Restaurant Customer Retention Statistics for 2027: 50+ Numbers

Last verified: October 12, 2026. Every number links to its source and is labeled as transaction data, a survey, or general business research.

Dutch Bros put a number in its annual report this year that most restaurant owners would love to know about their own business. In 2025, "approximately 72% of all transactions were attributable to Dutch Rewards members, up from approximately 68% in 2024," the coffee chain wrote in its Form 10-K, filed February 2026. By the first half of 2026 the share was 74% (Dutch Bros Q2 2026 results, August 2026). Nearly three of every four transactions now come from a guest the company can recognize, count, and invite back.

The big chains are all chasing the same thing. Starbucks says its Rewards members drove "nearly 60% of U.S. company-operated revenue in fiscal 2025" (Starbucks, January 2026). McDonald's told analysts that its average US customer visited 10.5 times in the 12 months before joining its loyalty program and 26 times in the 12 months after (McDonald's Q4 2025 earnings call, February 2026), though people tend to join when they're already visiting more, so that jump flatters the program.

Now picture a 60-seat taqueria (an illustrative example, not a real restaurant). The owner matches 90 days of online orders and loyalty check-ins by phone number and finds 1,240 unique guests, 372 of whom came at least twice: a 30% repeat guest rate. Is that good? It depends on whose 100 guests you compare it to, which is what this page sorts out, ending with the math to calculate your own.

The key numbers

Most rows come from companies that sell to restaurants, and each counts a different crowd over a different window, so check the method before you quote one. This chart sorts the headline numbers by how they were collected.

What percentage of restaurant customers come back?

Most first-time restaurant guests never come back. In 2026 data from Toast, Bloom Intelligence, and Upside, somewhere between 7 and 36 of every 100 first-timers return, and even among loyalty members, fewer than half make a second visit (Paytronix).

The spread is wide because each data set counts a different crowd over a different window:

• Toast (which sells POS and loyalty software) puts the baseline at 7%, "the share of first-time guests who return without any intervention," in a study of January to April 2026; loyalty enrollment moved that "from a 7% baseline to nearly 30%" (Toast, July 2026), though guests who enroll are already the keen ones.
• Bloom Intelligence (which sells restaurant marketing software) followed "nearly a million first-time guests across hundreds of restaurants" captured from January 2024 to early March 2026: "about 85% of first-time guests had exactly one recorded visit," so roughly 15 of every 100 came back, which Bloom rounds to about 1 in 7 (Bloom Intelligence, June 2026).
• Upside (a cash-back offers marketplace) found in card data from 1,498 restaurants, March 2022 to February 2025, that "64% of new customers shopping in their first month won't be back that year" (Upside, July 2026), so about 36 of 100 came back.
• DataDelivers (a restaurant data platform) found in card data on "more than 50 million guests" that "only 17% of first-time guests returned to that restaurant in the next 12 months" (RestaurantNews.com, March 2023).
• Paytronix (which sells loyalty software) reports a return rate of "less than 50%" after a member's first visit, rising to 95% after the fourth, from data on 800+ brands (Paytronix 2026 Loyalty Report). Its older visit-by-visit numbers for quick-service brands: second visits "only fifty-fifty," and third visits "only increase to 70%" (MediaPost, February 2017), which works out to 50 of every 100 first-time members making a second visit and 35 a third.
• Olo (which sells ordering and guest data software) says "about 65% of guests who visit a restaurant one year do not return the next," in sponsored content (Restaurant Business, March 2026).

Where "70% of first-time guests never return" comes from

The most repeated number in this field traces to a January 2015 Thanx analysis of 2014 data from 54 businesses in 16 industries, which found "more than 70% of retail and restaurant customers visit only once" within its data window (Thanx blog, archived). The press release headline turned that into "never make a return visit," later posts dropped the word "retail," and vendor pages still repeat it without a source. The measured 2026 figures above are better citations, and two of them, Toast's and Bloom's, are worse than 70%.

What it means for a one to three location independent: the second visit is the leak that matters. Bloom found guests who first ordered online came back at 19.9%, against 12.7% for walk-ins captured on Wi-Fi (same source), though part of that gap is how hard Wi-Fi guests are to recognize when they return.

How much is a repeat customer worth to a restaurant?

A regular is worth several times a one-time guest, mostly because regulars keep coming, not because each check is bigger. Square found regulars generate 6 times the annual revenue of one-time visitors, while PAR found loyalty members spent only about 2% more per visit than anonymous guests in 2025.

