My views on the market, tech, and everything else

The Rise of the Freemium Business Model


Freemium describes a business model in which a business gives one product away for free or at a subsidized price and then either: (1) sells another profitable product to this user base; or (2) sells access to that user base to third parties (e.g., advertisers). Three versions of the Freemium approach are:

  1. Available Forever- No premium versions are made available. Google and Facebook are examples that monetize with advertising.
  2. Premium Freemium: Users pay only for premium versions (e.g., more powerful features or greater use rights). Only the free baseline version is “available forever.” LinkedIn and TurboTax are an examples.
  3. Limited Freemium: The free version is made available but is limited by factors like time and/or capacity. One example would be a 30 day trial with no “available forever” baseline version.

The freemium business model is ancient. A bar giving away salty snacks to sell more  alcohol is nearly as old as the still that created the liquor. The practice of offering tapas in Spain is just one relatively modern example. Newspapers being used to sell political views is another example:

“Gerald J. Baldasty’s book, The Commercialization of News in the Nineteenth Century, makes a case clear as spring water that hard news has almost never been a mass commercial enterprise. The American newspapers of the 1820s and early 1830s were creatures of political parties, edited by zealots. Essentially propaganda sheets, these newspapers were “devoted to winning elections …Many subscribers simply did not pay for their newspapers,” Baldasty wrote.  “In 1832, one North Carolina editor estimated that only 10 percent of his 600 subscribers had paid for the paper.”

Many businesses have offered free services (free month of HBO) or free trials (magazines). The Venture capitalist Tomasz Tunguz of RedPoint has written about why the freemium business model works so well:

“At its core, freemium is a novel marketing tactic that entices new users and ultimately potential customers to try a product and educate themselves about its benefits on their own. By shifting the education workload from a sales team to the customer, the cost of sales can decrease dramatically… freemium startups leverage usage data to improve their product. The large amount of users using the product enables A/B testing with statistical significance, a non-trivial strategic advantage. Marketing teams can sift through the data to understand market segmentation and funnel efficiency, and product management can parse the data to improve the on-boarding experience. Third, freemium startups gather information about their customer base to prioritize their sales efforts.”

The cost of educating the potential customer about the benefits of the product can be dramatically lower once a potential customer has used the product in the free setting. The need for advertising to create awareness is lower and any expenses associated with a paid salesperson or sales support engineer patiently explaining the product to a potential customer has been replaced by self-education. Freemium also leverages other human tendencies like reciprocity and inertia, which further lower the cost of sale. John Vrionis of Lightspeed Venture Partners describes the less costly and more effective freemium sales process:

“Evangelizing a new religion is hard work, and more importantly it is expensive. But that’s essentially what selling proprietary software is like. Conversely, selling bibles to a group of believers is a lot easier… Sales people engage an account when they know that the prospect is well down the path of adoption and belief.”

In some settings the salesperson can be eliminated completely as the customers “upsell” themselves with self-service capabilities of the company web site.

As Tungus and Vrionis note, the essence of the Freemium model is to reduce the price of acquiring a customer. My post “Why is Customer Acquisition Cost (CAC) like a Belly Button?” is an introduction to the topic of CAC. Everyone with a business has CAC.

To illustrate, even comedians must incur costs to acquire customers. As an example, here a famous few sentences from a monologue form the comedian Louis C.K:

“Everything is amazing right now and nobody’s happy. In my lifetime the changes in the world have been incredible. When I was a kid we had a rotary phone. We had a phone you had to stand next to and you had to dial it. And then when, if you wanted money you had to go in the bank for when it was open, for like three hours.”

Those four sentences solidified Louis C.K.’s reputation as an A-List comic. The comedian’s monetization was indirect as is often the case today:

“Comedians today are more likely to gain fame via a YouTube skit or a bit that plays well on a talk show. (Louis C.K.’s most famous monologue, “Everything’s amazing and nobody’s happy,” happened in a sit-down interview with Conan O’Brien).”

Since Louis C.K. is a member of the Screen Actors Guild he would have received at least the union minimum wage to be on O’Brien’s show, but that is a tiny fraction of the value that monologue created for his career. Louis C.K. essentially gave away that monologue for free. In this case one activity (the Conan O’Brien appearance) is being used as a substitute for customer acquisition cost (CAC) for other products like his concerts. This cross promotional approach is not new. Especially for comedians. Not too long ago they generated most of their profit from albums (promoted on talk shows) and concerts (Johnny Carson used to tell people their dates).

Are other comedians using freemium like Louis C.K.? Yep. But the loss leader is not always a zero revenue gig, just lower revenue.

“Comics [have] realized how a constant presence on Netflix, as opposed to sporadic broadcasts on linear television, could help build a larger loyal audience. For most comedians, specials are a means to boosting their key revenue source: regular ticket sales on the road. TV specials have always translated into additional fans showing up on tour stops, but the lift from a Netflix special exceeded what a lot of comics were seeing from regular TV. “The amount of time it took to see the bump was drastically reduced from years to … six or seven months after an artist’s special is on Netflix,” says Volk-Weiss. “Their touring business goes up by hundreds of [percent].”

