Reviews
Ideas by profit model: a practical reference for 2027
Ideas by profit model compared as structures: eight ways to be paid, what cash timing does, what breaks under growth, and where margin actually comes from.
Two businesses can sell the identical thing to identical customers and be completely different to own. The difference is how the money is structured: when you are paid relative to when you spend, whether the same customer buys again, what happens to costs when volume rises, and whether the whole arrangement depends on you personally.
That structure is the profit model, and it determines more about daily life than the industry does. This page is a reference to the main models, what each one requires to work, and where each one hurts.
What to take away
- Most real businesses use one primary model with a secondary one attached.
- Each model has a different failure point under growth, and knowing yours tells you what to prepare for.
- Margin is usually treated as a pricing question.
- The best model on paper is worthless if it requires work you will avoid.
The eight structures almost everything reduces to
| Model | You are paid for | It works when | Where it hurts |
|---|---|---|---|
| Time | Hours, days, or sessions | Your hours are scarce and buyers value them highly | Income is capped by the calendar; efficiency reduces your earnings |
| Fixed-scope project | A defined outcome, priced up front | You can estimate the work accurately and hold the scope | A misjudged estimate is absorbed entirely by you |
| Retainer or subscription | Continuing access, availability, or a recurring service | The need genuinely recurs on a schedule | Customers leave quietly; value must be visible or it feels optional |
| Margin on goods | The difference between buying and selling | You can source well and move stock reliably | Cash sits in inventory; unsold and returned goods eat the margin on sold ones |
| Commission or marketplace | A share of transactions between others | Both sides need each other and cannot easily meet directly | Nothing works until both sides exist; participants route around you |
| Advertising or sponsorship | Attention you have assembled | The audience is large or unusually specific | Revenue arrives long after the work; you serve two masters at once |
| Licensing or royalty | Someone else's use of what you made | The thing is genuinely reusable and legally yours | Control sits with the licensee, and enforcement is your problem |
| Productised service | A standardized package delivered repeatedly | The work is similar enough each time to define tightly | Any customer who needs an exception breaks the economics |
Most real businesses use one primary model with a secondary one attached. The primary model is the one that would still exist if you removed the other.
Cash timing decides how it feels to run
This is the most practical distinction on the page and it gets less attention than it deserves. Profit and cash are different, and cash is what ends businesses.
Paid before you spend. Deposits, pre-orders, retainers, subscriptions billed in advance, memberships. The customer funds the work. This is the easiest position to grow from, because expansion generates cash instead of consuming it, and it is worth designing toward even when it means charging less.
Paid as you spend. Point of sale, immediate service, most consumer transactions. Comfortable and simple.
Paid after you spend. Invoiced work with payment terms, wholesale, stock bought before it sells, anything delivered before payment. Growth here consumes cash: the more you sell, the more money is tied up before it returns. Businesses fail in this position while profitable, which is a genuinely confusing way to fail.
If your model is the third kind, the levers are known and worth using early, because they are far easier to introduce at the start than to add to an existing relationship. Deposits on acceptance. Staged payments against milestones. Shorter terms. Stopping work on overdue accounts. Smaller initial stock orders even at a worse unit price.
Which models buy you a job
A business that runs without you needs three things: work that can be done by someone other than you, demand that arrives without you generating each instance of it, and revenue that continues through a period when you are absent.
Score the models against that. Time-based work fails all three. Fixed-scope project work fails the second and third unless a pipeline exists. Retainers and subscriptions pass the third, and pass the first if the service is documented. Goods and marketplaces can pass all three, at the cost of capital or of a hard start. Licensing and advertising pass all three once established, and take the longest to get there.
None of this makes time-based work a bad choice. It is the fastest to start, needs the least capital, and provides feedback immediately. The error is expecting it to become something else while changing nothing. Moving from selling hours to selling an outcome, then a package, then something the customer uses without you, is a deliberate sequence of decisions. It never happens by working harder at the current rung. The ideas by skill guide walks through that progression.
Repeat purchase changes everything
The single most useful question about any model: does the same customer buy again, and roughly how often?
If they do, every customer you win adds to a base, and your effort compounds. You can afford to spend more to win each one. Slow months are cushioned. Your reputation has somewhere to accumulate.
If they do not, every month starts empty. You are permanently selling, and the cost of finding customers is a permanent cost rather than an investment. Some excellent businesses work this way (high-value one-off work, services tied to rare events), but they need a channel that reliably produces strangers, and that channel is the whole business.
Where a one-off model can be given a recurring element, it is usually the highest-return change available. Maintenance after installation. Servicing after a sale. Consumables after equipment. A support arrangement after a project. The local service guide works through that conversion in trades where the customer is nearby and the visit recurs. The best moment to offer it is immediately after delivering something the customer is pleased with.
Recurring revenue has its own discipline, though. Subscription and retainer arrangements decay quietly: customers stop seeing the value, then leave, and you find out a month later. The work of making the ongoing value visible is real work, and it is why some retainers survive for years and identical ones do not.
What breaks when volume rises
Each model has a different failure point under growth, and knowing yours tells you what to prepare for.
Time-based work hits your calendar. The only responses are higher prices, faster delivery, or other people's hours, and the third changes the business into something you manage rather than do.
Fixed-scope work hits estimation. What worked across a handful of projects becomes a problem across many, because the losses on badly estimated jobs stop averaging out and start compounding.
Recurring services hit churn. Growth and departure run at the same time, and past a certain size you are refilling a bucket rather than adding to it.
Goods hit working capital and logistics. More sales means more money tied up, more storage, more returns, more that can go wrong between order and delivery.
Marketplaces hit balance and disintermediation. One side outgrows the other, and participants who met through you decide to deal directly.
Advertising and audience models hit relevance. The audience that grew around one thing does not necessarily want the next thing.
