A restaurant customer database is one list of every guest you can identify, with their contact details, consent status, visit history and spend, built by merging what your POS, reservation book, online ordering, guest Wi-Fi and wallet passes already collect. You can build a working one this week with a spreadsheet: the 7 steps, the 15 fields to capture, and a free copyable template are below.
The reason to bother is simple. In Toast's Regulars Report 2026 (Toast with Resy data, September 2026), 57% of orders on the Toast platform were anonymous, while just 7% of a restaurant's guest base, its multi-visit regulars, could drive up to 50% of order volume. If you cannot name those guests, you cannot thank them, notice when they stop coming, or invite them back.
What is a restaurant customer database?
It is a single table where each row is a guest and each column is something you know about them: who they are, how to reach them, whether they said yes to hearing from you, and how they behave (when they came first, when they came last, how often, what they spend, what they order).
Three things separate a useful database from a pile of exports:
- One row per person. The same guest shows up in your POS, your reservation book and your online ordering. Merge them on mobile number or email so visit counts are real.
- Consent lives next to contact. A phone number without a recorded text opt-in is a number you should not market to. Keep the yes, the date and where it came from.
- Last visit date is current. It is the field that tells you who is drifting. If it updates once a quarter, you find out too late.
How to create your restaurant customer database in 7 steps
Here is the order that gets you a usable list fastest, using tools you already pay for.
- Turn on the guest profiles inside your POS. Square, for example, creates an Instant Profile automatically when a guest's name is collected from their payment card and no matching profile exists, and adds later purchases on that card to the same profile (Square Support, checked November 2026). Toast's Guest Insights shows new, returning and lapsed guests, retention rate, time to return and top guests by orders and spend (Toast Central, checked November 2026). Start here, because it costs the guest nothing.
- Export your reservation guestbook. On OpenTable, the path is Reporting, then Guest Export, then download all guest profiles; OpenTable emails you the CSV, and the export carries name, mobile and email, with notes, tags and visit history recommended (Resy Helpdesk data transfer instructions, checked November 2026).
- Export your online ordering customers. First-party online ordering gives you name, phone, email and order history on every order. Pull it monthly.
- Merge on mobile number, then email. Normalize phone numbers to one format first (10 digits, no punctuation). Where two rows share a number, keep one and add the visit counts together.
- Add the consent columns. One for text, one for email, each with a date and a source ("Wi-Fi splash page, 2026-11-02"). If you cannot show how someone opted in to texts, mark them No.
- Add the three working fields. Last visit date, visits in the last 90 days, and average check. These drive almost every decision in the next section.
- Set a refresh day. Re-export and update weekly. A database that is a month stale flags lapsed regulars a month late.
One trap to know about: Square's help center states that a customer is automatically opted into email marketing if you manually type their email into their profile (Square Support, checked November 2026). Only type in emails from guests who asked to hear from you.
Restaurant customer data collection: the 5 capture points
Each source captures a different slice of your guests. The strongest databases combine at least three.
| Capture point | What you get | Who it misses | Effort for the guest |
|---|---|---|---|
| POS card matching | Name on card, visit dates, spend, items, daypart | Cash payers, guests who split cards, anyone with no contact details | None |
| Reservations | Name, mobile, email, party size, notes, visit history | Walk-ins, counter service, takeout | Already part of booking |
| Online ordering (first party) | Name, phone, email, full order history | Dine-in only guests, guests ordering through third-party apps | Already part of ordering |
| Guest Wi-Fi login | Email or phone, visit timestamp on each reconnect | Guests who never connect | One form on the splash page |
| Wallet pass (Apple or Google Wallet) | Phone or email at sign-up, plus every scan as a visit | Guests who decline the pass | One tap, no app download |
Every source has a blind spot. Merge them and the blind spots shrink.
POS transaction data
Your POS is the backbone because it records every visit with no effort from the guest. Its weakness is contact details: a card gives you a name and a history, not a way to reach the person. Treat POS profiles as the visit log and use the other capture points to attach a phone number or email.
