Types of Distribution Business: Models and Industry Playbooks

This hub is written for US wholesale operators. Permits, tax nexus, and recall law are state-plus-federal here; a UK cash-and-carry or an EU importer will need a different permit stack. If you are still choosing a category, do not start with a generic LLC-and-warehouse story. Start with the rulebook that will sit on every pallet: FDA food facility registration and FSMA 204 traceability for food; DSCSA and UDI for medical; MoCRA for cosmetics; Auto Care ACES/PIES for parts; MAP and authorization letters for electronics.

A $180,000 warehouse lease that looks cheap becomes expensive the week a buyer asks for lot genealogy, temperature logs, or an authorized-distributor certificate you cannot produce. Vertical choice also sets your gross-margin band. Foodservice specialty often lands 18–28% after freight if spoilage stays under 2%. Authorized electronics after MAP leakage and price protection can sit at 8–15%. Medical consumables can look like 22% on paper and print 14% after dating write-offs and GPO admin fees. Those ranges are directional — your SKU mix will move them — but they are why “any wholesale business” is a bad plan.

A distribution channel is who owns inventory, who owns the account, and who can set or break price. Direct manufacturer-to-buyer skips you. One-tier wholesale (you buy, you sell, you deliver) is the default US model. Two-tier adds a regional wholesaler or buying group under you. Hybrid keeps house accounts direct and parks the long tail with you. Exclusive, selective, and intensive are coverage rules on top of those paths — they are not synonyms for “premium” or “mass.”

Coverage rules change working capital before they change vanity revenue. Exclusive territories usually demand higher fill rates, training, and safety stock on a narrower line card. Intensive coverage demands more outlets, more promo leakage, and a MAP policy you can actually enforce. If you have not read distribution channels explained and exclusive vs intensive, stop here and do that first. Signing a supplier before you know whether you are the only door in a DMA or one of forty is how gray-market inventory shows up on Amazon with your MAP smashed.

Open food and beverage if your first accounts are c-stores, independent grocers, or independent restaurants and you can live with dated product, FEFO slotting, and a written recall drill. Open electronics if you already have authorized access or a realistic path to it — gray-market “deals” are not a launch strategy. Open beauty only if you can name the MoCRA responsible person and store temperature-sensitive actives correctly. Open auto parts if you can staff same-day routes and keep ACES fitment clean. Open medical supplies if you can pass a hospital vendor packet without inventing SOPs the night before.

If you cannot yet name a buyer list of 30 accounts who already buy the category weekly, you do not have a vertical yet — you have a product idea. The launch sequence that survives first-year cash pressure is: 1) confirm permits and insurance for that vertical, 2) lock supplier terms and authorization in writing, 3) stock a narrow A-list, 4) prove OTIF and fill rate on a tight route, 5) widen SKU count only after turns and spoilage or obsolescence are boring. Software comes after the rulebook; see distribution software only when lot, EDI, or pricing matrices are blocking invoices.

A second filter is insurance and credit, not Instagram. Product liability, cargo, and warehouse legal-liability limits that a grocery chain will accept are different from a salon or a clinic. Ask the underwriter about the category before you sign a supplier that ships aerosol, sterile product, or anything that lives below 40°F. The same $2 million COI that satisfied a gift-shop brand will bounce on a hospital packet. If the premium shocks you, that is the vertical talking — listen before you lease racking.

Most failed launches are not “wrong products.” They are thin routes and fat catalogs. A specialty beverage route with 40 doors and a $420 average drop can fund a sprinter van. The same inventory sprayed across 200 doors with $90 drops will lose money after fuel, driver time, and shrink even if invoice gross margin looks healthy. Density is an industry decision: food and auto parts punish thin routes immediately; electronics and beauty can hide the damage until a season turns and you are stuck with EOL or expired lots.

Work the cash conversion cycle before you celebrate a booked PO. Inventory days + receivable days − payable days is the number your lender already knows. A food distributor on Net 10 from a craft brand and Net 30 to retailers is funding 20–40 days of product plus spoilage. An electronics distributor who bought a closeout 90 days before a refresh is funding obsolescence. If that math does not fit a line of credit or owner cash, change the vertical or the terms — do not “hope for velocity.”

Write a one-page “kill criteria” before the first PO: spoilage or obsolescence over X, fill rate under Y for four weeks, or a permit you cannot obtain in 90 days. Operators who skip that page keep feeding a dead vertical because the van is already wrapped. The hub’s job is to send you into one spoke with a rulebook and a kill number — not to make every industry sound equally friendly.

A clean 90-day vertical test is boring on purpose. You can name the permit file, the lot or serial story, the top 30 doors, and last week’s OTIF without opening email. Purchasing can show turns on the A-list. Sales cannot override credit without a name on the exception. Returns have a cage and a clock. If any of those are still “we’ll formalize it after we get going,” you are still in a product hobby, not a distribution company.

Use the spokes as operating manuals, not as inspiration. Food is the dating-and-traceability manual. Electronics is the authorization-and-EOL manual. Beauty is the MoCRA-and-sell-through manual. Auto is the fitment-and-same-day manual. Medical is the quality-system manual. Channels and exclusive/intensive are the contract manuals that sit under all five. If you only read the vertical that matches your dream brand, you will still sign a coverage clause that bankrupts the route.

