Wilson Wu
Chief Revenue Officer, Snappy
You cannot control food prices, the weather, or the wage floor. What you can control is who walks back in during February. Operators across the region describe January through March as the hard stretch: a Fredericton owner told CBC it can be a tough time at his restaurant (CBC). The ones that come out intact are the ones that spent the fall building a list of regulars they can reach directly.
Halifax, Dartmouth and Fredericton operators all describe the same winter: costs that keep climbing and a dining room that empties out after the holidays. In January 2026, CBC News talked to six restaurants in the Halifax area and every one of them said it was struggling to turn a profit (CBC). This article separates the part of that you cannot touch from the four things you can. Our related piece on adapting this winter without a tax holiday covers the wider policy backdrop.
Summary
On this page
The Atlantic Canada restaurant winter slowdown is the yearly drop in guest visits and profitability that hits Nova Scotia and New Brunswick restaurants from January through March, driven by rising costs and a thinner customer base rather than one bad month. It is real, it is measurable nationally, and Atlantic Canada carries a version of it that bigger markets do not. Restaurants Canada’s Q1 2026 report found 71% of operators reporting declining profitability, 54% reporting fewer guests and 49% reporting lower sales, with 36% operating at a loss or merely breaking even, triple the levels from 2019 (Restaurants Canada).
By late summer the loss-making share had grown, though not every signal moved the same way. The Agri-Food Analytics Lab at Dalhousie reported in August 2026 that Restaurants Canada now puts 41% of operators at a loss or merely breaking even, up from 36% in March (Agri-Food Analytics Lab). Even as that share climbed, the Lab revised its own forecast of net closures down from about 4,000 to between 1,500 and 2,500, and it noted that accommodation and food-service insolvencies reached 360 in the first half of 2026, roughly 7.5% more than the same period a year earlier (Agri-Food Analytics Lab).
Going from 36% to 41% is a five percentage point move in the wrong direction over five months.
What makes the Atlantic version distinct is the shape of the customer base. The same analysis notes that smaller markets across Atlantic Canada and the Prairies “depend more heavily on seasonal traffic and narrow customer bases,” and that independent full-service restaurants remain the most exposed format in the country (Agri-Food Analytics Lab).
Narrow customer base is the phrase to sit with. In a city of two million, a quiet February is a rounding error against a huge pool of new faces. In Dartmouth or Bedford, the guests who stop coming in February are the same guests who were coming in November, and nobody is arriving to replace them.
The consumer side is not helping. Restaurants Canada’s November 2025 revised forecast cites an Angus Reid survey finding 74% of Canadians had reduced discretionary spending that year, with dining out at restaurants (56%) and restaurant takeout or delivery (50%) the top two areas of cutback, and it revised 2026 foodservice growth down to 2.3% from a previous forecast of 3.4% (Restaurants Canada).
Three costs sit outside your control: food prices, the minimum wage, and the buying power that favours chains over independents. Start by naming them, because time spent fighting a fixed cost is time stolen from the levers that actually move.
Food cost is the loudest one. Consumer food prices climbed 4.7 per cent year over year in November 2025, the highest rate in nearly two years according to Statistics Canada, with the biggest hikes in fresh or frozen beef, up nearly 18 per cent, and coffee, up almost 28 per cent (CBC). Owner Charbel Gerges of Yassou Souvlaki & Donair in Fredericton told CBC his lettuce ran $30 to $40 a case, went to $130, then settled back to $60 to $70. You can hedge that with menu design and supplier work. You cannot set the price.
Labour is the second fixed input. Nova Scotia employers must pay at least $16.75 per hour as of 1 April 2026, rising to $17.00 per hour on 1 October 2026 (Province of Nova Scotia). We covered what that means for scheduling and menu pricing in our breakdown of the Nova Scotia minimum wage increase. It is a number to plan around, not negotiate. For the broader Canada-wide labour cost picture, see our 2026 labour cost projections for restaurant operators.
Buying power is the third, and it is the one that separates you from the chains. Sylvain Charlebois of the Agrifood Analytics Lab told CBC that most of the restaurants expected to close would be independently operated, “just because they often don’t have the support that chains would receive when it comes to logistics, procurement and things like that” (CBC).
Operators who do have scale say the purchasing work is relentless. Mike Babineau, whose group runs nine establishments across New Brunswick and Nova Scotia with about 200 employees and three staff dedicated purely to purchasing, put it plainly: “You’re spending dollars to save pennies. Sometimes that’s what it feels like” (CBC).
If a nine-location group with three full-time buyers feels that way, a single independent chasing another two per cent on produce is not where the winter gets won.
The winter loss is a frequency problem before it is a traffic problem, and frequency is measurable per guest.
