled — the work is not the software, it is your menu mapping, reward structure, email templates, staff training, and signage. The compressed timeline (4 weeks) only works if you are already on the same vendor's POS and willing to launch with the default reward structure. For a 3-location concept moving to Punchh, expect a kickoff week, 2 weeks of menu/POS mapping, 2 weeks of campaign configuration and creative, 1 week of staff training across all locations, and 1-2 weeks of soft-launch testing before a public push. The biggest delay killer is creative — get loyalty signage, table tents, email templates, and POS screen designs done before you sign the contract, not after.

The successful NYC F&B membership model (Casa Cipriani at ~$3,800-5,000/year, ZZ's Club at ~$10,000-15,000/year initiation plus dues, Aman Club at $200,000 initiation) sells access to a private space, not points or discounts — the dues are the revenue, and F&B is the experience. For a $300/month tier, members typically get unlimited access to the space, member-only events, and 10-20% off F&B as the only transactional perk. Math the operator must hit: 200 members at $300/month = $720K annual recurring revenue, which covers fixed overhead on a 2,000-3,000 sq ft member-only space. The model breaks if you discount F&B more than 20% because dues and food revenue start cannibalizing each other. Pick a single perk model — either dues-funded space-access or transactional-discount, never both deeply.

The proven path is founding-member cohorts: announce 100 founding spots at 30-50% off lifetime dues, sell them to your existing top-spenders and known industry friends through 1:1 outreach over 60-90 days before opening. Soho House did exactly this — every NYC house launches with a curated founding cohort built from member referrals at adjacent locations. ZZ's Club, Casa Cipriani, and The Twenty Two NY all leveraged their parent restaurant or hotel guest lists as the seed pool. Public marketing for membership clubs is mostly counterproductive because exclusivity is the product — referrals and a waitlist signal scarcity. Budget $50-100K for the founding-member events and concierge outreach in the 90 days pre-open; the founding cohort then becomes your referral engine for years.

Three tiers max — more confuses guests and bloats the back-end. A clean structure: Local at $150/month (one guest pass per visit, 10% off F&B, member events), Standard at $300/month (unlimited guests with member, 15% off, priority reservations 7 days out), and Founders at $500/month (private dining room access twice a year, 20% off, lifetime locked dues). Cap the Standard and Founders tiers at a number you can actually serve — a 100-seat space cannot honor 1,000 unlimited-guest memberships without destroying the experience. The under-30 introductory tier (Soho House Cities Without Houses model) at $80-120/month captures younger members who graduate up. Re-price tiers every 18-24 months on new sign-ups only; never raise on existing members or you trigger churn waves.

Yes — initiation fees do three things: they fund opening capex, they signal exclusivity, and they reduce churn because nobody walks away from a $1,500-5,000 sunk cost casually. The NYC standard for sub-$500/month tiers is $500-2,500 initiation; for $500+/month tiers it climbs to $5,000-25,000; for ultra-tier ($1,500+/month) it can hit $50,000-200,000 (Aman). Structure initiation as non-refundable but creditable against the first year of dues to soften the ask. The waiver tactic — comp the initiation for founding members or referral-driven sign-ups — is more powerful than discounting dues because it preserves your published price for the next cohort. Track initiation revenue separately from dues for tax and reporting purposes — it is one-time and should not be amortized into MRR projections.

Target under 15% annual churn for a healthy paid F&B club; over 25% means the value is not landing. Soho House publicly reported around 7-9% pre-pandemic and roughly 10-12% post-pandemic across global houses, which is the gold standard. NYC-specific clubs that re-priced aggressively (Soho House raised dues 25% in early 2024) saw churn spike to 15-20% temporarily before stabilizing. The two leading churn predictors: visit frequency in the first 90 days (members who do not visit twice in 90 days churn at 4x the rate) and event RSVPs in the first six months. Build a 90-day onboarding sequence with three named events the new member must be invited to, and concierge outreach at day 60 if they have not visited.

The proven stack is Stripe Billing or Recurly for recurring dues ($2-5K/month at scale), Salesforce or Hubspot for CRM ($1-3K/month), Sevenrooms for reservations and member recognition ($500-1,000/month), and a custom or off-the-shelf member portal (ClubExpress, Glofox, or a Webflow + Memberstack build) running $500-2,000/month. Larger clubs (Soho House, Aman) build custom — under 1,000 members, off-the-shelf is fine and faster. The single most important integration is Stripe-to-Sevenrooms — when a member's card declines, the host stand needs to know before they walk in for dinner. Budget $80-150K for first-year stack including build, integrations, and one part-time membership manager. After year one, ongoing software cost should land at $4-8K/month depending on scale.

NY General Business Law §396-i prohibits any expiration date earlier than 9 years from the date of issue, bans monthly service or dormancy fees for the first 25 months, and requires that terms (issue date, fees, expiration) be printed on the card itself or made available in writing on the receipt. The federal CARD Act of 2009 sets the floor at 5 years and bans inactivity fees in the first 12 months — NY is stricter on both. As of 2026, NY also requires that you redeem balances under $5 in cash on request (per a 2018 amendment) and prohibits replacement fees for lost cards if the card is registered. Print this disclosure on every physical card and on the digital receipt for e-gift cards or you are exposed to per-violation fines starting at $250.

Under ASC 606 (revenue recognition), restaurants can recognize breakage income proportionally as redemptions occur, using historical redemption patterns to estimate the unredeemed portion. Industry breakage rates run 5-15% for restaurant gift cards (lower than the 10-20% retail rate because food gifts get used). For example, if your 3-year history shows 90% of cards are redeemed, you can recognize the 10% breakage as revenue proportionally over the redemption curve, not all at once. NYS escheat law (Abandoned Property Law §1315) generally exempts merchandise gift certificates from escheatment — you do not have to remit unredeemed balances to the state, which is why breakage is real income. Track gift card liability as a separate balance-sheet account and reconcile quarterly with your POS or gift card platform report.

