How coffee roasters find independent cafes to supply with their beans
A roastery grows by winning cafes that care about coffee. How to pick independent, well-rated cafes in your city as wholesale leads, and what to check before you visit.
Part of the guide Workflow automation
A coffee roaster finds wholesale customers most reliably by starting from the independent cafes in its delivery area, keeping the ones whose customers rate them highly, and visiting those first with a bag of samples. Chains buy centrally and rarely switch on a local pitch. A well-rated independent chooses its own beans, cares how its espresso tastes and can say yes in one conversation.
This is written for owners and sales people at small and mid-sized roasteries. It covers why rating is a better first filter than size, how many independent cafes a big city holds, a made-up week of prospecting, the traps that waste sample bags, and an AI employee that keeps the list current.
Why rating is the first filter, not footprint
A roaster is not selling a commodity. It is selling a flavour, a story about origin and the training that comes with it. The cafes that listen are the ones already proud of their cup. A high rating from many customers is the clearest public sign of that pride: it usually means the drinks are consistent, the staff are trained and the owner pays attention. A cafe rated 3.6 with long complaints about bitter coffee might need you more, but it is less likely to pay for better beans or to change the way it works.
Rating alone is not enough, so pair it with a few other signs:
- Independent, not a chain. Nearby places can keep only independents, so the head office of a chain never lands on a rep's list.
- Enough reviews to trust the score. A 4.9 from eleven reviews says little; a 4.6 from several hundred says a lot. Filter or Sort and limit can set a floor.
- A website or phone. Nearby places can keep only places with one, so the rep has a way to book the tasting.
- What the cafe says about itself. Many specialty cafes name their roaster on their own website. The Browser block can open a public page and bring back that line, where the site states it.
How many independents a big city holds
The pool is larger than most roasters expect, and the share of chains near a city centre is small. These figures come from the market pages built on xMap's places database in November 2026, counted within 10 km of each centre; the mix and ratings are worked out from the cafes nearest the centre, not the whole count.
| City | Cafes listed | Independent, of the nearest | Rated 4.5 or more |
|---|---|---|---|
| Toronto | 622 | 265 of 312 | 162 of the 297 rated |
| London | 3,111 | 310 of 342 | 152 of the 300 rated |
| Sydney | 2,573 | 304 of 334 | 138 of the 310 rated |
xMap's places database, November 2026, within 10 km of each city centre. Independents and ratings are of the cafes nearest the centre that each market page used.
Read the middle column carefully. It does not say 265 of Toronto's 622 cafes are independent; it says 265 of the 312 nearest the centre are. Further out the mix may differ. What the table does show is that a roaster in any of these cities could fill months of visits with well-rated independents alone, so there is no need to chase every cafe.
A made-up week for a roastery in Cape Town
Picture a roastery in Cape Town with one sales person and capacity for about ten new wholesale accounts this year. Everything in this paragraph is invented to show the shape of the work. On Monday the AI employee sends a list of 25 independent cafes within a 20-minute drive of the roastery, each rated 4.5 or more with at least 150 reviews, ranked by xMap LLM against the roaster's own notes: espresso-led menu, a brunch trade, no visible roaster partner on the website. The sales person books six tastings from the top ten by Wednesday. Two say they are happy with their current roaster. One asks for a filter coffee option. Three agree to a trial. The next Monday's list shows only cafes the flow has not sent before.
Notice what the list did not do. It did not decide who to call; it narrowed the hundreds of cafes in the area to a page a person can read with a coffee, and gave a reason for each one.
Mistakes that waste sample bags
- Pitching cafes that roast their own. Some cafes with excellent ratings have a roaster in the back. Their website usually says so; ask the AI to flag the phrase.
- Ignoring delivery days. A great cafe 45 minutes away in morning traffic costs more to serve than it earns. Draw the area by drive time with Travel area rather than a circle.
- Treating a new cafe as a weak one. A place open three months may have few reviews yet be the best lead on the list, because it is still choosing suppliers. Keep a second, smaller list of recent openings.
- Sending the same list twice. Without memory, a weekly search repeats itself and the rep stops reading. Only new fixes that.
The AI employee that keeps the list
- Schedule: every Monday at 07:00 in your time zone.
- Travel area: 20 minutes by car from the roastery, or several areas if you deliver to more than one district.
- Nearby places: cafes and coffee shops inside the area, independents only, with a phone or website, minimum rating 4.3.
- Filter: reviews of 100 or more, so the rating means something.
- Only new: cafes this flow has not sent before.
- xMap LLM: score each cafe out of 100 against your notes and write one line on why.
- Sort and limit: the 15 highest scores.
- Email: the list and a map to the sales person.
What lands in the inbox each Monday is a short ranked list: the cafe's name, rating and number of reviews, its address and phone, a link to its website, the fit score and a sentence such as "independent, espresso-led, rated 4.7 by 600 customers, no roaster named on its site". The Specialty cafes for my coffee roastery template starts with most of this in place; change the city and your notes in one sentence and it runs.
Once a cafe signs, keep it in a table in My data and add Match with my table before Only new, so customers never come back as leads.
Questions
Should a roaster ever approach a coffee chain?
Yes, but differently. A regional chain buys through its head office and usually through a tender or a long trial, so treat it as an account sale and keep it off the weekly cafe list.
Is a 4.5 rating the right line?
It is a sensible start in cities where cafes are rated highly on average. Lower it if the list comes back short, and always pair it with a floor on the number of reviews.
Can it tell me which roaster a cafe uses now?
Only where the cafe says so on its own public website. The Browser block can read that page and bring back the line; many cafes do not say, so treat a blank as unknown, not as no roaster.
How often should the list run?
Weekly suits most roasteries: new cafes appear a few at a time, and a sales person can work 10 to 15 leads a week properly. Daily lists tend to go unread.