• Concentration: in Toast's Q1 2026 transaction data, "just 7% of a restaurant's guest base accounts for multiple visits, yet this group could drive up to 50% of total order volume" (Toast, July 2026). Olo says 60% of restaurant revenue comes from repeat guests, based on "more than 100 million guest records" (Olo, November 2024).
• Annual value: Square's data from 2019 to 2025 shows "regular customers generate 6X more annual revenue than one-time visitors nationally," counting regulars as four or more visit dates a year at any local business type (Square, March 2026). Paytronix cites "27X higher customer lifetime value for 10+ visit guests versus one-time visitors" (Paytronix, April 2026).
• Per visit: PAR's 2025 data from more than 30,000 quick-service restaurants shows loyalty members "averaging $15.08 per visit versus $14.82" for anonymous guests, and guests with more than 10 visits a year "drive 61% of check-ins and 53% of sales" (PAR Technology, April 2026). SevenRooms' data from October 2025 to March 2026 shows "repeat diners spend 27% more than first-time diners" (DoorDash and SevenRooms, 2026 Restaurant Industry Trends Report).

One number to stop repeating: "repeat customers spend 67% more." It comes from a December 1999 Bain and Mainspring survey of online shoppers, which found the average repeat apparel customer "spent 67 percent more" in months 31 through 36 of the relationship than in the first six months (Bain and Mainspring, 2000). It compares the same online shoppers over three years and says nothing about restaurants.

I grew up in my family's small food business in India: no brand, no storefront, just a local market and regulars who trusted us. I learned margins counting every rupee. Here's what I think it taught me: regulars aren't a marketing segment. They are the margin.

Does "5% more retention, 25% to 95% more profit" apply to restaurants?

No. The line comes from Bain and Company consulting work published in the 1990s on banks, insurance brokers, credit cards, and auto service; it measured the value of an individual customer rather than a company's profit, and restaurants were not in it.

The original 1990 article by Frederick Reichheld and W. Earl Sasser said "companies can boost profits by almost 100% by retaining just 5% more of their customers," and its chart was titled "Reducing Defections 5% Boosts Profits 25% to 85%," with examples of "85% more profits in one bank's branch system, 50% more in an insurance brokerage, and 30% more in an auto-service chain" (Harvard Business Review, 1990). The "5%" was a 5 percentage point cut in the yearly defection rate (a credit card example went from 10% to 5%), the profit was the net present value of one customer's profit stream, and no method was published. The only restaurant is an anecdote: a Domino's franchisee who figured a regular was worth "more than $5,000 over the life of a ten-year franchise contract."

The exact "25% to 95%" wording first appears in a 2000 follow-up, "E-Loyalty," which restated the earlier work (Harvard Business Review, 2000). The Bain brief that a widely shared 2014 HBR.org article cited for it actually says "in financial services, for example, a 5% increase in customer retention produces more than a 25% increase in profit," with no 95% (Bain, 2001; HBR.org, 2014). Marketing scientist Byron Sharp of the Ehrenberg-Bass Institute calls the original "merely a thought experiment" about customer profitability, not company profitability (Ehrenberg-Bass Institute).

The companion claim, that a new customer costs 5 to 25 times more than keeping one, is softer still. The same HBR.org piece hedged it, "depending on which study you believe, and what industry you're in," and the authors of "Loyalty Myths" trace the five-times version to late 1980s research, asking what the extra cost really was of "dining at a new restaurant" (Ipsos excerpt, Loyalty Myths). Retention matters for restaurants; your own repeat rate proves it better than a 1990 bank.

Do loyalty programs actually increase visits?

Loyalty members do visit more, but heavy users are the ones most likely to join. Circana's receipt data shows members make 22% more restaurant visits a year than nonmembers and visit just as many chains, about 20, while giving 8% of their visits to a brand they belong to, against 4% for nonmembers (Circana, June 2025).

The same Circana data says loyalty traffic doubled from 2019 to 2024, members now account for 39% of restaurant visits, and loyalty traffic grew 5% in the latest year "even as overall restaurant traffic declined by 2%." The best behavioral evidence of what a program adds:

• Toast: at restaurants with active programs, "members were retained at roughly 2x the rate of new customers," and cashback rewards drove 90-day retention of 24% to 26% at casual spots (Toast, July 2026).
• PAR: a guest's "first reward redemption increases visit frequency by up to 38% within 90 days," and in 2025 loyalty transactions grew 28.5% while anonymous transactions fell 6.7% (PAR Technology, April 2026).
• Paytronix: frequency concepts kept active rates of 66% to 72% in 2025, but at bar and grill concepts "nearly three in four new members" never returned within 90 days (Paytronix 2026 report release).
• A caution from Sweetgreen: the chain said its switch from Sweetpass+ to SG Rewards "created near-term headwinds" as fourth-quarter 2025 traffic and mix fell 13.3% (Sweetgreen Q4 2025 results).

