A business model answers three questions: who pays you, what they get in return, and how the money and the goods actually move between you. The type of business model is just the name of a pattern other companies have already run, which matters because every pattern arrives with its own set of known problems.
That last part is the useful bit. Picking a model is not a branding exercise. It decides whether you get paid before or after you ship, whether you hold inventory or someone else does, and what it costs you to put one order in one customer’s hands. Those consequences are locked in the moment you choose, long before you’ve sold anything.
This guide walks through the twelve patterns that cover almost everything, what each costs to run when the thing you sell is physical, and how to tell which one you’re already in. If you’d rather see the same material as named companies with real numbers attached, the real-world business model examples cover that side.
The Bottom Line
- A business model answers who pays, for what, and how value reaches them. A revenue model only answers the pricing part, which is why “we’re a subscription” is a half-answer.
- Twelve patterns cover almost everything a product business does: retail, manufacturing, wholesale, direct-to-consumer, subscription, membership, marketplace, aggregator, franchise, freemium, advertising and razor-and-blade.
- Most small product businesses end up running two patterns at once, usually wholesale plus direct. That’s a hybrid, and it’s normal rather than a sign of indecision.
- The model decides your cost per order before you make a single sale. Wholesale ships pallets to a handful of addresses; direct-to-consumer ships one box to a thousand addresses, on the same revenue.
- Choose for survival, not for novelty. Of US establishments that opened in the year to March 2020, 80.9% were still trading after one year and 51.3% after five (Bureau of Labor Statistics).
Lower your delivery costs by 23%
How we reduce costs:
- No delivery vehicle expenses
- Optimized local routes
- Pay-per-delivery model
- Average 23% delivery cost reduction
What a business model actually is
The textbook definition is that a business model describes how a company creates, delivers and captures value. Three verbs, and businesses tend to be strong on the first and vague about the other two.
A practical version: a business model is the set of answers to six questions.
- Who is the customer, specifically enough that you could name ten of them?
- What problem do you solve, and what do they currently do instead?
- What do they pay, and on what trigger: per item, per month, per use?
- How does your product physically reach them? (If the honest answer is “somehow”, that’s the gap: business process modeling is how you turn “somehow” into a drawing of the actual steps.)
- What does it cost you to serve one customer once?
- What do you need to own to make that repeatable: a kitchen, a van, a supplier contract, a customer list?
Answer those six and you’ve described your model, whatever label you give it. Skip the fourth and fifth and you have a pitch, not a model. Those two are where most product businesses discover their chosen pattern doesn’t survive contact with shipping costs.
Business model vs revenue model: the difference that trips people up
A revenue model is one component of a business model: the mechanism by which money arrives. Subscription, one-time purchase, commission, licensing fee, advertising. Those are revenue models.
The business model is the whole machine around it. Two bakeries can both run a subscription revenue model and have completely different business models: one bakes in a commercial kitchen and drops boxes on 200 doorsteps every Friday, the other licenses its recipes to cafés who bake on site. Same revenue mechanism. Different customers, different costs, different everything.
This distinction matters when you’re choosing. “Let’s add a subscription” is a revenue model change and it’s usually easy. “Let’s sell direct instead of through distributors” is a business model change and it touches your packaging, your fulfillment, your customer service and your cash cycle at once. Most of what makes that switch survivable now is technology rather than strategy, which is the subject of our rundown of digital business trends and ideas.
Before you change models, work out what the change is likely to return. Mapping the work first is the cheapest way to find out, and the benefits of process modeling sets out what that exercise usually surfaces.