Use POS-native (Toast, Square, Resy) if you are single-location or 2-3 locations and 95% of redemptions happen in-house — it is free or near-free and the integration is bulletproof. Move to a dedicated platform like Givex, Valutec, or Paytronix once you have 3+ locations and need cross-location balance tracking, e-gift card e-commerce, and merchant-funded reporting. Givex pricing runs $50-150/location/month plus $0.50-1.50 per card issued; Toast Gift Cards is $0 monthly plus $0.50 per physical card. Paytronix is the heavyweight at $200-500/location for full loyalty + gift card stack, worth it only at 5+ locations. The most underrated feature: real-time balance lookup that the server can do at the table without leaving the POS — anything less creates redemption friction.

The proven NYC holiday play is bonus-card stacking: buy a $100 gift card, get a $20 bonus card valid Jan 15-March 15. The bonus card pulls traffic into your slowest months, the original card gets used at full margin during peak season, and the effective discount is 17% — but only on the bonus portion which is redeemed in low-fill-rate slots. Launch the promo Black Friday through Dec 24, with email sends weekly and signage at the host stand. Industry data shows operators see 30-50% lift in November-December gift card sales with bonus structures versus flat 10% off. Avoid percentage-off-card-purchase promotions ($100 card for $90) because they attack year-round margin and train the guest to wait for sales.

Build a B2B gift card sales channel with 5-10% bulk discounts at $5K+ purchase volume — corporate buyers (HR, EA, marketing) will absorb a $50-100 e-gift card to 50-200 employees easily, and your effective revenue is recognized immediately while redemption happens over 6-18 months. Use a platform like Givex, Toast for Restaurants Gift Cards Bulk, or a custom Stripe-powered checkout that lets a buyer place a single order, attach a custom message, and bulk-email cards to a CSV of recipients. Reach out to 50-100 NYC corporate concierge desks (Convene, Industrious, WeWork, AmEx Centurion concierge) and EA networks (NYC Executive Assistant Network on LinkedIn) starting October 1. A single corporate program can move $50-200K of gift cards in November alone with about 30 hours of sales work.

Three controls cut e-gift card fraud by 80-90%: enable AVS and CVV match at checkout (reject any decline), set a velocity limit of 3-5 cards per IP per 24 hours, and hold any card over $200 for manual review for 24 hours before activating. The most common fraud pattern is a fraudster buys a $500 e-gift card with a stolen card, redeems it within hours at the bar, and the chargeback hits 30-60 days later — you eat the merchandise and the chargeback fee ($15-25 each). Stripe Radar, Toast Gift Card fraud rules, and Givex all offer velocity controls but they are not on by default. Also avoid offering 'gift card to anyone via email link' with no recipient verification — require the purchaser to enter a recipient email and have the card delivered to that address only.

Skip Raise and CardCash unless you have a serious cash-flow gap — they take 8-15% of face value and the guest who buys a discounted card at 80 cents on the dollar is rarely your target customer. Groupon and similar daily-deal platforms are even worse, taking 50% of the deal and another 5-10% in processing, so a $50 voucher nets you $22-25 — under your food cost on a $50 check at a full-service restaurant. The only platforms worth considering are e-gift card marketplaces inside corporate-rewards programs (Tango Card, Tremendous, Reward Gateway) where corporate buyers pay full face value plus a small premium. The math on discount platforms: if you need cash this badly, your loyalty program and email list are not working — fix the demand engine first.

Coffee subscriptions work (Pret A Manger's Club Pret at $30/month for up to 5 drinks/day did $1.5B+ in subscriber revenue globally in its first 24 months, Panera's Unlimited Sip Club at $14.99/month has 50M+ enrolled). Meal subscriptions for full entrées have mostly failed in restaurants outside of fast-casual because the unit economics break — Sweetgreen's SG Rewards dropped its meal subscription test in 2023. NYC concepts that have made subscriptions work: Joe Coffee ($35/month unlimited drip), Joe & The Juice ($59/month unlimited beverages), and Brooklyn-based Devoción (small-batch coffee subscription at $24/month). Keep the math simple: COGS at 25-35% means you can give away 3-5 items per customer per month and still net positive, but only if the subscription drives at least 1.5x more visits than non-subscribers.

Price the membership at 75-85% of the cash value of the included dinners, so a 2-dinners-per-month prix-fixe at $75/dinner cash should sell at $115-130/month — the discount is the hook, but the recurring billing is the lock-in. The model works because the no-show or under-redemption rate runs 20-30%, meaning you collect dues whether or not the member shows up, and the fixed 5:30/9:30 PM seating slots assigned to members improve your floor utilization. Use Stripe Billing or Recurly for the recurring charge, and Sevenrooms for the seat-block management. The breakeven: if 30% of seats are held for members and 70% redeem on schedule, your effective revenue per held seat is roughly equivalent to a walk-in cover, with zero acquisition cost. Cap the program at 50-100 members per restaurant or you will choke your reservation grid.

Two formats work in NYC: the 'curated bottle' subscription ($75-150/month, member picks up 1-2 bottles selected by your sommelier with tasting notes) and the 'wine credit' membership ($100/month banks $120 of in-restaurant wine credit, expiring quarterly). NY State liquor law (ABC Law §63 and §64) restricts off-premise sales to licensees with the right permit class — a standard on-premise restaurant license cannot legally sell take-home bottles unless you hold a Restaurant Wine License with Off-Premises Endorsement (added 2021) or partner with a retail wine shop. Verify your license class with the SLA before launching. Wine credit memberships sidestep the off-premise issue entirely because the wine is consumed in-restaurant. Target 5-10% of your top 200 customers — a 50-person club generates $5-7.5K/month in pre-paid revenue and locks in 2-3 visits per member per quarter.

Expect 5-8% monthly churn in the first 6 months, settling to 3-5% monthly steady-state — meaning a healthy 12-month retention curve lands at 50-65%. Anything worse than 8% monthly steady-state means the value is not landing or your subscriber pool is too price-sensitive. The two churn killers: members who do not redeem in their first 30 days churn at 3-4x the rate (pre-empt with a 'your benefit expires' reminder at day 21) and members billed during a slow month with no recent visit churn at 2x (pre-empt with concierge outreach for any subscriber without a visit in 45 days). Track involuntary churn (failed cards) separately — it is usually 30-40% of total churn and recoverable with Stripe Smart Retries or a manual outreach playbook.