Surveys say guests care, with caveats. In DoorDash and SevenRooms' March 2026 survey of 3,001 consumers, 66% said they order more often where they're active loyalty members (DoorDash and SevenRooms). In PAR's September 2025 survey, a third were using restaurant programs more because of economic pressure, and 25% would switch restaurants for better perks (PAR Technology, December 2025). Upside counts the average restaurant member in 4.2 programs (Upside, July 2026), and Toast's survey of 1,500 frequent diners found 48% valued being remembered by name or usual order, "more than double the 22% who prioritize a points-based reward" (Toast and Resy, July 2026).

For a small restaurant, that adds up to a simple program whose first reward arrives within a few visits, plus staff who remember people. The design choices are in the restaurant loyalty programs guide.

Does ordering direct change how often guests return?

In the one restaurant-specific comparison published, yes. Paytronix's 2023 data showed first-party ordering customers returning every 33 days, against 43.1 days for customers who ordered repeatedly through third-party channels, though loyal guests may simply prefer ordering direct (Restaurant Dive, June 2023).

The same data found 41% of first-party orders came from loyalty members against 3% of third-party orders, with tips of $4.44 against $1.89. Paytronix's 2024 report added that guests ordering directly ordered 35% more per check, and guests who order both in-store and online have "35% more lifetime value than customers who only order in-store" (Paytronix, April 2024). On the stated-preference side, NCR Voyix's November 2024 survey found 58% of consumers prefer ordering delivery or carry-out on the restaurant's own website or app, against 16% for a multi-restaurant app (NCR Voyix 2025 report).

The bigger difference is whether you can see the guest at all. Toast reports that "57% of orders on the Toast platform are anonymous" (Toast and Resy, July 2026), and Olo's CEO told Restaurant Dive that with marketplace delivery "the brand gets zero data about who that guest is" (Restaurant Dive, January 2025). What each app shares is laid out in does DoorDash share customer data with restaurants.

I've watched neighborhood restaurants, a family taqueria, a pho shop, a pizza place, hand a large share of their revenue to platforms that don't share the customer's email address back to them. Here's what I think: a repeat guest you can't contact is a repeat guest you're renting.

Which messages bring guests back?

Personal, triggered messages beat blasts. In Klaviyo's 2026 benchmarks for food and beverage brands, automated SMS flows led to orders from 2.03% of recipients against 0.35% for one-off campaigns, and restaurant data firm DataDelivers reports that brands using one-to-one marketing retained guests at 2.7 times the rate of those that didn't, 35% against 14%.

Read the benchmarks with care: Klaviyo's "food and beverage" brands are online food sellers, not restaurants, and their SMS flows also beat campaigns on clicks (9.5% against 5.19%) and revenue per recipient ($1.83 against $0.34) (Klaviyo SMS benchmarks, 2026). DataDelivers' 35% against 14% comes from its 2026 report on 77 million guests (RestaurantNews.com, March 2026). For email, Constant Contact's December 2022 data put dining and food services at a 38.03% open rate and a 0.68% click rate (Constant Contact, archived), but opens are inflated by Apple's Mail Privacy Protection, which Mailchimp warns affects open-rate accuracy (Mailchimp). Track clicks and orders instead.

And drop the "98% SMS open rate." It traces to a 2010 Mobilesquared white paper, and in 2022 Mobilesquared itself said that figure no longer reflects "today's world of mobile messaging," putting the share of texts actually read at 55% (Mobilesquared, August 2022).

In practice: let messages fire on behavior (a first order, a quiet month, a birthday) with restaurant marketing automation flows, write them from the restaurant email templates, and follow the restaurant text message marketing rules before any text goes out.

Do reviews and service recovery bring guests back?

Better ratings bring more revenue and fuller dining rooms, and a well-handled complaint can keep a guest, but the research measures revenue, sell-outs, and stated intentions rather than actual repeat visits.