Licensing hits control. Your revenue depends on decisions made by people whose priorities are not yours, and the arrangement only exists if what you made is legally yours to license in the first place. The US Copyright Office's account of what copyright covers and who holds it is the place to start on that question.
Who your customers are matters as much as how many
Two businesses with the same revenue can carry entirely different risk, and the difference is concentration.
A venture where most of the income comes from one or two customers is not really a business yet. It is an employment arrangement without the protections, and the customer knows it. They can dictate terms, delay payment, and change direction, and losing them means starting over. This happens most often to service businesses that grow through one good relationship, and it feels like success right up until it does not.
A venture where income is spread across many customers is more stable and more work. Nobody can end you, but you are permanently selling, and the cost of finding customers never goes away.
Neither extreme is correct, and the useful habit is simply to know where you sit and to watch it. If a single customer accounts for a large share of your income, treat winning the next unrelated customer as more urgent than serving the existing one better. If your largest customer could leave without changing your life, you have bought that safety with continuous selling effort, which is a real cost.
The same logic applies to suppliers, to platforms, and to any single channel that produces most of your enquiries. Concentration anywhere hands somebody else a decision about your business.
Where margin actually comes from
Margin is usually treated as a pricing question. It is mostly not.
Price is one lever, and the one people reach for last because raising it is uncomfortable. It is also the fastest, since an increase falls almost entirely through to the bottom.
Scope is the lever that silently destroys margin. Work delivered beyond what was agreed is unpaid work, and in fixed-price and productised models it is the single most common reason a profitable-looking arrangement is not. Writing down what is not included is worth more than most cost-cutting.
Delivery cost is where operational improvement lives: doing the same work with less time, less waste, or fewer errors. It compounds quietly and it is invisible to customers.
Customer selection is the lever nobody names. Some customers cost far more to serve than others (more questions, more revisions, more chasing for payment), and they are frequently the ones who negotiated hardest on price. Choosing better customers improves margin without changing anything about the work.
Positioning underlies all of them. A general offer competes with everyone and is compared on price. A specific offer for a defined buyer is compared with fewer alternatives, and that is what makes a better price defensible in the first place.
Mixing models
Combinations are normal and usually good. A service business with a retainer alongside project work. A goods business with a subscription for consumables. A consultancy with a productised entry offer that leads to larger engagements.
They work when the second model serves the same customer with the same underlying capability, and when one clearly leads to the other.
They go wrong in two specific ways. The first is dilution: two models needing different channels, different customers, and different operations, run by someone with the capacity for one. The second is conflict, most often when advertising or sponsorship is bolted onto something the audience trusts. Then your paying customer and your reader are different people with opposing interests, and you will resolve it in favor of whoever pays. Choose knowingly.
Changing model later
It is done all the time, and it is harder than starting with the right one, because customers arrive expecting the arrangement they signed up for.
Moving from hours to fixed prices is the most common and the most rewarding. It requires enough delivered work to estimate confidently, and a scope written tightly enough to defend.
Moving from one-off to recurring requires a genuine ongoing need. Inventing one that nobody feels produces cancellations, not revenue.
Moving from service to product is the biggest change and the most often underestimated. Service revenue is generated by selling to one buyer at a time; product revenue requires reaching many people, which is a different job with a different skill set. Most attempts fail on distribution, not on the product.
Whichever direction you move, existing customers usually keep their old arrangement while new ones get the new one. Forcing everyone across at once is how you lose the base that was funding the transition.
Choosing one you can actually run
The best model on paper is worthless if it requires work you will avoid.
Retainers and subscriptions require ongoing relationship management and a tolerance for being asked things at inconvenient times. Fixed-price work requires firmness about scope, which some people find genuinely difficult. Goods require operational discipline: counting, ordering, chasing, packing. Marketplaces require patience through a long period when nothing works. Audience models require sustained public visibility. Licensing requires attention to contracts.
Match the model to the temperament as well as the market. A business you dislike operating will be operated badly, and that shows up in the numbers long before you admit the cause. Match it to the hours too: the part-time guide sets out which structures survive a week where the venture gets almost nothing.
Testing a model before committing to it
You can test the structure, not just the idea, and it is worth doing separately.
Offer the same thing under two arrangements to different customers and see which they accept more readily and complain about less. Ask for a deposit and see whether it is paid without argument, because willingness to pay in advance tells you a great deal about how much the buyer wants the outcome. Sell one fixed-price version of something you normally bill hourly, and measure what actually happened to your effective rate.
Then look at what people did, not what they said. The federal guidance on understanding a market before committing sets out what this stage is meant to establish.
Where to go next
The profit model framework covers selecting one for a specific venture, examples works through cases, metrics covers what to measure under each structure, and questions answers the recurring ones. For how the model interacts with what you can afford to start with, see the startup budget guide.
Common questions
Which model should a first venture choose?
Usually one paid on or before delivery, with little owed afterwards. It is the fastest to start, it commits you least, and it teaches you the most per month. Duration and recurring revenue can be added once you know what you can reliably promise.
Why am I profitable and short of money?
Because profit ignores timing. If you buy stock or deliver work before being paid, growth ties up more money than it returns, and the faster you sell the tighter it gets. The levers are deposits, staged payments, shorter terms and smaller initial orders, and all are far easier to introduce at the start.
How do I know if I am too dependent on one customer?
If losing them would mean starting over, you are. That situation is an employment arrangement without the protections, and the response is to treat winning the next unrelated customer as more urgent than serving the existing one better.
Can I change model later?
Yes, and it is harder than starting with the right one, because customers arrive expecting the arrangement they signed up for. Keep existing customers on their old terms and sell the new shape to new ones. Forcing everyone across at once loses the base that was funding the change.