NYC sales tax (8.875% combined) generally applies at the point of redemption for prepared food and beverage, not at the point of subscription billing — the dues are payment for future taxable goods, not a service. So a $100/month subscription that is redeemed for $120 of food triggers $10.65 in sales tax (8.875% on the actual $120 menu value, not on the $100 paid). NYS Department of Taxation and Finance guidance on prepaid restaurant cards (TSB-M-09 series) treats subscription dues like gift card sales for tax purposes — collect at redemption. The exception: if the subscription is structured as a 'fixed-price meal plan' with no menu choice, NYS may treat the dues as taxable on receipt. Get an opinion letter from a NYS sales tax CPA before launching anything beyond a simple credit-bank model — penalties for misclassification run 10-25% plus interest.

Yes, and it is one of the highest-margin plays an underutilized hotel restaurant can run — a $99/month subscription for unlimited weekday breakfast typically captures office workers within 4-6 blocks and fills empty 7-10 AM seats at near-zero marginal cost. The Beekman, The Greenwich, Fouquet's NY, and several boutique hotels have piloted neighborhood subscriptions with 100-300 subscribers each, generating $100-300K in annual recurring revenue from previously dead daypart capacity. Cap the subscription at 5 visits per week per subscriber and exclude weekends to protect your higher-margin in-house guest covers. Marketing channel: door-drop flyers within 5 blocks, LinkedIn ads targeting nearby zip codes by employer, and a launch event with the local BID. Operational watch-out: train staff to greet subscribers by name and never make them queue with walk-in guests, or churn spikes within 60 days.

Use Sevenrooms (or Resy CRM) if you are single-location full-service — the reservation data, guest profiles, visit history, and email automation are already in one platform and integration-free. Add Klaviyo on top once you have e-commerce gift cards, merchandise, or multi-channel revenue worth tracking, because Klaviyo's segmentation and automation engine is meaningfully more powerful (about $150-1,500/month depending on list size). Mailchimp is the cheapest at $13-350/month and fine for a basic newsletter but lacks the lifecycle automation that drives loyalty revenue. The single mistake: running both Sevenrooms email AND Mailchimp on the same list — guests get duplicate sends, you look amateur, and unsubscribe rates spike. Pick one as the system of record and use the other only as a downstream sender if you must.

Five flows cover 80% of lifecycle revenue: (1) post-first-visit thank-you with a return-soon offer, sent 24 hours after the meal; (2) win-back at day 90 of inactivity with a soft offer like a free dessert; (3) birthday email sent 7-10 days before with a redeemable in-restaurant offer (avoid same-day sends — guests are usually busy); (4) anniversary of first visit at the 12-month mark with a milestone gift; (5) reservation reminder at 24 hours with directions, parking info, and dietary confirmation. Each flow should be A/B tested for subject line and offer at 90-day intervals. Industry benchmark: lifecycle flows generate 25-40% of total email revenue with 5-10% of total send volume — they are the single highest-ROI tactic in restaurant CRM.

The sweet spot for a full-service NYC restaurant is 2-4 sends per month — one event/menu update, one mid-month soft promo, optional weekend special, and a monthly regulars-only segmented send. Above 6-8 sends per month, unsubscribe rates jump from 0.2-0.3% per send to 0.5-1% per send and inbox placement degrades. Fast-casual concepts can push to 6-10 sends/month because frequency-of-purchase justifies more touches. Always segment — sending the same email to your top 100 regulars and your never-came-back list is the fastest way to look amateur. Track unsubscribe rate per send and pause if it crosses 0.5% — that is your audience telling you to slow down. Delete-and-resend tactics (resending to non-openers with a new subject line 3-4 days later) work but should be capped at 1x per month.

Yes, but the TCPA (Telephone Consumer Protection Act) requires express written consent before any marketing SMS — capturing a phone number for a reservation is NOT consent for marketing texts, and a violation is $500-1,500 per text in statutory damages, frequently the basis for class actions. The 10DLC requirements that took effect mid-2023 also require you to register your brand and use case with The Campaign Registry through your SMS provider (Twilio, Klaviyo SMS, Attentive) at $50-100 setup plus $2-15/month per campaign. Best practice: add a clear opt-in checkbox (sample wording: I want to receive promotional SMS, msg and data rates may apply, reply STOP to opt out) at reservation, loyalty signup, or POS — and store consent timestamp/IP for 4+ years. Open rates on SMS run 90%+ vs email's 25-35%, so it is worth doing right; doing it wrong is a $500K-multimillion-dollar lawsuit waiting.

The four deliverability fundamentals that matter in 2026: SPF, DKIM, and DMARC records on your sending domain (Google and Yahoo's Feb 2024 sender requirements made these mandatory for any sender doing 5K+/day to Gmail), a clean list with sub-2% bounce rate, a balanced text-to-image ratio (avoid all-image emails which get filtered), and engagement signals (opens, clicks) above 15%. Set up these DNS records once with your IT vendor or marketing platform; most operators skip this and wonder why open rates are 12% instead of 28%. Klaviyo and Sevenrooms both walk you through the DNS setup; Mailchimp does it semi-automatically with a custom domain. Also clean your list every 6 months — purge anyone with zero opens in 12 months because they tank your sender reputation and reduce inbox placement for your engaged guests.

Use a unique offer code per send and reconcile redemptions in your POS report monthly — this is the only attribution model that holds up at a restaurant where most bookings come through Resy/Sevenrooms and not a click-tracked link. For Sevenrooms users, the platform attributes reservations back to email opens within 7 days (a soft attribution model — directional, not perfect). Klaviyo will show you click-to-purchase for any e-commerce or gift card revenue, which is exact. Realistic benchmark: a healthy restaurant email program generates 5-15% of total revenue in attributed bookings + redemptions; under 3% means the program is underperforming. Track three numbers monthly: revenue per email sent ($0.05-0.30 is the range), revenue per active subscriber per month ($1-5), and email-attributed cover share of total covers (5-15%).