Sherif Turk of USTA Turkish & Mediterranean Restaurant in Dartmouth told CBC that some customers he used to see twice a month now come once a month (CBC). Babineau reported the same pattern in New Brunswick: “Instead of going out two or three times a month, they’re going out once a month, sometimes once every other month” (CBC).
Run that as arithmetic on one household. Twice a month is 2 x 12 = 24 visits a year. Once a month is 12. That is 12 visits lost from a single regular, and nothing about it shows up as an empty room on a Saturday. The dining room still looks fine. The week does not.
The Agri-Food Analytics Lab makes the same point about independent full-service rooms: “A dining room can appear full on Saturday evening while the business loses money over the entire week” (Agri-Food Analytics Lab).
So the first question is not how to get new guests in February. It is whether you can name the guests you already had in November. Most independents cannot, because the visit lives in the cash drawer and the ordering app, not in a customer record. A loyalty program is the cheapest way to fix that: Snappy Loyalty collects points in-store and online against the same customer, and lets you collect names, birthdays and email addresses at signup (Snappy). Our guide to starting a restaurant loyalty program walks through the setup step by step.
Once a regular has a record, “she has not been in for six weeks” becomes a query instead of a hunch. That is the whole game of keeping customers coming back in February.
Do you know which regulars stopped showing up this fall?
Walk through your own fall visit numbers with us, guest by guest, and see exactly who has gone quiet before winter sets in.
Show me my regularsGift cards are a cash-flow instrument disguised as a marketing one. Money arrives in December, when your room is already busy, and the meal is served in January and February, when it is not.
That timing is the point. December sales are the easiest sales of your year because guests are already buying gifts, and the redemption lands in exactly the weeks the Agri-Food analysis flags as thin. You are moving revenue from your strongest month into your weakest without discounting anything.
Digital delivery matters more here than it looks. Snappy’s gift card and membership product sends cards instantly by SMS and email, offers digital and physical options, supports auto reload, and allows balance redemption at multiple checkout points (Snappy). A card someone can buy from their phone on 23 December is a card you actually sell.
Memberships work the same way with a longer tail. A paid membership is prepayment plus a reason to come back, which is precisely the frequency problem the Atlantic operators described. Set both up in the fall, not in December, so the promotion is live before the first gift-buying week.
One honest caveat: nobody should sell you a redemption or breakage number as a forecast. Treat prepaid revenue as deferred revenue, because the food still has to be cooked when the card gets used.
If your only channel to a guest is a marketplace app or a paid ad, you do not have a channel. You have a rental agreement that gets more expensive on the days you need it most.
Cutting what you pay to delivery apps starts with direct ordering. Snappy’s online ordering carries 0% commission fee on every online order and lets customers order directly from Google, with the restaurant owning all customer data (Snappy). On a February snow day, a direct order and a marketplace order look identical to the guest and very different on your P&L.
Owning the data is what makes the next step possible. Snappy’s CRM holds customer profiles and runs segmented campaigns, which means a winter message can go to one defined group rather than your whole list: guests who visited three or more times in the fall and have not been back since 15 January. Our guide to setting up marketing automation for a restaurant walks through building those segments.
Reviews are the other direct channel, and winter is when the risk concentrates. A short-staffed Tuesday produces the review that costs you March. Snappy Feedback puts a QR review funnel in the room and routes happy guests to Google while sending unhappy ones to internal feedback with real-time owner alerts, so the complaint reaches you before it reaches the public.
For the reviews that do land publicly, Snappy Reputation Management (AI) generates draft replies for Google reviews with a one-click review, edit and confirm workflow, flags low-rating reviews immediately, and sends notification alerts for critical reviews (Snappy). No response is auto-posted. The owner still approves every word.
Four levers fit a winter budget: a loyalty program, prepaid gift cards and memberships, a segmented CRM message to lapsed regulars, and a QR review funnel. The right sequence matters as much as the list, because three of the four only work if they were set up months earlier. This table sorts by what each lever costs to run once it exists, what it actually brings back, and when it needs to be in place.
| Lever | What it costs to run in winter | What it brings back | When to set it up |
|---|---|---|---|
| Loyalty points | Almost nothing once live. Rewards are funded by visits that would not otherwise happen | Frequency from guests you can now identify by name, in-store and online | Fall. It needs months of visit history before February to be useful |
| Digital gift cards and memberships | Card production and promotion only | December cash spent in January and February, plus a reason for the recipient to discover you | Early fall, so the promotion is running before gift-buying season |
| Segmented CRM campaign to lapsed regulars | Message send cost. No commission, no ad auction | The specific guests who dropped from twice a month to once, contacted directly | Any time, but the customer records must already exist |
| QR review funnel (Snappy Feedback) | QR stands and stickers in the room | Recovery of the guest who had a bad Tuesday, before the review is public | Immediately. This one works from day one |
| Direct online ordering versus a marketplace on a snow day | 0% commission on every online order through Snappy, versus a commission on every marketplace order | Full margin on the orders you were going to get anyway, plus the customer data | Before the first storm. Menu and Google ordering setup take lead time |
| AI phone answering (Snappy AI Phone General Inquiries, no prerequisite) | Runs whether or not a human is in the building | Every caller gets an answer instead of voicemail on a closed Monday, including catering and large-event enquiries | Before you cut hours, not after |
Read the last column first. Three of these six levers are worthless in February if you start them in February, which is why this is a September and October decision.
