How to build a farm-to-restaurant marketplace chefs order from every week.

Serve one food area, one delivery route and kitchens that already want local.

A farm-to-restaurant marketplace has two sides that both find the current system tiring. Chefs who want local produce juggle texts, emails and voicemails from five farms, each with its own availability list, order cut-off and invoice. Small farms spend their evenings sending price sheets and chasing orders, and a single missed message means a case of greens that never sells. Broadline distributors are reliable, but their local range is thin and the farm's name usually disappears on the way to the menu.

Pick the angle

• One metro area. Farms within a few hours' drive and restaurants in one city, so deliveries can share a route.
• One product line. Mushrooms, microgreens and specialty greens, or pasture-raised meat, where chefs care most about the source.
• One buyer type. Independent restaurants first; schools, hospitals and caterers buy differently and can come later.

Find out where the trade happens now. Ask chefs at the farmers market on a weekday morning, because many shop there for the kitchen. Talk to the local food hub if there is one, the farm bureau, and the regional chefs' association. A farmer who already delivers to six restaurants is your best interview: they know exactly which tasks they would hand over.

Chefs want fewer conversations, not more choice. One weekly order page with every farm on it is the product.

Take a small cut of each order, sized for farm margins.

Buyers pay at checkout and you keep a commission per sale. Farm margins are thin, so the commission has to be small enough that farmers do not quietly raise their prices to cover it, which would push chefs back to texting the farm directly. Many food hubs keep a share of each sale to pay for coordination and delivery; your share should reflect what you actually do. If you only run the ordering page, keep it low. If you also run a shared delivery truck, a larger cut is easy to defend.

A buyer service fee on each restaurant order is another option, and it keeps farm prices clean. Chefs accept it when the order saves them time. A cart with items from several farms in one checkout is the main reason a kitchen orders through you instead of five phone calls, so make sure it works well before adding fees.

Talk about payment terms early. Checkout takes payment when the order is placed, while many restaurants are used to an invoice paid a week or a month later. Tell kitchens plainly how payment works before they sign up.

Launch at the start of the growing season with one delivery day.

Timing matters more here than in most marketplaces. Launch a few weeks before the main growing season starts in your region, when chefs are planning spring menus and farms are deciding where their volume will go. Launching in midwinter means a thin page in colder climates, unless you lead with storage crops, meat, eggs and greenhouse growers.

Start small

• One city, one order cut-off and one delivery day. Add a second day only when the first is full.
• Ten to twenty restaurants that already buy local, signed up one by one. Visit between lunch and dinner service, never during it.
• A farm dinner or chef tasting with your growers, where chefs meet the people behind the listings.

Launch checklist

• Each farm has its delivery days, minimum order and cut-off filled in.
• Units and pack sizes are consistent across farms: case, pound, bunch, dozen.
• Rules for shorts, credits and late deliveries are written on one page.
• Payment at checkout explained to every kitchen before its first order.
• Availability updated weekly, on the same day, by every farm.

After the first season, ask chefs which product they still buy elsewhere and recruit a farm for it.

A worked example: Ridge and Hollow Larder, southwest Wisconsin.

Picture Ridge and Hollow Larder, a fictional farm-to-restaurant marketplace in the Driftless region of southwest Wisconsin. Its farms sit in the steep valleys between La Crosse and Madison. Its buyers are independent restaurants, two brewpubs, a hospital cafeteria and a handful of caterers in Madison, one to two hours east. Every order travels on a single refrigerated route that leaves the valley on Wednesday evening and reaches kitchens on Thursday morning.

How it started

The founder spent eight years as a sous chef in Madison and kept a spiral notebook of farm phone numbers. Every Monday she texted six growers, copied their replies into a spreadsheet and placed six separate orders. When she left the kitchen, three former colleagues asked whether she could keep doing the ordering for them. She signed eleven farms over one winter, most of them vendors she already knew from the Saturday market around the state capitol, and opened in late April with spring greens, eggs, frozen pork and cheese.

Categories

• Produce
  • Greens and herbs
  • Roots and storage crops
  • Seasonal specialties, such as squash blossoms and sweet corn
• Proteins
  • Pork and beef
  • Poultry and eggs
• Dairy
  • Cheese
  • Butter and cream
• Pantry
  • Flour and grains
  • Maple syrup and honey

Who it serves

The core buyers are chef-owned restaurants that print farm names on the menu and order between two and eight cases a week. The hospital cafeteria buys eggs, apples and storage crops in volume but needs paperwork on every farm first, so it joined in the second season. Caterers order in bursts around weddings and graduations, which fills the route in June.

How it makes money

Ridge and Hollow keeps a commission on every order, slightly lower for farms that drop product at the shared cooler in Viroqua themselves. Restaurants pay a small service fee per order that covers part of the contract courier. Two creameries pay for seller memberships because they list dozens of cheeses and want to appear first in dairy. In the second summer the founder added featured listings for the July and August tomato rush, when nine farms compete for the same chefs.