Five tactics in order of yield: (1) reservation autocapture via Sevenrooms or Resy (70-85% of cover names captured automatically with one consent checkbox), (2) host-stand iPad signup with a 'free dessert on next visit' offer (3-5% of walk-ins convert), (3) Wi-Fi captive portal that requires email for free Wi-Fi (works at high-dwell concepts like coffee, brunch — 40-60% capture), (4) gift card recipient emails (you already have them, just opt them in with consent at checkout), and (5) Instagram bio link to a giveaway landing page with email capture (slow drip — 10-50/month). Avoid paid email-list buys entirely — they are spam-trap-heavy and will torch your sender reputation in a single send. A NYC restaurant doing 80 covers/night should reach 5,000 emails in 6 months and 10,000+ in 18 months purely from organic capture.

Keep it under 30 active tags grouped into four families: dietary (vegan, GF, nut allergy, kosher, halal — 6-8 tags), occasion (anniversary, birthday, business, first date — 4-6 tags), preference (window seat, quiet table, sommelier service, fast service — 5-8 tags), and status (VIP, regular, press, owner-friend, do-not-comp — 4-6 tags). More than 30 and your hosts stop using them, fewer than 15 and you lose actionable detail. Sevenrooms and Resy both let you build these as auto-tags (e.g. third visit triggers Regular tag) plus manual tags from server notes. The single most valuable tag for revenue is Anniversary-of-First-Visit because it triggers a one-year milestone email that converts at 15-25% open and 5-8% redemption — far above lifecycle averages. Audit and prune the tag list quarterly.

Capture allergies at the point of reservation (Resy and Sevenrooms both have a notes field) and create dedicated allergy tags that flag the host stand and kitchen the moment the guest checks in. The four high-risk allergens to track separately because they trigger anaphylaxis — peanut, tree nut, shellfish, sesame — should be hard tags with a manager-acknowledgment workflow before the cover is seated. Train every server to verbally confirm the allergy at the table (do you still have that peanut allergy noted from your last visit?) because guests often forget they noted it. NYC has no specific allergen-disclosure law for restaurants beyond the FDA Food Code which NYC adopted, but in a serious incident your CRM record showing the noted allergy plus the kitchen confirmation is your single best defense against a $250K-2M negligence claim. Review allergy tags every visit; do not assume they carry forward unchanged.

The honest answer is that LTV tracking in restaurants is approximate, not exact, and you should plan around 60-80% attribution accuracy at best. Sevenrooms and Resy both ingest POS check totals and assign them to the reservation holder, which is correct for 1-2 person tables but breaks for groups of 6+ where one corporate card pays for the whole table. For party tracking, assign LTV to the reservation holder for parties under 5 and to the host plus a flag for split parties at 6+. Hotels with a PMS (Opera, Mews, Cloudbeds) can assign all in-house F&B charges to the room folio, which gives near-perfect LTV per guest. The number to chase is the trend, not the absolute — a guest whose 90-day LTV is rising 10% quarter-over-quarter is your real signal for VIP elevation, regardless of whether the absolute dollar amount is precisely captured.

The 60-second pre-shift drill is the single most underused tool in NYC hospitality: 30 minutes before service, the host pulls the reservation grid, opens each profile, and reads tags + last-visit notes for the next two hours of bookings. Flag VIPs to the GM, alert the kitchen on allergies, and pre-assign tables based on noted preferences (quiet, window, banquette). At the door, the host greets repeat guests by name (not Mr/Ms — first name signals warmth), references one detail from last visit (welcome back, was the lamb everything you remembered?), and walks them to a table that matches a noted preference. This three-touch recognition lifts repeat visit frequency 15-25% in operator data, and costs zero dollars beyond the discipline of the pre-shift huddle. If your CRM shows a guest has visited 3+ times and the server still asks 'is this your first time with us?' — your CRM is decoration, not a tool.

The four POS fields that should sync per check: total spend, item-level detail (what they ordered), discount/comp amount, and timestamp. Sevenrooms and Resy both pull these via direct integrations with Toast, Square, Aloha, and most major POS systems — the sync runs every 1-15 minutes depending on the connector. Audit it monthly: pull the POS daily sales report, sum by reservation, and reconcile against the CRM lifetime spend total — they should be within 5%. If they diverge by 15%+, your sync is broken and you need to escalate immediately because you are running blind on guest value. The most common failure modes are POS menu items not mapped to CRM categories, walk-in covers not assigned to a profile, and split checks landing in the wrong guest record. Set a monthly recurring task on the GM calendar to spot-check the reconciliation.

Yes — Sevenrooms, Resy OS, and most high-end CRMs support a guest photo field, and the legal posture in NY is that internal-use photos for hospitality service are permitted as long as the guest is informed (a brief notice at reservation that we may keep a profile photo to recognize you on return) and you do not share or publish the photo. The biometric privacy traps are NY's SHIELD Act (data security obligations) and the NYC POPPA (Persona Photos and Privacy Act considerations) which apply mostly to public-facing surveillance, not concierge profiles. Practical approach: only photo-tag opted-in VIPs at the $5K+ annual spend tier; never photo-tag without consent; and grant access only to GM, host lead, and senior managers — not the full server roster. Steakhouses and member clubs in NYC have run photo-recognition for decades; the difference today is documenting consent. Avoid third-party facial-recognition tools — that crosses into Illinois BIPA-style risk if you have any IL-resident guests and is reputationally toxic.

Six segments cover 90% of revenue: (1) Top 10% by 90-day spend — your VIPs, get private invites and zero discounts; (2) Visited 1x in last 90 days, no return — the win-back segment, send a soft return offer; (3) New guest in last 30 days — the welcome flow and second-visit nudge; (4) Birthday this month — the birthday offer; (5) Lapsed 6+ months — a re-engagement segment that gets one bigger offer then is suppressed if no response; (6) Never opened an email in 12 months — suppress entirely from active sends to protect deliverability. Each segment should get a different message and offer. The one segment most operators miss is the Once-A-Year Anniversary cohort — guests who came once last March; segment them and send a we-miss-you note in the next March cycle.

