Pattern · Marketing & customer
Free and Freemium
Also known as Free as a business model, Freemium, Advertising-funded free. Anchored to Chris Anderson set out the family; the freemium variant was named by Jarid Lukin in a comment on Fred Wilson's blog., 2009.
Where this is contested
There is no single originator and the two halves of this pattern have different histories. Free funded by a third party is as old as the advertising-supported newspaper. Free as a deliberate customer-acquisition mechanism, where a proportion of non-paying users are expected to convert, was named freemium in 2006, in a comment on a Fred Wilson blog post, by Jarid Lukin of Alacra. Anderson's 2009 book assembled the family and gave it its economics. The entry is anchored to Anderson because his is the citable treatment, with the naming credited where it belongs.
Giving something substantial away at no charge so that somebody else pays for it: an advertiser, a subset of users who upgrade, or the same user later, which works when the free thing costs almost nothing to reproduce and fails quietly when it does not.
Plate · The shape
How it works
There are three distinct engines here and they are constantly confused with one another, usually by people who have adopted one and are quoting evidence from another.
In the advertising version, a third party pays for the attention of the people getting the thing free. The product is genuinely free to the user and the business is really in the audience trade, which means it is a multi-sided platform wearing different clothes.
In the freemium version, a free tier is a marketing cost carried in the hope that some proportion of users upgrade. Conversion is typically low, in the low single figures as a percentage, so the arithmetic only works if the free tier is nearly costless to serve and the paying tier is worth a great deal. Both conditions have to hold. One alone produces a business that grows enthusiastically and never makes money.
In the cross-subsidy version, one thing is free and a connected thing is charged for, which is the family the bait and hook pattern belongs to.
Anderson's underlying claim was about marginal cost. Where reproduction and distribution cost approaches zero, price tends towards zero, so a business that clings to charging for the reproducible thing will be undercut by one that gives it away and charges for something else. That argument is strongest in software and information and gets weaker with every physical component involved.
The parts of the model
The free thing
Substantial enough to be genuinely useful on its own. It has to solve a real problem completely, or it is a trial rather than a free tier and will be treated as one.
Signals
A user could stay on it indefinitely and still be glad they found it · It is recommended by people who have never paid · It is not time-limited, because a time limit makes it a trial
The wall
The boundary between free and paid. Everything depends on where it sits: too generous and nobody upgrades, too mean and nobody stays long enough to want to.
Signals
The reason to upgrade is a natural consequence of success with the free tier, such as more usage, more people or more responsibility · The wall is not a feature the user needed on day one · Users describe upgrading as the obvious next step rather than as being blocked
The payer
Whoever actually funds the model: an advertiser buying attention, a converting minority, or an institutional buyer paying for what individuals use free.
Signals
You can name the payer in one sentence without hedging · The payer's willingness to pay rises as the free population grows · Revenue is concentrated in a small proportion of the total user base
The cost to serve the free tier
The number that decides whether any of this works. Support, storage, compute, moderation and the human attention free users consume.
Signals
Cost per free user is known to the penny rather than assumed to be negligible · Support load from free users is measured separately from paid · The business could absorb a tenfold increase in free users without a funding conversation
How you know you are in it
You do not fill a pattern in. You recognise yourself in it, or you do not. Read these as a list about your own business rather than as a definition.
- You have a large and enthusiastic user base and a revenue line that does not reflect it, and the plan for closing the gap is that conversion will improve.
- The upgrade prompt appears at a moment of frustration rather than at a moment of success, and your conversion rate reflects that.
- Support tickets from people paying you nothing outnumber those from people paying you something, by a lot.
- You cannot say what a free user costs you per month, and the last time anyone tried the answer was described as basically nothing.
- The free tier was set at its current generosity by an early decision nobody has revisited, and quietly reducing it now would cause an outcry.
- Your competitors are free too, and none of you can remember agreeing to that.
The numbers that decide it
- Conversion rate from free to paid, tracked by cohort rather than in aggregate, because a blended figure hides the fact that conversion usually happens early or never.
- Cost to serve a free user per month, fully loaded with support and infrastructure. Multiply by the free population and compare to gross profit from the paying minority. That single comparison decides whether the model is a business or a hobby with users.
- Lifetime value of a converted user against the total acquisition cost of everyone who did not convert. This is the real acquisition cost and almost nobody calculates it that way.
- The ratio of free to paid users the model can sustain. Every freemium business has a break-even ratio, and knowing it turns generosity from a matter of taste into a decision.
- Referral rate from free users, which is the main way the free tier earns its keep beyond conversion. If free users do not bring others, the free tier is a cost with no return other than eventual upgrades.