Reservations
Reservation guests hand you clean data because they want the table. The Resy side of Toast's 2026 report found that regulars booked 83% of their visits in advance, while 52% of non-regular visits were walk-ins (Toast Regulars Report 2026, via Stacker). Your reservation book over-represents your best guests, which is exactly why it belongs in the database.
Wi-Fi
A Wi-Fi splash page that asks for an email or phone number turns every reconnect into a recorded visit. Keep the marketing opt-in as its own unchecked box rather than bundled into the login, so the consent column stays honest.
Online ordering
First-party ordering is the richest source per guest, because every order is tied to a name, a phone and an item list. It also proves the list pays: in Toast's 2026 data, 86% of orders driven by email marketing and 83% of orders driven by SMS came from existing diners, at restaurants using those tools (Toast Regulars Report 2026, via Stacker).
Wallet passes
A wallet pass is a card that lives in Apple Wallet or Google Wallet. The guest adds it with a tap at the table or counter, without downloading an app, and every scan logs a visit against their row. The pass can also carry messages: Google's Wallet developer documentation lets a business add a message that triggers a push notification to everyone holding the pass, capped at 3 notifying messages in a 24-hour period (Google Wallet developer docs, checked November 2026). That cap is a feature: it keeps wallet push notifications rare enough to stay welcome.
Restaurant guest database: the fields to capture
You do not need 40 columns. These 15 cover the decisions you will actually make.
| Field | Why it matters | Where it comes from |
|---|---|---|
| Guest ID | Lets you merge sources without duplicates | You assign it |
| First and last name | Personal greeting, staff recognition | POS card, reservation, ordering |
| Mobile number | Text messages, merge key | Reservation, ordering, Wi-Fi, wallet pass |
| Email campaigns, second merge key | Reservation, ordering, Wi-Fi | |
| Text opt-in (Y/N, date, source) | Required before marketing texts | Sign-up form, wallet pass, Wi-Fi |
| Email opt-in (Y/N, date, source) | Clean list, fewer complaints | Sign-up form, ordering checkout |
| Birthday (month and day) | Birthday offer | Sign-up form |
| First visit date | New versus returning | POS |
| Last visit date | Lapse detection, the most important field | POS, wallet pass scans |
| Total visits | Regular versus occasional | POS |
| Visits in last 90 days | Current habit, not old history | POS |
| Average check | Value of the guest | POS |
| Favorite item | Offer the thing they already love | POS item history |
| Usual daypart and day | Time the message to their habit | POS timestamps |
| Channel (dine-in, takeout, delivery, reservation) | Pick the right offer | POS, ordering, reservations |
Skip the birth year. Month and day is all a birthday offer needs, and collecting less is safer. Square's own guidance is to never store sensitive personal data in the notes field (Square Support), and that is a good rule for any system.
Free restaurant customer database template (copy this)
Copy the header row into the first line of a Google Sheet or Excel file. The example row shows the format.
guest_id,first_name,last_name,mobile,email,sms_opt_in,sms_opt_in_date,sms_opt_in_source,email_opt_in,email_opt_in_date,birthday_mm_dd,first_visit,last_visit,total_visits,visits_last_90d,avg_check,favorite_item,usual_daypart,channel,segment
G0001,Maria,Lopez,3035550142,maria@example.com,Y,2026-09-14,wallet pass,Y,2026-09-14,04-22,2025-11-03,2026-10-28,19,6,38.50,carnitas tacos,Fri dinner,dine-in,RegularThen add three formula columns after the last header (columns U, V and W) and fill the segment column with a formula, so the sheet does the sorting for you:
| Column | What it calculates | Example formula (Google Sheets) |
|---|---|---|
| T: segment | New, Developing or Regular | =IF(N2=1,"New",IF(N2<=5,"Developing","Regular")) |
| U: days_since_last_visit | How long since they came in | =TODAY()-M2 |
| V: normal_gap_days | Their usual time between visits | =(M2-L2)/MAX(N2-1,1) |
| W: overdue | Gap is more than 2x their normal | =IF(N2<2,"",IF(U2>2*V2,"Overdue","OK")) |
Column letters assume the header order above. Adjust them if you reorder.