The 12 types of business models, compared
Here’s the whole set at a glance before the detail. “Who pays” is the party that hands you money, which isn’t always the person using the product.
| Business model | Who pays you | What you hold | Cash timing | Fulfillment load |
|---|---|---|---|---|
| Retail / reseller | End customer | Finished inventory you bought | Paid at sale | Medium, many small orders |
| Manufacturer / maker | Retailers or end customers | Raw materials and WIP | Varies by channel | Medium to high |
| Wholesale / distribution | Other businesses | Bulk inventory | Net 30 to net 60 | Low, few large drops |
| Direct-to-consumer | End customer | Finished inventory | Paid at sale | High, one box per order |
| Subscription | End customer, recurring | Enough stock for a known cycle | Paid before you ship | High but predictable |
| Membership | Member, for access | Little or nothing | Paid up front | Low |
| Marketplace | Both sides, via commission | Nothing | On transaction | None, or coordination only |
| Aggregator | End customer | Nothing; you contract supply | On transaction | You own the promise |
| Franchise | Franchisee, fees and royalties | A system and a brand | Up front plus ongoing | None, franchisees carry it |
| Freemium / SaaS | The minority who upgrade | Software and servers | Monthly or annual | None |
| Advertising | Advertisers | An audience | Monthly, in arrears | None |
| Razor-and-blade | End customer, repeatedly | Two linked products | Thin now, better later | High, consumables repeat |
The patterns in the top half of that table are where most food, floral, catering and wholesale businesses live. The bottom half is where the famous examples come from, which is why generic business model guides feel oddly irrelevant to anyone who makes a physical thing.
Two variants sit outside the table and are worth naming. A hybrid combines two or more rows, usually wholesale plus direct, and it’s the most common real-world answer. And a donor-funded model, which is what nonprofits run, separates the payer from the beneficiary entirely: the person who gives the money is not the person who receives the goods, which is why these organisations manage relationships through donor management software rather than a sales pipeline. If you supply food to a charity programme, you are serving that model even if you don’t run it.
Product models: retail, manufacturing, wholesale and direct-to-consumer
These four are the backbone of any business that sells goods. They’re distinguished by what you make versus buy, and who you sell to.
The retail or reseller model
You buy finished goods and sell them at a markup. Margins are set by what you can negotiate on the buy side and defend on the sell side, so the whole game is purchasing and turns.
It’s the most common pattern in business and the least differentiated. A reseller’s advantage has to come from somewhere other than the product: location, range, service, speed, or the fact that you’ll bring it to the customer when the competition won’t.
The manufacturer or maker model
You convert inputs into something worth more than the inputs. A bakery, a roaster, a jam producer, a candle maker. Your margin comes from the conversion, which means you own both the quality upside and the spoilage downside.
Makers nearly always face a channel decision rather than a model decision: sell your own output direct, sell it wholesale, or both. That decision is usually what people actually mean when they say they’re “choosing a business model”.
It’s also the model with the heaviest equipment bill, so the choice is partly a financing question. Ovens, cold storage, packing lines and vehicles are capital spending rather than running cost, and our guide to the types of capital expenditure explains how that distinction changes both your tax position and your cash planning.
The wholesale and distribution model
You sell in bulk to other businesses who resell. Order values are large, order counts are small, and your customer is a buyer with a spreadsheet rather than a shopper with a craving.
It’s also the largest channel in the economy by volume: US merchant wholesalers turned over about $840.1 billion in a single month in September 2026 (Census Bureau Monthly Wholesale Trade data).
Two things define wholesale economics. Margins per unit are thin, and you get paid on terms. Net 30 is standard, which means you’ve financed your customer’s inventory for a month. In exchange, fulfillment is cheap: a week’s revenue might leave your building on four pallets to four addresses.
The direct-to-consumer model
You sell your own product straight to the person using it, through your own website, your own counter, or your own delivery. You capture the retail margin instead of handing it over, and you own the customer relationship and the data that comes with it.
The channel is still growing faster than retail as a whole. E-commerce reached 17.9% of total US retail sales in the third quarter of 2026, up 10.4% year over year against 5.6% for retail overall (US Census Bureau). Selling direct online has its own sub-taxonomy, covered in our breakdown of the types of ecommerce business models.