For non-subscription restaurants, churn means a previously-active guest has not visited in N days where N is 3-4x your typical repeat-visit interval. For a fast-casual concept where regulars visit weekly, churn at 30-45 days of no-show. For a full-service where regulars visit monthly, churn at 90-120 days. For a tasting-menu spot where regulars visit quarterly, churn at 270-365 days. The wrong move is using a single 90-day window across all concepts because it under-counts churn at high-frequency concepts and over-counts at occasion-driven ones. Calculate it monthly: take all guests who visited in the trailing 6-month window before your churn cutoff, then check how many have visited since — the gap is your churned cohort, and your retention rate is the inverse. Track this number monthly because it is the single best leading indicator of revenue decline 60-90 days out.

Start with a simple 12-month cohort table: for each month a guest first visited, track what percentage came back in months 1, 3, 6, and 12. A healthy NYC full-service restaurant shows roughly 30-40% return in month 1, 50-60% by month 6, and 65-75% by month 12 — the difference between the 30-day and 12-month numbers is your slow-burn return rate, and it is where most retention investment pays back. Compare cohort acquisition channels: walk-ins typically retain 10-20 points lower than reservation-driven covers, and OpenTable Points-driven covers retain 15-25 points lower than Resy or direct bookings. The action: if your walk-in retention is bad, your door experience is the problem; if Resy retention is bad, your menu or service is the problem. Sevenrooms and Resy both offer cohort reports built-in — use them quarterly, not just annually.

Win-back emails to lapsed guests (90-180 days since last visit) typically convert at 2-5% to a return visit within 30 days of send — far below an active-guest send (8-15%) but at near-zero marginal cost. The right offer is concrete and time-bound: a free starter or dessert with any entrée, redeemable in the next 21 days, beats a percentage discount which feels generic. Avoid blanket 20% off — it cheapens the brand and trains your guests to wait for win-back offers to come back. Industry benchmark: a win-back send to a 2,000-person lapsed list at 3% conversion brings back 60 covers worth $4-6K in revenue at a roughly $200 send cost — 20:1 return. Run it quarterly, not monthly, and exclude anyone who has redeemed a win-back offer in the prior 9 months so you do not train the cycle.

Guests who do not return within 90 days of their first visit are 4-5x more likely to never return at all — the 90-day mark is the cliff in restaurant retention. Improve it with three plays: a thank-you email at 24 hours after first visit with a one-touch personal note from the GM (not a generic template), a 14-day reminder with a new menu item or seasonal teaser, and a 60-day nudge with a soft return offer if no second visit yet. Aim for 35-45% second-visit rate within 90 days at a full-service restaurant; under 25% means the food, the service, or the price-value relationship is not landing on first impression. Survey first-time guests at day 7 with a single NPS question — if scores trend below 7, you have a service problem to fix before more lifecycle spend makes sense.

Build an at-risk regulars report that flags any guest whose typical visit interval has slipped by 50%+ — for a monthly regular, that means no visit in 45-60 days; for a weekly regular, 14-21 days. Sevenrooms calls this Lapsed Regulars; Resy has a similar segment. The intervention that works best is a 1:1 outreach from the GM (not a marketing email, not a discount blast) — a short note saying we have not seen you in a while, would love to have you back this month, here is a complimentary glass of cava on your next visit. This personal touch wins back 25-40% of at-risk regulars within 30 days. The mistake is over-comping — a free $15 cocktail is generous; a free $80 dinner devalues your brand and trains the regular to expect comps. Run the at-risk report weekly and assign each name to a specific manager for outreach.

Always compare year-over-year, not month-over-month — NYC has predictable seasonal swings (August Hamptons exodus, January-February post-holiday slump, July 4 + Labor Day holes) that look like churn but are calendar effects. A guest who visited monthly Sept-June and disappears in July-August is on vacation, not churned. The right cohort comparison is the same 12-month window year-over-year (Aug 2024 vs Aug 2025) — if your cohort retention is down 15%+ YoY in the same calendar period, that is real churn. Build your seasonality baseline from at least 24 months of POS data. Concepts at risk of misreading seasonality as churn: Upper East Side fine dining (high Hamptons exposure), Wall Street lunch (summer-Friday emptiness), and any restaurant in a tourist-heavy zone (FiDi, Times Square) where weekend vs weekday patterns swing 50%+.

Yes — Net Promoter Score (NPS = % Promoters minus % Detractors on a 0-10 scale) is the single most useful single-number guest health metric, and a healthy NYC full-service restaurant runs 50-70, fast-casual 30-50, hotels 40-60. A score under 30 means you have a real service or food problem; over 70 means you are best-in-class (think Le Bernardin, Per Se, Eleven Madison Park territory). Send the survey via email 12-24 hours after the visit (Sevenrooms and Resy automate this) — never on the same day, never from a tablet at the table. Track NPS monthly and segment by visit type (first-time vs regular, dinner vs brunch, dine-in vs delivery) because the average can hide ugly subgroup scores. The actionable signal is the Detractor verbatim — read every single 0-6 response, because they tell you exactly what is broken.

Expect 8-15% response rate on a single-question NPS sent 12-24 hours post-visit; longer surveys (5+ questions) drop to 3-6%. Lift it three ways: keep the email subject line short and personal (How was your visit? — not Please rate your experience at...), put the survey scale inline in the email body so guests can click 8/9/10 without leaving inbox, and send from a real human address (gm@yourrestaurant.com) not a no-reply alias. Avoid offering a discount or comp for completing — it biases the score upward and trains guests to fish for offers. Mobile-only design is mandatory; 70%+ of restaurant survey responses come on phones. If response rate is below 5%, your email deliverability is the issue, not the survey design.

Personal GM outreach within 24 hours of the response — a phone call beats an email, an email beats nothing, and silence is the worst option. The script is simple: thank them for the feedback, ask one open-ended question (what could we have done better?), listen, do not defend, and offer a specific make-good (a return visit with a $50-100 credit on the GM, not a generic discount code). Industry data shows 50-70% of recovered detractors return within 90 days and many become promoters — the recovery is more powerful than the original good experience. Track every detractor and the recovery action in your CRM with a tag (Detractor-Recovered or Detractor-No-Response). Patterns matter: if 30% of detractor scores cite the same issue (slow service Friday night, undercooked steak, rude bartender), you have a systemic problem to fix, not a one-off recovery.