Wait!” You are probably saying, “People have done free promotions forever.” That’s true, but with the digitalization and networking of the world this freemium phenomenon has been put on steroids. Ben Thompson succinctly describes one root cause of why the Freemium phenomenon is on the rise as a customer acquisition technique as the world become more digital and networked:

“The defining characteristic of anything digital is its zero marginal cost. Take apps for example: What makes the software market so fascinating from an economic perspective is that the marginal cost of software is $0. After all, software is simply bits on a drive, replicated at the blink of an eye. Again, it doesn’t matter how much effort was needed to create said software; that’s a sunk cost. All that matters is how much it costs to make one more copy – $0. The implication for apps is clear: any undifferentiated software product, such as your garden variety app, will inevitably be free. This is why the market for paid apps has largely evaporated. Over time substitutes have entered the market at ever lower prices, ultimately landing at their marginal cost of production – $0.”

In addition to being reproducible and distributable at virtually zero marginal cost, digital goods typically do not have the two qualities described here by Brad DeLong:

“Excludability: the ability of sellers to force consumers to become buyers, and thus to pay for whatever goods and services they use.

Rivalry: a structure of costs in which two cannot partake as cheaply as one, in which producing enough for two million people to use will cost at least twice as many of society’s resources as producing enough for one million people to use.”

If I make a digital copy of your digital music, you still have your music (the music is non-rival). If I steal your phone you will no longer have a phone (a phone is rival). So-called “public goods” are non-rival and non-excludable. A lighthouse is a public good and in many settings so is software. If a business can’t charge customers a fee directly for a good or service since it is a public good giving it away is “sleeves off its vest.” In other words, the real opportunity cost of the sales incentive is zero if the business can’t charge a fee for it anyway.

One confusing element of a freemium business model is the accounting involved. If giving away goods or services for free or with a subsidy is not a marketing expense, what exactly is it? Marketing costs are typically shown below the gross margin line on an income statement. But the cost of freemium offerings can be considered part of costs of goods sold or COGs. Different companies treat freemium costs in different ways. Some companies even split the difference: the cost of non-converted non-paying customers is a marketing expense and the cost of serving paying customers is COGS. In any event, giving away free services is an expense however you look at it.

Small startups with little cash on hand or that desire not to dilute their ownership too much by raising more cash have found freemium to be particularly attractive. If they know how to write software they can use the free service to avoid spending cash on customer acquisition. As an added benefit, customers acquired though freemium tend to value the product more. They churn less from the service and tend to have better credit. Of course freemium can be taken too far, especially if there isn’t a profitable complementary good that can be paired with the service. With internet scale services, while the incremental cost of a new customer/user is not zero (as per Ben Thompson’s description of digital goods) it is effectively zero as there is so little incremental cost. An exception would be highly resource intensive services such as live streaming, email/file storage, or translation all of which use significant bandwidth, storage, or compute per incremental use.

Bill Gurley describes an increasingly important phenomenon facing every business today: “If a disruptive competitor can offer a product or service similar to yours for ‘free’ and if they can make enough money to keep the lights on, then you likely have a problem.” Every businesses today must be prepared for competitors to give away what they sell as an incentive for customers to buy something else. That fact thatmany of the free services are digital and have close to zero marginal costs has caused the freemium strategy to spread like wildfire. The implications of the freemium business model are massive since businesses now face competition from companies that they never previously thought of a competitors. Fred Wilson believes: “when network effects matter, when your users are creating the content and the value, free is the business model of choice.” If the network effects benefits of size are large getting big fast is hard given customers acquisition cost. And the best way to lower customer’s acquisition costs is often free. Making the approach even more attractive is the fact that even free users add to network effects since they are using the systems and its formats.

Using a freemium approach can be tricky in an enterprise setting. Version One ventures points out that freemium can create problems and is not a panacea:

“Offering an app/service for free can send the wrong message. Here, free can be equated with low quality. Free isn’t necessarily sustainable with B2B. Acquiring business users may prove too costly, forcing start-ups to raise incredible amounts of money to finance aggressive sales and marketing efforts for a free app. In the enterprise, freemium models generally work in two situations: (1) You target a large enough user base and (2) The product becomes more valuable over time.”

Freemium works best when there are millions of potential customers since the conversion rate will not be high. Best practices include:

  1. Avoid selecting “free” items that have significant variable COGs. Select software-based free services that have almost zero marginal cost works better.
  2. The best free services have network effects.
  3. Make sure that the item you select as free side of the marketplace has complementary services that might be sold at a profit. Free checking at a bank is good example of free service that makes upselling natural (many types of loans).
  4. The best forms of Freemium make it natural for the users to register revealing their personal data and set up a credit card on file.
  5. The best complementary products are services that people are accustomed to paying real money for (convincing people to spend money in a new category is not easy),
  6. The best free services have low churn since they are sticky)

What sort of conversion rates are typical in a freemium model? One estimate from Totango is:

“For a B2B SaaS company with free trial, I assume the following ball park figures:

3%-5% – average and means good business operations
8% and above – excellent conversion rates
below 3% – below average.”

