Selling to restaurants: how to find them and when to approach

Restaurants are easy to find and hard to sell to, because the owner is busy every hour the doors are open and decides most purchases in a few short windows. The best time to approach is just before or just after a restaurant opens, when its kitchen is hiring, or when something is visibly going wrong, such as a rating that keeps slipping. The best way to approach is a short, specific message to the owner, sent outside service hours, followed by a visit if you sell something they need to see.

This article is for anyone whose customers are restaurants: point-of-sale and booking software, food, wine and coffee suppliers, linen, cleaning and pest control, delivery platforms and equipment dealers. It covers who decides, the moments that open the door, how to read a city's restaurant market, and an AI employee that finds those moments for you.

Who actually decides

In an independent restaurant, the owner decides almost everything, often with the head chef for anything that touches the kitchen. In a chain, the restaurant manager decides very little: menus, suppliers and software are chosen at head office. Before you spend a week calling branches, find out which kind of market you are in. xMap's places database shows how different cities are:

3,792Restaurants in London

326Nearest counted

50Chains

276Independent

Within 10 km of the centre of Sydney the same database lists 1,640 restaurants, and of the 303 nearest the centre only 10 belong to a chain. Toronto sits between the two: 2,125 listed, and 41 chains among the 290 nearest the centre. In a market like Sydney's nearly every door you knock on belongs to the person who decides; in London one call to the right head office can be worth dozens of branch visits. The Nearby places block marks every place as a chain or independent, so your list can hold only one kind.

The moments that open the door

• The weeks around opening. A new restaurant is choosing a till, a booking system, suppliers, linen and cleaners all at once. Arrive after the contracts are signed and you wait a year. Nearby places can keep only places that opened in the last 6 or 12 months.
• Hiring in the kitchen. A restaurant posting for a head chef or sous chef is changing its menu or about to open a second site. The Job postings block finds those postings by title and city.
• A strong rating with no website. A well-loved place with no site of its own is a natural lead for online ordering, booking or marketing services.
• A rating that keeps falling. Slipping scores often mean trouble with service, cleanliness or supply, all things someone sells a fix for. Approach with care and with something useful, not a pitch about their reviews.

Each of these is a field or a filter, not a guess, which is why it can be watched every day by an AI employee instead of by a rep scrolling a map.

When to send, call or visit

Timing within the day matters as much as timing within the year. Lunch and dinner service are the worst moments to call, and the hour before service is not much better. Mid-afternoon on a weekday is usually the quietest. An email or WhatsApp message lets the owner read it when they choose; the Outreach block sends only between 08:00 and 18:00 on weekdays in the restaurant's own time zone, and waits for your approval on each first message. A visit works best once they have answered, with something they can taste, touch or try.

Reading a city before you start

Before choosing where to begin, compare a few districts. Look at how many restaurants each has, how many opened in the last year and how many are rated 4.5 or more. Of the 326 London restaurants nearest the centre, 143 are rated 4.5 or more; of Toronto's 290 nearest, 103 are. Twenty-two of those 290 in Toronto opened in the twelve months before November 2026, against 9 of the 303 nearest in Sydney and 19 of the 326 in London. A city or district with many recent openings is where suppliers and software firms find the most first-time buyers; one with many highly rated places is where premium products land best.

An AI employee that finds the moment

Here is one flow that combines the opening and hiring signals. Say, as a made-up case, you supply coffee and equipment to independent restaurants in Mexico City:

1. Schedule: every Monday and Thursday at 15:00, after lunch service.
2. Area of interest: the districts you deliver to.
3. Nearby places: restaurants and cafes, only independents, opened in the last 6 months, only places with a phone or website.
4. Only new: each restaurant reaches you once.
5. xMap LLM: "Score 0 to 100 how likely this restaurant is to buy specialty coffee and a commercial machine; give one reason."
6. Sort and limit: highest score first, keep 15.
7. Email to you, or Outreach to the restaurants, approved first.

A second, smaller flow can run Job postings for "head chef" and "sous chef" in the same city every weekday and send those restaurants to the same person. Together they give a short list twice a week of places at a buying moment, each with a reason.

Mistakes that waste the pipeline

• Treating every restaurant alike. A chain branch and an independent need different approaches and different people. Split them before you start.
• Arriving after the opening rush. A place that opened eight months ago has signed its contracts. Weight the newest openings highest.
• Pitching on a bad review. Owners rarely thank strangers for pointing out their low scores. Offer a fix, not a diagnosis.
• Long first messages. Three sentences, one question, your name. Anything longer is read after closing, if at all.
• Calling during service. Nothing ends a conversation faster.

Ready-made flows exist for several of these jobs: new restaurants to sign up the week they open, new independent cafes and restaurants for point-of-sale software, restaurants hiring chefs, as kitchens to equip and restaurants for a wine distributor. The restaurants and cafes page collects more. For writing to them with follow-ups that stop when they reply, see outreach that waits for your approval.