What you also own is the fulfillment. The same revenue that moved on four pallets in wholesale now moves in six hundred individual boxes, each needing a pick, a pack, an address and a successful handoff. That cost is real and it’s the single most common reason a DTC launch underperforms its spreadsheet. Before committing, map the order journey end to end, so the hidden steps show up while they are still free.
Recurring-revenue models: subscription, membership and razor-and-blade
These three all sell the future rather than a single transaction, and all three live or die on retention rather than acquisition.
The subscription model
The customer pays on a cycle and receives on a cycle. Coffee every two weeks, flowers monthly, a lunch plan every weekday.
For a product business the attraction isn’t the revenue, it’s the forecast. Knowing you have 340 deliveries on Friday changes how much you buy, how many hands you roster and how tight your route can be.
The catch is the inverse: a subscription promises a date, and a missed date doesn’t just annoy a customer, it cancels a contract. Churn is the number to watch, and it’s unforgiving: e-commerce subscriptions lose about 4.25% of subscribers a month on current benchmarks, roughly a third of which is involuntary and caused by failed payments rather than unhappy customers (Recurly Research). Subscriptions move delivery reliability from a nice-to-have to the core of the product. The other half of running one is billing: recurring charges, pauses, skips, failed cards and dunning all have to work without manual intervention, and buyers’ guides to subscription and SaaS billing software are a reasonable place to understand what that layer has to do before you choose one.
The membership model
Members pay for access, status or price rather than for a specific item: a warehouse club, a buyers’ club, a standing trade account with preferential pricing. The fee is high-margin because it has almost no cost of goods behind it.
The razor-and-blade model
Sell the durable thing cheap, make your money on the consumable that follows. Printers and cartridges, coffee machines and beans, dispensers and refills.
For local producers this works better in reverse than people expect: sell the equipment or starter kit at cost to a café, and own the ongoing supply. You’ve converted a one-off sale into a standing order, which also means a standing route.
Platform models: marketplace, aggregator and advertising
Platform models don’t sell a product. They sell access to a transaction, and they’re the reason most business model lists are full of software companies.
The marketplace model
You connect buyers to sellers and take a commission. You hold no inventory, and your real asset is liquidity: enough sellers that buyers find what they want, enough buyers that sellers stay.
Marketplaces are brutal to start because you need both sides at once, and extremely durable once started, which is why the winners are so large. Third-party marketplaces now account for 83.4% of global e-commerce gross merchandise value, up from 81.0% in 2023, leaving own-store sales with the remaining 16.6% (ECDB). As a seller, that’s the arithmetic you’re facing when you decide whether to list or to build your own store.
The aggregator model
Similar, with one important difference: the aggregator owns the customer promise. Buyers transact with the aggregator’s brand, and suppliers are interchangeable behind it. It’s a stronger position than a marketplace and a heavier obligation, because when a supplier fails the aggregator eats the complaint.
The advertising model
You build an audience and sell its attention. Free to the user, paid for by advertisers. It needs scale to work, which is why it’s rarely a primary model for a local business, though a well-followed local brand can run a small version of it through sponsored placements and paid partnerships.
Licensing and asset-light models: franchise, freemium and dropshipping
The franchise model
You’ve built something repeatable, and you sell the right to repeat it: brand, playbook, supply relationships, training. Income arrives as an up-front fee plus ongoing royalties, and growth no longer needs your capital because franchisees fund their own units.
It’s a bigger slice of the economy than most people assume: the International Franchise Association projects about 845,000 US franchise establishments in 2026, contributing $558.4 billion to GDP and employing nearly 8.9 million people (IFA).
The requirement is that the thing is documented. You cannot franchise knowledge that lives in your head, which is why the operators who get there have written their processes down long before they thought about selling them.
The freemium and SaaS model
Give a usable version away, charge for the version serious users need. Only a minority of free users ever convert, so the model requires the free tier to be cheap to serve and the paid tier to be clearly worth it. The economics of this one are being redrawn as AI raises both the cost of serving a free user and the capability of the paid tier, which is visible in the current AI adoption statistics.