• Ratings and revenue: Harvard Business School's Michael Luca found that a one-star increase in a restaurant's Yelp rating leads to a 5 to 9 percent increase in revenue, an effect "driven by independent restaurants," with no effect on chains, using Seattle restaurant revenue from 2003 to 2009 (HBS working paper).
• Ratings and full rooms: economists Michael Anderson and Jeremy Magruder found "an extra half-star rating causes restaurants to sell out 19 percentage points (49%) more frequently" in San Francisco reservation data (The Economic Journal, 2012, author copy).
• Complaint handling: a Cornell Hospitality Quarterly study of 802 casual-dining guests found "a positive and significant association between guests' reported satisfaction with the outcome of their complaint and their repeat patronage intentions" (Susskind and Viccari, 2011).
• The recovery paradox: a meta-analysis found a great recovery lifts satisfaction but has no significant effect on repurchase intentions (de Matos and colleagues, 2007). Fixing a mistake well helps; not making it helps more.

What will change for restaurant retention in 2027?

In 2027, holding on to a guest will mean holding on to someone who visits a little less, orders a little less per trip, and compares value harder, so each return visit is worth more and is easier to lose on price.

The traffic numbers already show it. In the National Restaurant Association's August 2026 tracking survey, 46% of operators reported lower customer traffic than a year earlier and 43% higher, even as 51% reported higher same-store sales (National Restaurant Association, September 2026), so checks, not visits, are carrying the growth.

Guests are trading down, not disappearing. In Revenue Management Solutions' Q2 2026 consumer survey, 33% said they're spending less of their disposable income on restaurants than a year ago, and of those, 47% are ordering cheaper items and 45% are choosing more affordable restaurants (FastCasual, August 2026). Technomic found 46% of consumers visiting restaurants less often because of rising gas prices (Nation's Restaurant News, July 2026).

Baskets are shrinking for a second reason. Gallup found 11% of US adults say they currently take a GLP-1 drug for weight loss, up from 3% in 2024 (Gallup, July 2026). Circana's receipt and survey data show GLP-1 users' visit frequency "remains stable" while they order 1% fewer items per trip (Circana, January 2026), and Cornell researchers using Numerator transaction data on about 150,000 households found spending at limited-service restaurants fell about 8% after a household member started the drug (Cornell Chronicle, December 2025).

What it means for a one to three location independent: track visit frequency and check size as two separate numbers. If your regulars still come every other week but order one item fewer, the fix is a smaller portion or a two-person bundle, not a deeper discount, and the restaurant industry outlook for 2027 covers the costs moving underneath it.

How do you measure your own retention rate?

Your repeat guest rate is the share of unique guests in a set window who visited or ordered at least twice: guests with two or more visits, divided by all unique guests, times 100. Your cohort return rate is the share of one month's first-time guests who came back within a set number of days.

The published numbers above use different definitions and windows, so the benchmark that matters most is your own, measured the same way every quarter.

Step 1: Pick the window and the ID. Ninety days suits a takeout-heavy menu; a special-occasion restaurant may need 180 or 365. Count guests by phone, email, or loyalty ID, because a guest you can't recognize twice can't count as a repeat.

Step 2: Count unique guests. Export the window's orders from your ordering system, loyalty program, and POS customer records, and remove duplicates.

Step 3: Count the repeaters. Count the unique guests with two or more orders or visits in the same window.

Step 4: Divide, then track a cohort. Divide step 3 by step 2. Then take one month's first-time guests and count how many came back within 60 days, and how many made a third visit.

In the illustrative taqueria, 372 repeat guests divided by 1,240 unique guests is a 30% repeat guest rate. Of the 400 guests whose first order came in July, 112 came back within 60 days (28%) and 61 made a third visit (15.3%), so most of the loss happens between the first and second visit.

Write all five next to your sales each quarter; a falling cohort return rate with steady sales means your regulars are carrying you.

On one of my September demo calls, a multi-location family restaurant group had a different online ordering vendor at each location and a group website that was down, so its menus and customer data were scattered across vendors. Here's what I think: you can't count a repeat guest you can't recognize, and you can't recognize a guest across systems that don't share a customer record.

What should a restaurant do with these numbers?

Use them to decide where to look first, then let your own data decide what to fix. The moves cost more attention than money.