None of that worked until the routine did: availability updated by Sunday at 6 p.m., orders closed Tuesday at noon, deliveries on Thursday before 10 a.m. Chefs learned the rhythm within a month.

The first 90 days: from the first farm visit to a steady weekly order.

The first 90 days of a farm-to-restaurant marketplace should end with real weekly orders, not a finished website. Count backward from the date the first spring greens are ready in your region, because that is when chefs rewrite menus and are most willing to try a new supplier. In cold climates the plan often starts in January or February, while farmers have time to talk and are still deciding how much to plant for wholesale.

Two rules set the order of work. Nothing goes live until the weekly routine is fixed: when farms update availability, when orders close and when the deliveries arrive. And kitchens are signed one at a time, in person, because a cold email from a new platform lands beside a dozen sales pitches every week.

Measure each phase with a number you can count: farms with complete listings, kitchens with an account, orders placed in a week. If week 8 comes and ten farms are not ready, move the launch rather than open with a thin page. A late launch with a full cooler beats an early one that teaches chefs to ignore you.

Ways to build it, tested against a real week of kitchen orders.

Whatever you choose has to cope with what makes farm ordering different: quantities that change every week, meat priced by the pound with variable weights, a minimum order per farm, and one cart that pays several farms at once. Before you commit, run a real week of orders from one restaurant through each option you are considering.

Building it yourself with developers gives you full control over order cut-offs, route planning and printed pick lists for drivers. It also means paying for every change, and farm ordering changes constantly in the first season. It makes sense later, once the routine is proven and you need something no product offers, such as standing orders or invoices on terms.

Installing a marketplace script or a WordPress plugin on your own server costs less at the start, but multi-vendor checkout, per-farm delivery charges and payments to each farm usually need several add-ons working together. Someone has to update them, keep the server secure and fix the cart when it breaks the week before holiday orders.

A hosted builder such as Yclas gives you a cart across sellers, shipping per seller and Stripe checkout without a server to run, so your hours go into farms and kitchens. The trade-off is working within the features it has: you shape your routine around them instead of writing custom logistics.

On any route, make sure you can export listings and orders. The farm and chef relationships are the business; the software is replaceable.

What a farm-to-restaurant marketplace costs to run.

The software is the smallest part of the bill. Most of the cost is your time in the first months: finding the first listings, answering questions and checking what gets posted.

Expect some spending on food safety paperwork checks, a cooler or insulated totes if you aggregate orders, and possibly a shared delivery vehicle or courier for farms that cannot drive into the city. Liability insurance for a business that handles food is worth asking a broker about.

Mistakes that empty a farm-to-restaurant marketplace.

• Opening in the dead of winter with a thin page. A chef who sees four items in February decides the site is not worth checking in May. Open just before the first spring harvest, or lead with storage crops, meat, eggs and greenhouse growers if you must start earlier.
• Letting every farm invent its own units. One farm sells greens by the pound, another by the clamshell, a third by the case of unknown weight. Chefs cannot compare and stop trying. Agree a short list of units and pack sizes before the first import and hold every farm to it.
• Signing restaurants before the delivery works. A missed Thursday drop costs a kitchen a dish on Friday night, and chefs do not give second chances easily. Run two or three test weeks with friendly kitchens before you sign anyone you need to impress.
• Setting a commission farms pass straight to chefs. If tomatoes cost more on your site than on a text to the farmer, chefs will text the farmer. Check prices on your site against what farms charge directly, and lower your cut or add value until they match.
• Leaving availability stale. Product listed on Monday may have sold at Saturday's market. When a chef orders what a farm no longer has, the short lands on your reputation. Make the weekly update a condition of staying on the site, and hide farms that miss it.
• Ignoring the kitchen clock. Calls during lunch service, an order cut-off on the day the chef does inventory, deliveries during prep rush. Ask each kitchen when it orders and receives, and set your schedule around the majority. A cut-off two hours after most chefs finish their weekend inventory will collect far more orders than one set for your own convenience.
• Hiding short orders. A farm that shorts quietly leaves a chef finding out at the back door. Require notice before the truck leaves, refund automatically, and track shorts per farm.
• Building a vegetable-only catalog. Greens and tomatoes are easy to recruit, but a kitchen also wants eggs, a meat producer and a cheese maker on the same order. Without them the chef still makes three other calls, and your cart saves little time. Recruit at least one grower for each part of the plate before launch, even if their range is small.
• Expanding to a second city too early. Two half-full routes cost more than one full one. Fill the first delivery day before you add a second day, and fill both before you add a new city. A full route also gives you the numbers to persuade farms in the next area.

Kitchen and farm terms a founder needs to speak both languages.

Chefs and farmers each have their own vocabulary, and your site sits between them.