For 95% of NYC operators, the built-in Sevenrooms or Resy survey is sufficient because it auto-triggers from the reservation, lands in the guest profile, and feeds your existing reporting at zero extra cost. Step up to a dedicated tool like Medallia or Qualtrics ($1-5K/month enterprise) only if you are running 10+ locations and need cross-location benchmarking, multi-language survey support, or text-analytics on open-ended responses at scale. Delighted ($224-449/month) is the middle option for 3-10 location operators who want better verbatim analysis. The one feature worth paying extra for is verbatim text-analytics that auto-categorizes detractor comments — this turns 200 raw verbatim responses into a ranked list of top 5 issues you can actually act on. Below 3 locations or 200 surveys/month, the manual read is faster than any tool.

12-24 hours after the visit hits the inbox is the optimal window — fresh enough to remember the experience, late enough that the guest is not still at dinner or arguing with their date about the bill. Same-day sends (within 2 hours) feel pushy and have 30-40% lower response rates. Sends 48+ hours out start losing detail in the response (guests forget specifics). For brunch, send Monday morning; for Friday/Saturday dinner, send Sunday afternoon; for hotel F&B, send at checkout-plus-12-hours. Sevenrooms and Resy let you set the delay window in the survey automation. If the guest had a flagged issue (allergy, complaint, comp), suppress the auto-survey and trigger manual GM outreach instead — sending a generic survey to a guest who already complained is tone-deaf and torches your NPS.

Yes, and it is the highest-yield review-generation tactic — promoters (NPS 9-10) who get a one-click ask to share on Google convert at 8-15%, far above any other channel. The compliant design is to ask everyone for the survey, then conditionally show the review link only to 9-10 scorers; do not ask 0-8 scorers because that is gating and Yelp/Google have terms-of-service prohibitions on review-gating that can get your business penalized or de-listed. Sevenrooms and Resy both have a built-in toggle for this conditional flow. Direct the click to the Google review URL (use the Google Business Profile place ID short link) for highest yield because Google reviews matter most for local search ranking; route a smaller share to Yelp if you need volume there. A NYC restaurant with 5,000 surveys/month and a 60 NPS can generate 50-100 net new Google reviews/month from this funnel alone.

Run a quarterly verbatim review where the GM, chef, and FOH lead read all detractor comments and the bottom 20% of passive comments together — pattern-spot the top 3 recurring issues, assign one owner per issue, and set a 60-day fix deadline. Common patterns NYC operators surface: noise levels too high, server attentiveness inconsistent during peak, host stand greeting cold, bathroom condition inconsistent, dessert menu weak. Each of these has a tactical fix — acoustic panels at $5-15K, scheduled section walk-throughs, host-stand script update, hourly bathroom check log, dessert menu revamp. Tracking: send the same NPS survey 90 days after a fix and measure whether the relevant theme dropped from the verbatims. If you are reading 200 verbatims a month and not making any operational changes, you are wasting the survey program — better to not run it than to ignore it.

Google reviews matter most by a wide margin — they drive local search ranking, show up in Maps, and are the first signal a guest sees when searching your restaurant name. Resy reviews matter second for full-service restaurants because they appear in-app at the moment of booking and influence reservation conversion 10-20%. Yelp matters third and is declining in NYC influence (national Yelp traffic has dropped roughly 20-30% since 2019 as Google has eaten the local search layer) but still drives some neighborhood walk-in traffic. Tripadvisor matters mainly for hotels and tourist-zone restaurants in Times Square, FiDi, and the Theater District. OpenTable reviews matter only if you are on OpenTable for reservations. For a NYC neighborhood restaurant: Google + Resy is 80% of review value; Yelp + Tripadvisor combined are 20%.

Target 4.4+ stars with at least 200-300 Google reviews to be competitive in NYC neighborhood search; the top-tier spots (Carbone, Don Angie, Ci Siamo) carry 1,000-3,000+ reviews at 4.5+. Below 4.0 stars you are functionally invisible in Google Maps neighborhood ranking; below 3.5 stars you are losing 20-30% of would-be walk-ins to the search result alone. Velocity matters as much as count — Google's ranking algorithm rewards a steady drip of fresh reviews over a single old burst, so target 10-30 new reviews per month. The 4.4-4.6 sweet spot is more believable than 5.0 (which reads as fake to sophisticated NYC diners). Aim for a review-to-cover ratio of about 1 review per 50-100 covers — a 100-cover/night restaurant should generate 30-60 reviews/month with a working ask program.

Yes — respond to 100% of negative reviews (1-3 stars) within 24-48 hours and at least 30-50% of positive reviews to signal you are paying attention. The tone for negatives: thank them for the feedback, acknowledge specifically what went wrong (do not deflect or argue), apologize, and invite them to contact the GM directly via email — never debate publicly. Industry data shows responding to negatives lifts overall rating by 0.1-0.3 stars within 6 months because future readers see you as accountable. For positives, a short personal thank-you (we loved having you, can't wait to welcome you back for the spring menu) is plenty — avoid copy-paste templates which read as bot-generated. Assign review response to one person (GM or marketing manager) and put 30 minutes per day on the calendar — anything ad-hoc gets dropped during weekend service.

Yelp's algorithm filters reviews from accounts with little Yelp activity, single-review accounts, or geographic patterns it deems suspicious — roughly 20-30% of reviews land in the Not Recommended bucket and the algorithm is intentionally opaque. You cannot directly fix it; soliciting reviews from friends/family or one-time Yelpers makes the problem worse because those accounts get filtered. The compliant tactics: encourage already-active Yelp users (the Elite ones) to review, never offer incentives, and never ask anyone to write a review on Yelp specifically (Yelp's TOS Section 6 prohibits actively soliciting reviews). The harder truth is that Yelp's filtering is a structural feature you have to live with — focus your review-generation energy on Google where there is no filter and the activity bias is far weaker. Paying for Yelp Ads does not unlock filtered reviews despite what the sales rep implies.