How to collect restaurant guest data without a loyalty program
You do not need points to get a phone number. Guests share contact details when the exchange is clear and immediate. Four ways that work without a points program:
- Recognition. In Toast's April 2026 survey of 1,500 U.S. adults, 48% said being remembered by staff makes them feel most valued, against 22% who prioritize points-based rewards, and 32% named feeling recognized as a main reason they return (Toast Regulars Report 2026, via Stacker). "Can I get your number so we remember your usual?" is an honest ask.
- A free item on the next visit. A QR code on the table that offers a free dessert or a free side next time, in exchange for a phone number and a text opt-in. It gives a reason to come back and a record when they do.
- Digital receipts. Offering a text or emailed receipt attaches contact details to the card already on file in your POS.
- A wallet pass instead of a punch card. It holds the free item, logs each visit, and replaces the paper card guests lose.
Simple beats clever here. The same Toast survey found over half of diners have left a loyalty program because rewards were too hard to earn or not worthwhile (Toast Regulars Report 2026, via Stacker). A single free item is easy to understand and easy to redeem.
How to use restaurant customer data: segments, win-back, birthdays
Data earns its keep when it changes what you do on Monday. Start with three segments and three plays.
| Segment | Rule | Play | Example free-item offer |
|---|---|---|---|
| New | 1 visit | Second-visit nudge within 7 to 14 days | "Come back this week and your fries are on us" |
| Developing | 2 to 5 visits | Build the habit around their daypart | "Your Friday coffee is free after 3 more visits" |
| Regular | 6+ visits | Recognize, protect, and catch drift early | "We saved you a free slice of the new pie" |
| Overdue | Gap over 2x their normal | Win-back | "We miss you. Your favorite appetizer is free this week" |
| Birthday | Birthday this month | Birthday offer | "Happy birthday. Dessert is on us any night this month" |
Second visit first. The first return is the hardest one to earn. In Bloom Intelligence's 2025 guest retention analysis, a vendor study of its own restaurant clients from January 2024 to October 2025, 25% of first-time visitors returned within 90 days (Bloom Intelligence, 2025). Treat that as a directional number, since the study does not publish its sample size.
Win-back by personal rhythm, not the calendar. A guest who visits weekly is overdue after two weeks. A guest who visits monthly is not. The overdue formula in the template uses each guest's own gap, which is the difference between a timely nudge and spam.
Offer the item, not a percentage. A free appetizer has a known cost to you and a clear value to the guest. A percentage off resets what guests expect to pay. Use the favorite item column: the free item that lands best is the one they already order.
Measure it. Square's 2026 Local Economy Report defines a regular as someone who visits at least four times a year, and found regulars generate six times more annual revenue than one-time visitors (Square via eCommerceNews, March 2026). Count how many guests cross that line each month. For the benchmarks to compare against, see our guide to restaurant customer retention rate.
Where restaurant customer data already lives
Most restaurants already own more customer data than they think, spread across systems that do not talk to each other:
- POS: visit history, spend and items for card payers.
- Reservation platform: contact details and notes for your planned diners.
- Online ordering: full contact and order history for takeout guests.
- Gift card sales: buyer names and emails, often never exported.
- Catering and private events: organizers with big checks and contact details in someone's inbox.
- Third-party delivery: usually the thinnest, because the platform holds the relationship.
The work is not collecting more. It is pulling what you have into one list and keeping it current.
Restaurant CRM database: when a spreadsheet stops working
A spreadsheet works until about the point where weekly exports and merges eat more than an hour, or you have several hundred active guests and the overdue list changes daily. That is when software that reads your POS directly and keeps the database current earns its fee.