A vertical is a floor plan. Food needs a cooler with a FEFO face, dry that does not block the cooler door, and a quarantine cage for temperature-abused inbound. Electronics needs a serial cage, an RMA bench, and a demo closet you can age on purpose. Beauty needs climate that does not cook actives and a tester cage that is not sellable stock. Auto needs a will-call counter, a hot-pick zone for the 200 parts that pay rent, and a core cage with a clock. Medical needs sterile away from dirty returns, dated short-face locations, and a locked shelf for anything an auditor will ask to see. Twelve thousand feet of undifferentiated racking “until we know” means you will reslot under live orders and fail the first chain walk.

Buy locations for the first 400 SKUs plus air for the A-list you do not know yet — not a mezzanine a broker promised you would grow into. Staging belongs at the door that loads the van. Every vertical needs a returns cage on day one: dated food, serialized electronics, beauty testers, auto cores, medical complaints. A shared returns pallet by the trash is how lots vanish and how a FSMA or DSCSA drill dies. Walk the empty box with the spoke you chose and put tape on the floor before racking ships. Layout is cheaper than a second lease and cheaper than a failed mock recall.

The vertical table does not show cash conversion, and that is where the industries split. Food c-stores can be weekly COD or Net 7 while a craft brand wants Net 15 — you fund 8–20 days plus spoilage. Electronics VARs want Net 30–45 and the brand wants cash or Net 10 on a last-time-buy. Beauty salons will ask for Net 30 on a launch you already paid for. Auto shops live on Friday statements and stretch you when a fleet is late. Medical clinics and GPOs can sit 35–70 days after a packet that took a month. Inventory days plus DSO minus payable days is the number your lender already runs. Fifty-five days on $180,000 of stock is $180,000 plus the receivable, not “a good relationship at the bank.”

Write credit before the first invoice: application, limit, personal guarantee on thin doors, COD until three clean pays, and one named person who can override. Sales that can ship around credit will ship around credit. Call the 15-day bucket while the buyer still remembers the drop. Reserve 2–4% for bad debt on independents and shops; reserve more if you chase an IDN that pays when AP feels like it. If the vertical’s normal DSO cannot fit owner cash or a line, change terms or change vertical. Do not paper a 70-day GPO with a 15-day supplier and call it a launch.

The first meeting is a service promise, not a catalog. Food: I can keep these 12 SKUs in date twice a week; here is last week’s fill and a lot I can trace by lunch. Electronics: I am authorized; here is the letter, how MAP tickets close, and the EOL calendar. Beauty: here is the responsible-person name, the back-bar I will not overstock, and education on Tuesdays. Auto: here is fill on the 200 parts that strand a bay, the next run, and how a core comes back. Medical: here is the recall SOP, the COI, the storage log, and the 30 SKUs that stop a clinic day. If you cannot say those sentences without notes, you are still shopping for a vertical.

Ask the buyer three numbers: current supplier fill, current drop days, and the last time they were left short on an A-item. Write the answers down. Your offer sits against those numbers, not against a rebate that trains them to wait for a deal. Leave a one-page terms sheet — cutoff, minimum, returns, credit — and walk from doors that will not sign a minimum. Thirty named accounts who already buy the category weekly beat a hundred “send me info” emails. If you cannot name thirty, stay on this hub and pick a smaller geography or a different spoke.

The deaths rhyme across industries: fat catalog, thin route, unwritten house accounts, and a permit that was going to be handled in month two. Food dies on dating and a van that is not cold. Electronics dies on a gray pallet and an EOL closeout a rep needed. Beauty dies on a third brand with 90 shades and a salon that dumps onto a marketplace. Auto dies on dirty interchange and 11% outs. Medical dies on a packet you cannot finish and a sterile vanity SKU that expires. Write a kill number before the first PO — spoilage or obsolescence over X, fill under Y for four weeks, permit not in hand in 90 days — and honor it when the van is already wrapped.

A quieter death is buying theater. A WMS with lot, EDI, and a portal before 40 doors spends the cash that should have been inventory and a driver. The opposite death is a spreadsheet after a chain asks for a lot list at 10:14 a.m. Open the software spoke only when a control is blocking invoices or an audit. Until then a boring ERP with inventory and AR, plus a paper recall drill that actually prints doors, beats a demo week you will not implement. If a vendor needs discovery to tell you food needs lots, this hub already told you.

Do not start in software. Start with the rulebook on the pallet. You need lot and expiry when a buyer or an FSMA 204 drill cannot be answered from receiving photos and a binder. You need serials when warranty and gray-market fights require them. You need fitment when a binder is costing $180 returns. You need UDI capture when a clinic item master will not match your invoice. You need MAP monitoring when street price is teaching independents to stop pre-ordering. Those are triggers. “We should be more digital” is not a trigger.

When you buy, buy the module that matches the spoke: FEFO and lot for food, lifecycle and authorization for electronics, shade/lot and an adverse-event file for beauty, ACES/PIES and cores for auto, storage logs and a recall print for medical. EDI arrives when a named account makes 850/856/810 a gate. If two modules would fix the block, implement one and watch fill, aging, and DSO for 60 days. This hub’s job is to send you into one spoke with a layout, a credit policy, and a kill number. Software is a consequence of that choice, not the choice.