Nothing creates a base from which to try to up-sell users like a free service.  There is no question that a price of zero is magical for humans. Work by Dan Ariely proves this point:

“In one study we offered students a Lindt Truffle for 26 cents and a Hershey’s Kiss for 1 cent and observed the buying behavior: 40 percent went with the truffle and 40 percent with the Kiss. When we dropped the price of both chocolates by just 1 cent, we observed that suddenly 90 percent of participants opted for the free Kiss, even though the relative price between the two was the same. We concluded that FREE! is indeed a very powerful force.”

Ariely concludes: “‘Free’ is kind of an incredibly tempting human hot button. And sometimes it’s great and sometimes it gets us into trouble.”   

The freemium approach is now scalable globally as never before. The implications of this shift are huge and still being sorted out.

P.s., How much is being given away by producers in consumer settings as a result of the freemium phenomenon? Since free products have no price they are not easy to measure, One indication that the number is large is the fact that 98% of App Store revenue is coming from freemium apps. Services like mobile games and Craigslist have free elements.

  • A Deloitte Report summarizes some of the research in consumer settings:

“Researchers at the University of Michigan found that they saved 15 minutes by using the internet compared to the university library to answer a list of specific questions. Assigning a monetary value to these time savings Google’s Chief Economist, Hal Varian, estimated that the search feature of the internet generates a consumer surplus in the US of $65-130 billion a year. This is likely to significantly underestimate the true level of saving because in this experiment the internet was compared with a university library – an asset that few consumers can access. Another approach is to estimate the time people spend on the internet. The reasoning runs that if I choose to spend an evening searching the web it demonstrates that I prefer this activity to all others available to me. Using this approach Erik Brynjolfsson and Joo Hee Oh at MIT estimated America’s consumer surplus from the internet at $564bn in 2011. Inclusion of this surplus in America’s GDP numbers would have boosted annual growth by a sizeable 0.4 percentage points – equivalent to a 15% uplift in trend growth.  The nature of the internet and the way we interact with it makes estimating the consumer surplus from it incredibly difficult. Crowdsourced websites such as Twitter, Wikipedia, TripAdvisor or the review section of Amazon are particularly challenging. How does one measure the addition to the consumer surplus from our receiving advice, ideas and information?”

These estimates do not include the value of open source and other public goods for businesses such as free cloud compute and storage for commercial customers. For example, every cloud company now has a free tier. Azure also has free credit for new customers and a free service trier.  Amazon has both an always free trier and a first-12-months-free offer. Google also has a free to use tier for its cloud services. In 2015 GetApp, a company that tracks over 3,000 business apps, reported that 22% of the total have adopted a freemium model. Business like DropBox, Slack, MailChimp, SoundCloud, EverNote, ZoHo, SurveyMonkey, Skype, HootSuite, Moz, Weebly, Salesforce, Optimizely, and Box all use a freemium approach. Open source examples include Linux, Git, MySQL, Node.js, Docker, Hadoop, Elasticsearch,  Spark,  MongoDB, Selenium, NPM, Redis,  Tomcat, Jenkins, Vagrant, Postgres, Gradle, Nginx, Ansible,  Kafka, GitLab, and Chef.

Open source software has enabled the creation of many freemium business models. By combining software that is given away as part of an open source community with related proprietary services that are not given away, a new business model is created. The related profitable and proprietary services must “complement” the free service like mustard complements a hot dog for the model to work well. Many entrepreneurs and investors have realized that the best path in creating a new business is to invent an open source project that is widely contributed to and then build a service around it after hiring all the people who worked on it. Businesses that have been built around open source, like Databricks, Mesosphere, and Docker, exemplify this trend.


Louis C.K. https://www.youtube.com/watch?v=q8LaT5Iiwo4

Ben Thompson https://stratechery.com/2014/cost-bitcoin/

Deloitte: http://blogs.deloitte.co.uk/mondaybriefing/2015/09/weve-never-had-it-so-good.html

DeLong  http://delong.typepad.com/hoisted_from_the_archives/2007/11/j-bradford-delo.html

John Vrionis: http://venturebeat.com/2015/12/06/its-actually-open-source-software-thats-eating-the-world/

Jack Shafer: News never made money, and is unlikely to.  http://blogs.reuters.com/jackshafer/2013/08/15/news-never-made-money-and-is-unlikely-to/

Tomasz Tunguz http://tomtunguz.com/when-to-go-freemium/

Version One: http://versionone.vc/freefreemium-work-enterprise/#ixzz4eNwf7zE0

Fred Wilson: http://avc.com/2012/07/in-defense-of-free/

Categories: Uncategorized