What to look for: a direct POS connection, automatic merging of duplicates, consent tracking per channel, and segments that update themselves. We compare the options in restaurant CRM software for 2026, and if what you mainly want is a reason for guests to sign up, the best restaurant loyalty software comparison covers points, wallet passes and apps.
Regulr is one option for restaurants on Square: it gives guests an Apple or Google Wallet pass with no app to download and runs win-back and birthday campaigns off each guest's own visit rhythm, paying out in free items rather than discounts. It is a fit if you are on Square; the CRM comparison above covers the rest.
Privacy and consent basics: TCPA, CAN-SPAM and state privacy laws
This is general information from the primary sources, not legal advice. When in doubt, ask a qualified professional.
| Rule | What it covers | What you need to do |
|---|---|---|
| TCPA (FCC rules, 47 CFR 64.1200) | Marketing calls and texts sent with an autodialer or prerecorded voice | Get prior express written consent, never make it a condition of purchase, honor opt-outs within 10 business days, no solicitation calls before 8 a.m. or after 9 p.m. local time |
| CAN-SPAM (FTC) | Commercial email | Accurate headers and subject lines, a physical postal address, a clear opt-out honored within 10 business days |
| CCPA (California) | Businesses over its thresholds | Privacy notices and consumer rights, if you meet a threshold |
| Other state privacy laws | Varies by state | Check your state's thresholds, some are low |
Texts (TCPA). The FCC's rule requires the prior express written consent of the recipient for advertising or telemarketing calls made with an autodialer or prerecorded voice, states the person cannot be required to agree as a condition of purchasing goods or services, and accepts an electronic signature where valid under federal or state law. Guests can revoke consent by any reasonable method, and you must honor it within a reasonable time not to exceed ten business days (47 CFR 64.1200, Cornell LII). In practice: a sign-up form with clear language about marketing texts, an unchecked box, and a record of the date and source.
Email (CAN-SPAM). CAN-SPAM does not require opt-in. It requires no misleading headers or subject lines, disclosure that the message is an ad, a valid physical postal address, a clear opt-out that keeps working for 30 days after sending, and opt-outs honored within 10 business days. Each violating email can draw penalties up to $53,088 (FTC CAN-SPAM compliance guide).
State privacy laws. California's CCPA applies to a for-profit business that does business in California and has gross annual revenue over $25 million, buys, sells or shares the personal information of 100,000 or more California residents or households, or earns 50% or more of annual revenue from selling personal information (California Attorney General). Most independent restaurants fall under those lines, but not every state's law is that high: MultiState counted twenty states with comprehensive privacy laws on the books as of February 2026, and Rhode Island's applies to entities processing data of at least 35,000 consumers (MultiState, 2026). Multi-location groups should check every state they operate in.
Three habits keep you safe in any state: collect only the fields in the table above, record consent per channel, and make opting out one step.
Your first week
Turn on your POS guest profiles, export your reservations, merge on phone number, add the consent columns and paste in the template formulas. By the end of the week you will have a short list of regulars who have not been in lately, and a free item ready to invite them back. That list is the start of every retention number worth tracking, and our restaurant customer retention rate guide shows how to measure the result.
Free: Customer Retention Checklist
A printable checklist with the strategies from this article, plus message templates you can copy-paste today.
No spam. Unsubscribe anytime. Your email stays private.
Get weekly retention tips
One actionable idea every Tuesday. No fluff, no spam.
Join 2,400+ local business owners. We respect your inbox.
Founder of Regulr & City Curated
Regulr is the customer retention layer for local businesses. It plugs into your POS, learns every customer's behavior, and runs personalized retention campaigns automatically — SMS, email, wallet pass updates, and RCS sentiment routing. Built for restaurants, coffee shops, salons, med spas, fitness studios, and other independent local businesses where every customer is a name and every visit matters.