When this shape works
- Marginal cost of serving one more free user is genuinely close to zero, which in practice means digital, self-serve and low-support.
- The paying tier is worth a lot per customer, so a low conversion rate still produces a real business.
- Value grows with use, so the natural moment to upgrade arrives from the user succeeding rather than from being obstructed.
- The free tier does work you would otherwise pay for: distribution, word of mouth, or building a habit that makes switching unattractive later.
And when it doesn't
- Serving a free user costs real money, in time, materials, support or licensed third-party capacity. This is why free rarely works for services delivered by people.
- The paying tier is cheap, since a low conversion rate on a low price produces revenue that never catches the cost of the free population.
- There is no natural upgrade trigger, so the wall has to be an artificial obstruction, which users experience as a bait and resent accordingly.
- The business needs revenue soon. Free models are slow, and the gap between growth and revenue is funded by somebody, which for an owner-managed firm means the owner.
How this pattern dies
Growth without revenue
The user chart goes up and to the right for two years and the revenue chart does not follow. Everyone points at the user chart. The underlying arithmetic never worked, and it was visible from the first cohort if anyone had looked at conversion by cohort rather than at the total.
The wall in the wrong place
Set too generously, the free tier serves the whole need and nobody upgrades. Set too meanly, users never reach the moment where paying makes sense. Both errors look identical in the conversion number and require entirely opposite fixes, which is why they are so often fixed the wrong way.
The free tier that cannot be withdrawn
Years in, the free tier is unprofitable and cannot be reduced without a public row, because users who paid nothing have nevertheless acquired an expectation. Anyone who has watched a software company narrow its free plan knows how this goes. The lesson is to be less generous at the start than feels comfortable.
Free by imitation
The whole category gives its core product away because a funded competitor started it, and nobody in the market can now charge for the thing customers most value. Everyone competes on the periphery. This is a market-level failure that no individual firm can escape alone.
Support as the hidden cost
The infrastructure genuinely is nearly free and the humans are not. Free users ask questions, and a support function sized for the free population is the cost that turns a viable model into an unviable one. It rarely appears in the original model.
In the wild
The Metro newspaper
UK, and the purest surviving advertising-funded free product: free to the reader on the train, funded entirely by advertisers buying that commuting audience. The economics of a free newspaper are the economics of an audience business, and the paper is the delivery mechanism.
Spotify
The canonical freemium case and an honest illustration of the difficulty: an advertising-supported free tier feeding a subscription tier, with royalty costs that do not fall to zero however many free users there are, which is why the model took so long to reach profitability.
The accountancy firm's free first meeting
Owner-managed, UK, and the version most small businesses actually run. It works where the meeting is short, standardised and genuinely useful, and it fails where it becomes an hour of unpaid advice given to people who were never going to buy. The wall belongs at the point where advice becomes specific to that business.
The Strategy Toolkit's own Tier 0
Closer to home, and worth stating plainly. Ninety-two entries, working templates and no sign-up, with the gate placed at the analysis engine and playbook runs. The free tier is genuinely complete for a reader who only wants the reference, and the wall sits where the value becomes personal to their situation.
The owner-managed version
Most owner-managed businesses cannot run this pattern, and the ones that try usually do so by accident. The reason is simple: your free thing is almost always your time, and your time has a real cost that does not fall as more people take it.
So the test I would apply is whether the free thing is reproducible. A guide, a template, a calculator, a recorded assessment, a diagnostic that runs itself. Give those away by all means, and give them away properly rather than as a teaser, because a mean free offer damages you more than no free offer. A free thing that solves someone's problem completely is the best marketing most small firms will ever do.
What you should not do is give away the thing that costs you an afternoon. The free consultation, the free audit, the free proposal that takes two days to write. Those are not a free tier. They are unpaid work with a hopeful expression on its face, and the people most likely to accept them are the people least likely to buy.
Where the wall belongs is easier to see than people think. Free up to the point where the advice becomes specific to that business. General is free, specific is paid, and customers understand that boundary immediately because it matches how they think about it anyway.
One last thing. Be less generous at the start than feels comfortable. You can always add. Taking something back from people who have got used to having it for nothing is one of the least pleasant conversations in business, and it does not get easier with time.
Changing out of this shape
Moving into this pattern deliberately means fixing the arithmetic before the generosity. Calculate the fully loaded cost of serving a free user, decide the free-to-paid ratio the business can carry, and set the wall from that rather than from what feels fair. Place the wall at a success moment rather than an obstruction, which usually means it triggers on volume, on additional people or on responsibility rather than on a feature. Then track conversion by cohort from the first week, since a blended rate will hide a failing model for a year. Getting out is the harder direction: reducing an established free tier is possible but should be done by grandfathering existing users and applying the new boundary only to newcomers, which costs revenue and preserves the goodwill the free tier was bought with.