The dropshipping and on-demand model
You sell the product and someone else holds and ships it. Capital requirements are near zero, and so is your control over the thing that determines whether a customer comes back. It’s a reasonable way to test demand and a hard way to build a brand, because the part of the experience you’ve outsourced (the box arriving on time, intact) is the part customers judge you on.
The fulfillment cost each business model hides
The models above differ far more in delivery cost than in anything else, and that difference rarely shows up in the comparison tables.
Run the same $20,000 month through three patterns and the shape changes completely:
- Wholesale: 10 to 25 drops, mostly palletised, scheduled on the buyer’s receiving window. Cost per dollar shipped is low. Your risk is concentrated in a few accounts.
- Direct-to-consumer: hundreds of individual orders, each with its own address, time expectation and chance of a failed first attempt. Cost per dollar shipped is the highest of the three, and customer expectations are set by national retailers.
- Subscription: similar order count to DTC, but known in advance. Same cost drivers, far better route density, and a date you’ve contractually promised.
Geography does as much to that arithmetic as sales volume does. Three hundred orders inside a dense metro, where local delivery in Chicago might mean twelve stops a mile, is a different job from three hundred spread across three counties, even though the invoice total is identical.
A hybrid of wholesale and direct is the most common real-world answer, and it’s the most operationally awkward, because one chain wants pallets and forklifts and the other wants parcels and doorbells. Businesses that run both well usually separate the two flows rather than trying to make one process serve both. Whichever mix you land on, the routing and scheduling layer is doing more work than the pricing layer.
How to choose a business model for a product business
Work through it in this order, because each answer constrains the next.
- Start from the customer you can reach repeatedly. Not the biggest market: the one you can actually get in front of this quarter. Wholesale means finding buyers; direct means finding shoppers. They’re different jobs with different skills.
- Check the cash cycle against your bank balance. Wholesale pays in 30 to 60 days and you’ll fund production in the meantime. Subscriptions pay before you ship. If cash is tight, that gap decides more than margin does.
- Price the fulfillment before the product. Work out, roughly, what one order costs to get to one customer in each model. Include packaging, labour, mileage, failed attempts and returns. This is the number that kills models.
- Ask what happens at ten times volume. Some patterns get cheaper per unit as they grow (wholesale, subscription routes). Others get linearly more expensive (unplanned same-day DTC). The second kind can still be right, but you should know you’ve chosen it.
- Then write the process down. A model is a claim about how work will flow. The process is the flow itself, and writing it down is what turns the claim into something you can staff and cost.
Can you run more than one business model at once?
Yes, and most established product businesses do. A roaster sells wholesale to cafés, direct through its own site, subscriptions to households, and takes a markup on branded equipment. That’s four patterns under one roof.
The practical limit isn’t strategic, it’s operational. Each model adds a different promise (a receiving window, a shipping cutoff, a delivery date), and promises have to be kept by the same staff, in the same building, on the same day. Businesses that add models faster than they add process are the ones where Friday becomes a crisis.
Two rules keep it manageable. Keep the flows separate where they physically differ, and keep the promises honest where they overlap. Everything else is detail.
Common questions about business model types
What are the main types of business models? Retail, manufacturing, wholesale, direct-to-consumer, subscription, membership, marketplace, aggregator, franchise, freemium, advertising and razor-and-blade. Most lists are variations or combinations of these twelve.
Which business model is most profitable? As a percentage, the asset-light ones (franchise royalties, memberships, software), because there is no cost of goods. For a business that makes a physical product, direct-to-consumer usually has the best gross margin and the worst fulfillment cost, so the net answer depends entirely on your order density.
Can a business change its model? Regularly, and usually by addition rather than replacement. The common path is wholesale first for volume, then direct for margin, then subscription for predictability. Each addition is a real operational change, not a pricing announcement.
What’s the simplest model to start with? Whichever one matches a customer you can already reach. For most makers that’s selling direct at a market or through a handful of local accounts, because both give you cash and feedback in the same week.