Both Google and Yelp let you flag reviews for removal, but the success rate is honest: Google removes about 20-30% of flagged reviews (those that clearly violate policy — naming staff, profanity, conflict of interest, or off-topic), Yelp removes maybe 10-15%. Build your case in writing: cite the specific policy violation (Google review policy categories include conflict of interest, harassment, off-topic), provide context (we have no record of this reservation), and submit through the Google Business Profile appeal flow or Yelp's flag-this-review form. For a serial harasser or ex-employee defaming the business, escalate to a defamation attorney — NY has a one-year statute of limitations for defamation claims (CPLR §215), so do not sit on it. Do not respond emotionally in public — every response stays visible forever and a defensive response makes you look worse than the original review.

For 90% of single-location NYC restaurants, no — the built-in Sevenrooms or Resy survey-to-review funnel does the same job at no extra cost. Tools like Birdeye ($299-499/location/month), Podium ($289-549/location/month), and NiceJob ($75-200/month) are worth it at multi-location concepts (5+ NYC locations) where centralized review monitoring across Google, Yelp, Tripadvisor, and Facebook in one inbox saves 10-20 hours/week of staff time. The features that justify the spend: AI-suggested response drafts for negative reviews, cross-platform sentiment analytics, and unified inbox routing to specific managers. Avoid any tool that promises to write fake reviews or mass-message contacts — that is a fast path to platform suspension and FTC scrutiny. The FTC's 2024 fake review rule carries fines up to $51,744 per violation as of 2025 enforcement.

Resy in-app rating (the 5-point scale shown next to the restaurant in the app and on Resy.com) moves booking conversion 10-20% in operator-shared data — moving from a 4.4 to a 4.6 in Resy can generate 50-150 incremental cover bookings per month at a busy NYC restaurant. OpenTable's Diner Reviews work similarly but OpenTable has a smaller share of high-end NYC reservations than Resy (Resy was acquired by AmEx in 2019 for ~$70M and now dominates NYC's premium reservation share). The reviews that move the score most are the most recent 90 days because Resy weights recency heavily. Treat Resy in-app reviews with the same urgency as Google: respond to negatives within 48 hours, encourage ratings via post-visit email, and watch your trailing-90-day average like a hawk. Sevenrooms and Resy both surface the in-app rating in operator dashboards.

Send 7-10 days before the birthday with an offer redeemable in a 30-day window (15 days before to 15 days after the actual date) — same-day birthday emails miss because the guest is already booked, and offers requiring redemption on the actual birthday have low yield. The two formats that work: a complimentary dessert or glass of cava with any entrée (low cost, high perceived value, requires a paid visit) or a 20% off your check up to $30 max (capped to control downside). Avoid free entrées or unconditional comps — they get redeemed by birthday-only diners who never return. Industry redemption rate is 15-25% on a well-designed birthday email, and the average birthday party brings 2-4 covers, so a 5,000-person email with birthdays this month at 20% redemption generates 50-200 covers in the month. Capture birthdays at reservation, loyalty signup, and POS prompt — most operators have birthday data on under 30% of guests when it should be 60%+.

Two anniversary programs both pay back: wedding anniversary (capture date at reservation when booking 2-tops, romantic occasion tags) with a complimentary glass of champagne and a personalized card at the table, and first-visit anniversary (12 months from first booking date) with an email saying we cannot believe it has been a year, here is something for your return visit. Wedding anniversary recognition at the table is essentially zero cost and lifts NPS by 5-10 points from already-happy couples. First-visit anniversary emails convert at 15-25% to a return visit when paired with a small offer like a complimentary amuse or aperitif. Sevenrooms and Resy automate both — set up the workflows once and let them run. The mistake: sending anniversary emails to guests who have visited 5+ times in the past year (it reads as tone-deaf because they are already regulars) — segment them out.

Budget 2-4% of revenue for total comps including VIP, manager comps, and service-recovery comps; over 5% means the program is being abused or the operations are chronically under-delivering. Of that, allocate 30-50% to VIP-specific comps (sommelier amuse, comp dessert for a regular, complimentary aperitif for a 10-visit guest) and 50-70% to service-recovery and manager-discretion comps. Track every comp in the POS by reason code (VIP-Recognition, Service-Recovery, Manager-Discretion, Owner-Friend) and review the monthly comp report by server, by manager, and by guest profile. The single biggest abuse pattern: managers comping friends or themselves under VIP-Recognition codes — set a rule that any single comp over $50 requires a second-manager approval and a written reason in the POS note. Comp data should also flag any guest receiving 3+ comps in 90 days for senior review.

Use a combined criteria: lifetime spend over $2,500 OR visit count over 12 in trailing 12 months OR press/influencer/owner-tagged. Pure spend misses high-frequency low-spend regulars (the daily lunch crowd) who are equally valuable for room atmosphere and word-of-mouth. Pure frequency misses the corporate diner who books a $1,500 dinner monthly. The composite gives you 3-7% of your guest base as VIP, which is the right cohort size to give meaningful recognition without diluting the tier. Sevenrooms and Resy both let you set composite VIP rules. Re-calculate quarterly because some VIPs lapse and some new regulars graduate in. Concrete VIP perks should be 3-5 named items: priority reservation 14 days out, named-guest greeting at the door, sommelier amuse on dinner visits, birthday recognition at the table, and one annual GM-thank-you call or note.

Write a one-page press/influencer policy and stick to it: comp anonymous critics from major outlets (NYT, Eater, Resy editorial, Infatuation) automatically and without comment; comp named press/influencers based on a signed media-host agreement spelling out coverage expectations; politely decline DM-based 'I'd love to come in for content' asks unless the audience is 50K+ engaged followers in your demographic. The FTC requires influencers to disclose comped meals (16 CFR Part 255 Endorsement Guides), and you should state in writing that disclosure is required as a condition of the comp. NYC operators report that maybe 1 in 20 'influencer' comp requests results in actual coverage that drives traffic; the other 19 are content for the influencer with no measurable benefit to the restaurant. Track every comped meal with an attribution flag and review quarterly to see what coverage actually showed up — kill the channel if ROI is invisible.