Where to go from here
Frameworks that work inside this pattern
- Value-Based Pricing
A pricing discipline that anchors price to the economic value an offer creates for a defined segment, with cost setting only the floor, and that tests willingness to pay before launch using tools such as Van Westendorp's Price Sensitivity Meter and Good-Better-Best tiering.
- Unit Economics
Tests whether a business makes money on each unit sold and each customer acquired, tracing revenue through variable costs to contribution and comparing the cost of acquiring a customer with the lifetime value they return. If the unit loses money, scale multiplies the loss.
- AIDA
The oldest working model in marketing: move a prospect through Attention, Interest and Desire to Action. Empirically shaky as a theory of buying, yet still the most useful checklist ever written for auditing an advert, a landing page or a sales pitch.
- Crossing the Chasm
Moore's account of why technology products stall between visionary early adopters and pragmatist mainstream buyers, and the beachhead strategy for getting across: pick one narrow segment, build the whole product it needs, and win the references pragmatists actually trust.
- Jobs to be Done
A lens for understanding demand: customers hire products to make progress in a specific circumstance, and whether they switch is governed by four opposing forces. Study the job and the forces around it rather than the customer's attributes or the product's features.
- Net Promoter Score
A customer loyalty metric built on one question, how likely are you to recommend us, scored 0 to 10. Subtract the percentage of detractors (0 to 6) from the percentage of promoters (9 and 10) for a single figure that is easy to track, easy to compare, and easy to abuse.
Playbooks that work the problem
- Pricing with Confidence
A sequence for putting a number on your work without flinching: find the job customers actually hire you for, price the value rather than the hours, check the volume the price must sustain, then imagine the new price list failed and find out why before it does.
- Launching a Product
A launch sequence built on one hard-won lesson: work out what customers would actually hire your product to do before your enthusiasm spends the budget. Job, proposition, beachhead, then a rhythm of cheap tests once you are live.
Neighbouring patterns
- Bait and Hook
Sell the durable thing cheaply, sometimes below cost, and make the money on whatever the customer must keep buying afterwards, which works precisely as long as you can stop anyone else supplying the refill.
- The Multi-Sided Platform
A business that makes money by bringing two groups together who each need the other, where the decisive question is not what to charge but which side to charge, because the side you subsidise is the side that makes the whole thing work.
- Subscription and Recurring Revenue
Charging a regular fee for continuing access rather than a one-off price for a transaction, which converts selling from an event into a relationship and converts profit from something you earn once into something you have to keep deserving.
What the critics say
Gladwell's review of Free attacked its central move directly: Anderson treats near-zero marginal cost as though it were near-zero total cost, when the fixed costs of creating the thing being given away are exactly the costs that have to be recovered. The businesses Anderson cited as proof were in most cases either loss-making or funded by someone else's capital at the time he cited them.
Gladwell, M. (2009) 'Priced to Sell', The New Yorker, 6 July.
Freemium conversion rates are far lower than founders assume, typically in the low single figures as a percentage, which means the model requires either an enormous free population or a very high price on the paid tier. Kumar's analysis found that most freemium failures were not marketing failures but arithmetic ones, set at the point the free tier was designed.
Kumar, V. (2014) 'Making Freemium Work', Harvard Business Review, 92(5), pp. 27-29.
Free is not a neutral price. Behavioural work shows that moving a price from a very small amount to zero changes decisions disproportionately, which cuts both ways: it drives adoption far beyond what the price change would predict, and it attracts a population whose intent to pay is materially lower than a paying trial would produce. Treating free users as prospects with a discount overstates their value.
Shampanier, K., Mazar, N. and Ariely, D. (2007) 'Zero as a Special Price: The True Value of Free Products', Marketing Science, 26(6), pp. 742-757.
Sources and further reading
- Anderson, C. (2009) Free: The Future of a Radical Price. New York: Hyperion.
- Shampanier, K., Mazar, N. and Ariely, D. (2007) 'Zero as a Special Price: The True Value of Free Products', Marketing Science, 26(6), pp. 742-757.
- Kumar, V. (2014) 'Making Freemium Work', Harvard Business Review, 92(5), pp. 27-29.
- Osterwalder, A. and Pigneur, Y. (2010) Business Model Generation. Hoboken, NJ: John Wiley and Sons, pp. 88-107, which treats free as a business model pattern and credits Anderson.