• Measure before you spend. Pull your repeat guest rate and cohort return rate first, so you know whether the leak is the second visit or regulars drifting away.
• Fix the second visit first. Give every first-time guest a reason and a date to come back, such as a thank-you message with an offer that expires in two weeks, sent only to guests who agreed to hear from you.
• Make the first reward reachable. A program whose first reward lands within a few visits gives a new guest something to come back for.
• Remember people. In Toast's survey, being remembered beat points, so put names and usual orders where staff can see them.
• Protect the basics. Check every bag, quote honest pickup times, and keep the recipe the same on a slow Tuesday as on the Friday rush, because the second order has to be as good as the first.
• Answer complaints fast. The Cornell research ties a well-resolved complaint to a guest's intent to return.
• Grow the check, not just the count. When baskets shrink, a two-person bundle or a well-placed add-on matters; see how to increase average order value.

How to cite this page

Suggested citation: Pankaj Avhad, "Restaurant Customer Retention Statistics for 2027: 50+ Numbers," DirectOrders, updated October 12, 2026, https://www.directorders.com/blog/restaurant-customer-retention-statistics.

This page is an aggregation. When you quote a single number, please also cite the original study or report linked next to it, and keep its date and method attached (transaction data, survey, or general business research), because the same number means different things in each.

Where DirectOrders fits

DirectOrders is commission-free online ordering for independent restaurants on a flat monthly fee: Pro is $249 a month, and Pro + Voice is $349 with 500 AI phone minutes included, with a 14-day free trial (pricing). Every plan includes a branded website with online ordering plus SMS and email marketing, and the restaurant owns its customer data, which is what makes the retention math above possible: a guest who orders on your own site is a guest you can recognize and, with consent, invite back. It runs alongside Toast, Square, Clover, Lightspeed, and Revel rather than replacing them.

Sources

All checked October 12, 2026. Vendor sources are labeled in the text with what the company sells.

• Company filings and results: Dutch Bros Form 10-K, February 13, 2026; Dutch Bros Q2 2026 results, August 5, 2026; Starbucks Investor Day release, January 29, 2026; McDonald's Q4 2025 earnings call transcript, The Motley Fool, February 11, 2026; Sweetgreen Q4 2025 results, February 26, 2026
• Return and repeat data: Toast, loyalty return rate, July 21, 2026; Toast, restaurant customer retention, July 30, 2026; Toast and Resy, The Regulars Report 2026, July 30, 2026; Bloom Intelligence, June 5, 2026; Upside, Escaping the Loyalty Plateau, July 2026 and release, July 13, 2026; DataDelivers via RestaurantNews.com, March 21, 2023 and March 19, 2026; Paytronix 2026 Loyalty Report and release, April 6, 2026; MediaPost on Paytronix data, February 16, 2017; Olo sponsored content, Restaurant Business, March 12, 2026; Olo, November 15, 2024
• The 70% claim: Thanx blog, January 29, 2015 (archived); Thanx press release, January 29, 2015
• Value and loyalty: Square 2026 Local Economy Report, March 18, 2026; PAR 2026 QSR Operational Index, April 21, 2026; PAR loyalty survey, December 17, 2025; DoorDash and SevenRooms, 2026 Restaurant Industry Trends Report and landing page; Circana, June 12, 2025; Bain and Mainspring, 2000
• The 5% claim: Reichheld and Sasser, Zero Defections, HBR, September 1990; Reichheld and Schefter, E-Loyalty, HBR, July 2000; Reichheld, Bain Brief, 2001; Amy Gallo, HBR.org, October 29, 2014; Byron Sharp, Ehrenberg-Bass Institute; Keiningham and colleagues, Loyalty Myths excerpt, Ipsos
• Direct ordering and data: Restaurant Dive on Paytronix, June 5, 2023; Paytronix, April 1, 2024; NCR Voyix 2025 Commerce Experience Report; Restaurant Dive on NCR Voyix, January 30, 2025
• Messages: Klaviyo 2026 SMS benchmarks; Constant Contact, December 2022 (archived); Mailchimp email benchmarks; Mobilesquared, August 24, 2022
• Reviews and recovery: Luca, HBS Working Paper 12-016; Anderson and Magruder, The Economic Journal, 2012 (author copy, archived); Susskind and Viccari, Cornell Hospitality Quarterly, 2011; de Matos, Henrique, and Rossi, Journal of Service Research, 2007
• 2027 shifts: National Restaurant Association, September 30, 2026; FastCasual on Revenue Management Solutions, August 6, 2026; Nation's Restaurant News on Technomic, July 21, 2026; Gallup, July 7, 2026; Circana, January 14, 2026; Cornell Chronicle, December 19, 2025