Build at least 6-8 comp reason codes, not just one generic Comp button: VIP-Recognition, Service-Recovery (food error), Service-Recovery (service error), Manager-Discretion, Owner-Friend, Press-Influencer, Birthday-Anniversary-Promo, and Loyalty-Reward. Toast, Square, Aloha, and Toast all support custom comp reason codes — set them up in the POS once and require servers to pick a reason on every comp. The monthly report should slice comps by reason, by server, by manager, by check, and by guest profile. The single most important variance to watch is the ratio of Service-Recovery to total covers — if it is over 2-3% of covers, you have a service or food problem, not a comp problem. Audit any single comp over $100 and any comp where the same server appears more than 5x/month as the comping authority.

Outstanding gift card balances are a real liability on your books — you owe goods to the cardholder until the card is redeemed or breakage is recognized. A single-location NYC restaurant doing $200K/year in gift card sales typically carries $40-80K in outstanding liability at any given moment. The risks: a sudden surge of redemptions (post-holiday, after a viral moment) can hit COGS and labor without the matching new revenue; if you sell or close the business, gift cards generally must be honored by the new owner per NY GBL §396-i and the asset purchase agreement. Reconcile gift card liability monthly against the gift card platform report; recognize breakage proportionally per ASC 606; and stress-test cash flow against a scenario where 30% of outstanding balances redeem in a single month. If you are heading toward a sale, refinancing, or shutdown, get advice from a restaurant CPA on gift card treatment 6+ months ahead.

Loyalty fatigue is when your members stop engaging because the program feels too noisy, too slow to reward, or too cluttered with offers — symptoms include redemption rate dropping below 10%, email open rate among members under 20%, and net member count flat or declining despite ongoing enrollment. The cure is to do less, better: trim from 6-8 monthly emails to 2-3, simplify reward structure (one clear way to earn, one clear reward at one threshold), and stop sending undifferentiated promos to your top 20% of members who do not need a discount to come back. Audit your program annually: ask 50 active members in a 1-question email what they would change about the program. If the top answer is 'fewer emails' or 'simpler rewards,' you are running a fatigued program. The other tell is when your top spenders unenroll quietly — they got enough generic offers to feel insulted.

Run a monthly comp-by-server report and flag any server whose comp rate is more than 1.5x the restaurant average — at a healthy operation, individual servers should run within 50% of the mean. Patterns to investigate: same guest receiving comps from the same server repeatedly (potential kickback or friend), comps applied at end-of-shift without manager re-approval, and comps that match exactly to a non-cash tip pattern (server comps an item, guest tips bigger). Toast and Square both have comp anomaly reports that surface these. The legal angle: under NY Labor Law §198 and §195, wage-theft investigations can subpoena POS comp data going back 6 years. Document every comp policy in writing, train managers annually, and fire on first verified abuse — comp abuse rarely happens in isolation, and tolerated abuse signals to staff that other rules are negotiable too.

GDPR applies if you market to or collect data from EU residents (your tourist guest list almost certainly qualifies), CCPA/CPRA applies if you have $25M+ revenue or process data on 100K+ California residents, and NY SHIELD Act (effective March 2020) applies to ANY NY business with NY-resident data — it requires reasonable data security, breach notification within 60 days, and an incident response plan. The practical compliance moves: post a privacy policy on your website, add a cookie banner if you run any tracking pixels, give guests a simple unsubscribe and data-deletion request path (privacy@yourrestaurant.com), and require your CRM and email vendors to sign data processing agreements (Sevenrooms, Resy, Klaviyo all offer them). NY also passed the NY Privacy Act framework which is still moving but trending toward CCPA-style requirements by 2026-27. Budget $2-5K to have a privacy attorney write the foundational policies once; do not copy a template from another restaurant.

NY SHIELD Act requires notification to affected NY residents and the NYS AG within a reasonable time (effectively 60 days from discovery is the operative standard), and if the breach affects 5,000+ NY residents you also notify the major NY credit reporting agencies. Card data breaches additionally trigger PCI-DSS notification to your card processor and possibly Visa/Mastercard fines of $5K-100K+ depending on volume. Have an incident response plan in writing before you need it: who calls the cyber attorney, who calls the platform vendor, who drafts the guest-facing notification, who handles press. The platform vendor (Sevenrooms, Toast, Klaviyo) carries primary technical responsibility but you carry primary notification responsibility because the data was collected in your name. Cyber liability insurance ($1-3K/year for a small restaurant policy) covers notification costs and forensic investigation — most operators do not have it and find out the hard way that the bill is $50-200K.

Give 60-90 days written notice via email and in-store signage, allow members to redeem all earned points/credits during the wind-down window, and convert any unredeemed balance into either a cash-equivalent gift card (cleanest) or a transition credit on a successor program (acceptable). Never let earned points expire silently or be voided in a TOS update — that is the fastest way to generate viral 'this restaurant stole my points' social posts and Yelp/Google retaliation. Legally, points are generally not regulated as gift cards (NY GBL §396-i exempts loyalty rewards explicitly) so you have flexibility, but reputationally the cost of shutting badly is 10x the cost of an orderly wind-down. Communicate the why (we are switching to a simpler program / our vendor is being acquired / we are streamlining) honestly. Most operators who shut down a loyalty program quietly lose 10-20% of their email list to unsubscribes within 30 days; an honest, generous wind-down loses 2-5%.

The realistic answer is you cannot perfectly unify it without engineering, but you can get 80% of the value with a $200-1,000/month customer data platform (CDP) like Bikky (built specifically for restaurants), Wisely (Olo-acquired in 2021 for $187M, hospitality-focused), or a lightweight Zapier-Airtable build that pipes the four key fields (email, last visit, lifetime spend, tags) into one master record. Bikky and Wisely both start around $300-800/month per location and integrate with the major POS, reservation, and email platforms out of the box. The full enterprise CDP path (Segment, mParticle, Twilio Segment) is overkill at single-location and runs $2-10K/month. Before spending: write down the three reports you actually need from unified data — if they are 'who is my top 100 spenders, who is at-risk of churn, and what is my email-attributable revenue' — then make sure the tool you pick can answer those three questions on day one, not on a future roadmap.