Cutting a shift is the fastest way to save money and the fastest way to teach a regular that you might be closed. Do the cheaper thing first.
Closing on slow days is already common. Janick Cormier, vice-president Atlantic at Restaurants Canada, told CBC that some restaurants have been shaving operating expenses by closing on slower days of the week or shortening the workday, and predicted more of it (CBC). It works. It also removes the option for the guest who was going to come Tuesday.
Before you close the night, price it properly. The Snappy Team App gives live sales tracking with month-over-month comparison, staff clock-in and clock-out, and inventory counts with COGS, which is what turns “Tuesday feels dead” into a labour cost per cover you can compare against Wednesday.
Then reduce the cost of staying open. Fewer front-of-house hours with QR ordering at the table, a phone answered automatically instead of by a manager who is also expediting, and orders arriving direct rather than through a commission channel all shrink the cost of a slow night without closing the door.
The phone deserves its own line. Snappy’s AI Phone General Inquiries has no prerequisite and answers every call from the merchant’s own FAQ, so a caller asking whether you are open on a stormy Monday gets an answer whether or not anyone is standing at the host stand. That is the difference between reduced hours and lost customers.
Cutting hours is still a legitimate move. It should be the fourth thing you try, not the first, and it should be a decision made against your own labour and cover data rather than the feeling of a quiet room.
Demand falls while costs do not, and the regional customer base is thinner. Consumer food prices were up 4.7 per cent year over year in November 2025 per Statistics Canada (CBC), Nova Scotia’s wage floor rises to $17.00 per hour on 1 October 2026 (Province of Nova Scotia), and smaller Atlantic markets “depend more heavily on seasonal traffic and narrow customer bases” (Agri-Food Analytics Lab).
Repeat visits. Operators in both provinces describe regulars dropping from two or three visits a month to one (CBC), which is a frequency problem you can address with loyalty, prepaid gift cards, direct outreach to lapsed guests and a feedback funnel. None of those require the price of beef to fall.
They move cash from your busiest month to your quietest one, which is a genuine timing benefit for January and February. Snappy’s gift cards and memberships deliver instantly by SMS and email with auto reload and multi-point balance redemption (Snappy). Treat the money as deferred revenue, because the meal still has to be cooked later.
That is your call, made against your own numbers, not an industry average. Restaurants Canada has confirmed operators are closing slower days and shortening the workday (CBC), but cut the cost of being open first: fewer front-of-house hours, calls answered automatically, and orders arriving direct.
They bring volume and they take a commission on every order, which hurts most when volume is already thin. Snappy’s online ordering charges 0% commission fee on every online order and lets guests order directly from your Google listing (Snappy), so the same snow-day order keeps its margin and its customer data. Many operators run both and push regulars to the direct link.
The Atlantic Canada restaurant slow season is a fixed-cost problem wrapped around a frequency problem. Food prices, the wage floor and the weather are inputs you plan around. The guests who came twice a month in November and once in February are the variable you can still move, and moving them is worth more than another two per cent on produce.
Four things fit inside a winter budget: a loyalty program that turns anonymous visits into named regulars, prepaid gift cards and memberships sold in December, a segmented CRM message to the guests who have gone quiet, and a review funnel that catches the bad Tuesday before Google does. Three of the four have to be built in the fall to matter in February.
Nova Scotia and New Brunswick operators are not short on effort. Nine-location groups with dedicated buyers still describe it as spending dollars to save pennies. The leverage left is on the demand side, with the people who already like your food. If you want the whole regional picture, our explainer on Newfoundland and Labrador’s Restaurant Guarantee Program covers how another Atlantic province is approaching the same squeeze.
Build your winter regulars list before the first storm. Talk to our team about getting loyalty, gift cards and direct ordering live while you still have a busy fall to collect names.
Book a walkthrough
Build your winter regulars list before the first storm
Walk through what needs to be live this fall, loyalty, gift cards and direct ordering, so you can reach regulars by name in February.
No-obligation walkthrough
Know Who Stops Coming Back This Winter
Regulars are quietly cutting visits in half. See what a fall list of names could catch before February.
On this page
Get The Latest Restaurant Data, Trends & Tips
